Routledge Library Editions: Development, Mini-Set B: Aid 0203840186, 9780203840184, 9780415584142, 9780203840351, 9780415592765, 0709938292

Routledge Library Editions: Development will re-issue works which address economic, political and social aspects of deve

227 55 74MB

English Pages 1179 [1180] Year 2022

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Volume Cover
Volume 7
Cover
Half Title
Title
Copyright
Original Title
Original Copyright
Contents
List of Tables and Figures
Preface
1. THE POLITICAL ECONOMY OF US FOREIGN AID: PAST AND PRESENT
Introduction
The Blurred Lines of Foreign Aid Policy
Contemporary Foreign Aid Programs
The Evolution of US Postwar Foreign Aid
2. CRITICAL PERSPECTIVES ON FOREIGN AID
Introduction
Harsh Evaluations From The Political Spectrum
3. THE BUREAUCRATIC ROLE CONFLICT MODEL
Introduction
The Need For Policy Models
Results of Foreign Aid
Decision Premises or Goals
Constraints: Rules and Repertoires
Opportunities: Role Conflict
Conclusion
4. US AID TO LATIN AMERICA
Introduction
Results
Goals
Constraints
Role Conflict
Conclusion
5. US AID TO ASIA
Introduction
Goals
Results
Constraints
Role Conflict
Conclusion
6. US AID TO THE MIDDLE EAST
Introduction
Results
Goals
Constraints
Policy Actors
Conclusion
7. CONCLUSION
Introduction
Summary and Conclusion
Recommendations
Index
Volume 8
Cover
Title
Copyright
Original Title
Original Copyright
Contents
General editor's introduction
Tables
Figures
Acknowledgments
Introduction
1 Aid: motives
2 Aid: allocation principles
3 Aid and economic growth I
4 Aid and economic growth II
5 Debt and the terms of aid
6 Aid tying: trade and resource allocation effects
7 Some questions and further issues
Appendix: Resource gap models: statistical estimation of the parameters
Further reading
Volume 9
Cover
Half Title
Title
Copyright
Original Title
Original Copyright
Contents
List of Table
Acknowledgements
Introduction
1 The Background
2 Aims and Achievements Since Independence
The Agricultural Sector
Manufacturing
African Participation
The Role of Government
The Current Situation
3 The Facts of Aid
British Capital Aid
British Technical Assistance
The Commonwealth Development Corporation
Aid From Other Sources
The Costs of Aid
4 British Aid Policy and Administration in Kenya
The Early Years, 1963–69
Aid Administration Today
Policy Issues
Other Donors’ Policies
The Commonwealth Development Corporation
5 Aid in Action
Part One: The Land Transfer Programme
1960–65
1966–74
The Effects of British Aid on Land Settlement
The Evaluation of Aid for Land Transfer
Appendix: A Model of Expropriation
Part Two: The Mumias Sugar Company
History of the Project
The Project in Prospect and Practice
The Outgrowers
Mumias and Rural Development
Mumias and ‘Appropriate Technology’
Other Considerations on Mumias as an Aid Project
Part Three: The Special Rural Development Programme
SRDP in Kwale
The Lessons of SRDP for Aid Donors
SRDP and District Development
6 The Influence of Aid
Fungibility
Public Finance
Other Government Policies
Administration
Political Economy
Conclusion: Incomes and Income Equality
7 Policy Implications for British Aid
8 Controversies Over Aid
Index
Volume 10
Cover
Half Title
Title
Copyright
Original Title
Original Copyright
Contents
List of Tables
Preface
Glossary
Introduction
1 India’s Aid Resources in Macroeconomic Context
Introduction
Measurement of aid flows
Size of aid flows to India
Impact of aid on economic growth
India’s future capital inflows: aid or commercial lending?
Conclusions
References
2 Aid and Poverty in India
Dilemmas in using policy against poverty
Scale and structure of aid: impact on poverty
Aid to agriculture: poverty-oriented?
Aid other than to food and agriculture: poverty impact
Learning from India
References
3 Policy Dialogue
Surprising facts and aid processes
From bilateral leverage to multilateral sector dialogue
Before leverage
Towards the crisis of bilateral macro-leverage
Aid 1965–85: towards multilateral sector dialogue
The problem of ‘ideology’ in policy dialogue
Risk of rigidity
Changing fashions
Choosing sectors and topics for dialogue
Can the dialoguer deliver?
References
4 The Systemic Effects of Aid and Donor Procedures
Introduction
Many donors: help or hindrance?
Systemic aspects of multilateral aid
Systemic aspects of bilateral aid
Local cost finance
Donor procedures
References
Chapter 5 Project Aid to India
Role and evaluation of project aid
The World Bank Group’s project sample
Some projects from other donors
Appendix: Does India’s project aid matter?
References
6 Resource Management, Institution Building, and Technical Assistance
Aid’s contribution to resource management and institution building: the organized sector
Resource management and institution building: the rural sector
Aid to remove administrative and manpower deficiencies
References
7 Aid and Market Forces
The ‘control syndrome’ and the market critique
The effects on private foreign investment
The control syndrome and domestic business
Donors and the control syndrome
How much aid goes to the private sector?
Aid and the private sector: recent policy changes
References
Appendix Conclusions of the India Aid Effectiveness Study
Notes
Index
Volume 11
Cover
Half Title
Title
Copyright
Original Title
Original Copyright
Contents
Preface
Acknowledgements
Introduction
1 Background to Malawi
Colonisation and After
Malawi at Independence, 1964
Poverty
Federal Effects
Lack of Skilled Manpower
Physical Resources
The Economic Infrastructure
Economic Dependence: Needs and Prospects
The Development Strategy
2 Malawi’s Economic Performance after Independence
Crisis and Consolidation
General Economic Performance
Output
Employment
Investment, Consumption and Savings
Public Finance and Administration
Balance of Payments and External Trade
Development of the Infrastructure
Development of the Social Services
Land and Population
Conclusion
3 Britain as Donor
General British Aid Policy
Britain’s Aid Programme to Malawi
Consistency of Objectives
British Aid to Malawi: the Facts
Volume
Terms and Conditions of Aid
Scope for Leverage
Conclusions
Appendix: CDC’s Role in Malawi
4 Budgetary Assistance
Government Finance and Budgetary Aid
Contribution to Development
An Expenditure Cut
Financing the Deficit without Budgetary Aid
British Policy on Budgetary Aid
The Basis for the Elimination Policy
Why Malawi Agreed
Was the Rate of Elimination Appropriate to Malawi’s Needs?
The Gentleman’s Agreement
5 Aid to Malawi’s Development Account
Malawi’s Development Programme
British Aid
British Influence on Malawi’s Development Choices
Direct Effects of British Project Aid
Shunting
Assessment of British Project Aid
Natural Resources
Transport
Education
Other Aided Projects
6 Britain’s Technical Assistance to Malawi
Supplementation Schemes
Malawi’s Localisation Policy
The Effects of the Supplementation Schemes
Without Expatriates in the Public Service
Obtaining Expatriate Public Servants without British Aid
Alternative Approaches to Localisation
Volunteers
Overseas Training and Education
7 Conclusions
Aid and Development
Implications for Future Aid Policy in General
Implications for Aid to Malawi
Index
Volume 12
Cover
Halftitle
Title
Copyright
Original Title
Original Copyright
Contents
List of Tables
Foreword
Introduction: The Role of the OPEC Member Countriesin Financing Third-World Development
1 The Establishment and Evolution of the OPEC Fund
2 The OPEC Fund’s Experience (mid-1976 to mid-1983): Its Approaches and Procedures, the Magnitude of its Assistance, and its Impact on Development
3 Agriculture and Energy in the OPEC Fund’s Activities—A Sectoral Perspective
4 The OPEC Fund and Africa—A Geographical Perspective
I African Development and the Role of the OPEC Fund
II Supporting Sub-Saharan Africa: the Need for Concerted Strategy
5 The OPEC Fund and the Least Developed Countries: An Orientation Towards the Poor
I The OPEC Fund’s Assistance to the LLDCs
II The Need for Commitment
6 Innovative Forms of Cooperation: The OPEC Fund’s Approach
I Institutional Cooperation and Co-financing with Other Donor Agencies
II New Partnerships in the 1980s: Co-financing with Commercial Sources of Funds
7 The OPEC Fund and the North-South Dialogue—SomePersonal Reflections
Annex 1: The OPEC Fund’s Loans and Grants, 1976-April 1983, by Region
Africa
Asia
Latin America and the Caribbean
Loans and Grants Signed to April 15, 1983
Grants for Technical Assistance, Research, etc.
Annex 2:Tables 9–20
Detailed Statistical Data Referred to in Chapters 3 and 5
Annex 3:Acronyms
Index
Recommend Papers

Routledge Library Editions: Development, Mini-Set B: Aid
 0203840186, 9780203840184, 9780415584142, 9780203840351, 9780415592765, 0709938292

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

ROUTLEDGE LIBRARY EDITIONS: DEVELOPMENT

THE POLITICS OF UNITED STATES FOREIGN AID

THE POLITICS OF UNITED STATES FOREIGN AID

GEORGE M.GUESS

Volume 7

LONDON AND NEW YORK

First published in 1987 This edition first published in 2011 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 This edition published in the Taylor & Francis e-Library, 2011. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk. Routledge is an imprint of the Taylor & Francis Group, an informa business © 1987 George M.Guess All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 0-203-84018-6 Master e-book ISBN DOI: 10.4324/9780203840184 ISBN 13: 978-0-415-58414-2 (Set) eISBN 13: 978-0-203-84035-1 (Set) ISBN 13: 978-0-415-59276-5 (Volume 7) eISBN 13: 978-0-203-84018-4 (Volume 7) Publisher’s Note The publisher has gone to great lengths to ensure the quality of this reprint but points out that some imperfections in the original copies may be apparent. Disclaimer The publisher has made every effort to trace copyright holders and welcomes correspondence from those they have been unable to contact.

The Politics of United States Foreign Aid GEORGE M.GUESS Institute of Public Administration Georgia State University Atlanta, Georgia

CROOM HELM London & Sydney

This edition published in the Taylor & Francis e-Library, 2011. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk. © 1987 George M.Guess Croom Helm Ltd, Provident House, Burrell Row, Beckenham, Kent BR3 1AT Croom Helm Australia, 44–50 Waterloo Road, North Ryde, 2113, New South Wales, Australia British Library Cataloguing in Publication Data Guess, George M. The politics of United States foreign aid. 1. Economic assistance, American— Developing countries I. Title 338. 91’73’01724 HC60 ISBN 0-203-84018-6 Master e-book ISBN

ISBN 0-7099-3829-2 (Print Edition)

CONTENTS

List of Tables and Figures Preface 1.

2.

3.

4.

THE POLITICAL ECONOMY OF US FOREIGN AID: PAST AND PRESENT

1

Introduction

1

The Blurred Lines of Foreign Aid Policy

2

Contemporary Foreign Aid Programs

9

The Evolution of US Postwar Foreign Aid

18

CRITICAL PERSPECTIVES ON FOREIGN AID

35

Introduction

35

Harsh Evaluations From The Political Spectrum

36

THE BUREAUCRATIC ROLE CONFLICT MODEL

52

Introduction

52

The Need For Policy Models

52

Results of Foreign Aid

56

Decision Premises or Goals

59

Constraints: Rules and Repertoires

62

Opportunities: Role Conflict

71

Conclusion

86

US AID TO LATIN AMERICA

87

Introduction

87

Results

88

Goals

96

Constraints

103

Contents

5.

6.

7.

vii

Role Conflict

109

Conclusion

115

US AID TO ASIA

117

Introduction

117

Goals

119

Results

127

Constraints

133

Role Conflict

138

Conclusion

143

US AID TO THE MIDDLE EAST

145

Introduction

145

Results

147

Goals

156

Constraints

159

Policy Actors

162

Conclusion

166

CONCLUSION

167

Introduction

167

Summary and Conclusion

168

Recommendations

174

Index

187

TABLES AND FIGURES

Tables 1.1 3.1 3.2

Presidential Request For Fiscal Year 1986 Budget Authority For Foreign Aid Budget Authority Requests and Appropriations For US Foreign Aid: 1975– 1986 ($000) Congressional Earmarks For Economic Support Fund, FY 86 ($000)

11 75 79

Figures 1.1 1.2 3.1 3.2 3.3 4.1 5.1

US Foreign Military Assistance, 1946–1981 Distribution of US Foreign Assistance by Region, 1946–1982 The Bureaucratic Role Conflict Model USAID’s Planning and Budget Process, Fiscal Year 1986 Guardian and Spender Role Conflict in US Foreign Aid Fiscal Year 1986 Request For Economic and Military Assistance Programs (Scaled To Program Size) Allocation of Bilateral Foreign Aid By Major Region, FY 1987

14 20 56 64 74 98 118

And I will shake all nations, and the desire of all nations shall come. Messiah, George Frederich Handel I wish the Americans would just leave us alone, one Guatemalan declared, if we want to kill each other off it’s our business. The US has no right to interfere. Inevitable Revolutions, Walter LaFaber.

To Regula, Andy, and Marty

PREFACE

Despite its historical inability to attain few complete successes or even unconditional failures, foreign aid is still the most controversial component of US foreign policy. In 1976, LaFaber notes that 10,000 machine guns, provided to the El Salvadorean military as part of the earlier Alliance For Progress “Civic Action” program, were proffered as surplus to East Coast mobsters in the US (1984:203). Not all foreign aid produces results this dramatic or lucid. Most projects lead to rather mundane results like waterholes for livestock, land titles, planted trees or crops and trained teachers and nurses. But recently it was discovered that former Philippine dictator Ferdinand Marcos illegally diverted nearly $2 million in interest from US Agency For International Development (USAID) provided funds for his unsuccessful re-election bid in February, 1986 (Rupert, 1986). Why does the US provide aid to such “allies” when it is clear that “development” however defined will be an unlikely result? How do multiple political pressures from the US Congress, Presidency, Defense Department and State Department affect the scope and purpose of US foreign aid? Are the processes and results of US aid uniform across geographic regions? What explains the differences between aid project successes in countries like Taiwan, Brazil, and Egypt, and other failures in the Philippines, Vietnam and Nicaragua? Are there comparative lessons that can be drawn from aid country and program results that can be used to improve future development prospects? This book argues that US aid is very much the product of a distinctive process of American bureaucratic politics. This process, in turn, relates to the hybrid status of foreign aid as a component of both domestic and foreign policy. Foreign aid programs are perhaps the main tools of US foreign policy. Yet they are planned and administered like a domestic US policy. Foreign aid is one of the few domestic programs whose beneficiaries are abroad and domestic constituents are quiescent unless an international crisis affects one of their interests. Functionally, US foreign aid is a highly politicised domestic program conceived and applied through a host of large institutions subject to federal laws and administrative rules. Recipient interests are often ignored in both planning and implementation because their needs cannot be defined consistently with US bureaucratic routines. The vacuum left by absence of powerful domestic interest groups clashing over substantive foreign aid issues in Congress, shifts excessive political power to the Executive branch with its defense-driven foreign policy machinery. Congress reacts in an almost negative programmatic fashion by adding more restrictions to the program. For instance, in Fiscal Year (FY) 1986, foreign aid resources were cut by congressional appropriations, reduced further by the new Gramm-Rudman-Hollings “sequestration” (an automatic withholding of 4.3% of the funds), and limited by selected earmarks (purely political allocations by Congress) (Nowels, 1986:18). Under these conditions, it is not surprising that US foreign aid is often ineffective; the wonder is that it ever succeeds by any criteria. This book contributes more ink to the well-covered field of foreign aid by attempting to explain major trends in program-project results in Latin America, Asia, and the Middle

xii Preface East with a framework composed of factors which experience suggests are important. “Bureaucratic Politics” frameworks have been in vogue for some time now because they claim to go farther than the suggestion that politics caused it all. They also fit nicely with the real perception that since the world is becoming more institutional and bureaucratic (in a neutral sense) the insights of such frameworks must explain something. The usual problem with such models is that their contents are either atheoretic in the sense of being a hodgepodge of unrelated or unweighted factors, or they simply add up to another “black box” into which one may fit anything institutional to explain the phenomena at hand. Here it is stressed that, as in most policy areas, some institutional actors seek more resources and other actors try to control their behavior and it is this interaction that explains a great deal of the results of foreign aid. If the actors represent real interests (as opposed to mere personal agendas) the conflict can lead to institutional trust and healthy results. But in foreign aid, the roles (behavioral expectation attached to position) tend to be several layers removed from either recipient or US country interests. Conflict is often superficial, a reflection of underlying institutional distrust which also tends to increase it. This often leads to highly bureaucraticised programs with lots of unintended consequences that encourage recipient-donor dependency and produce eventual “ward” status in places like the Philippines. In particular, the absence of a clear and forceful “guardian” role, played for example by the foreign aid authorizing committees in Congress, means that the annual competition for budgetary resources is reduced to a power play among many fragmented agency “spenders”. Those with the bulk of power in this process, the President, Department of Defense (DOD), and particular members of Congress, tend to control the process and often reduce the effectiveness of both USAID and the foreign aid program it administers. The proposed model suggests that these tendencies are explicable by combinations of foreign policy goals, bureaucratic rules and repertoires and actor role conflicts. Superficial conflict allows many bureaucratic advocates to dominate foreign aid policy-making over a few ill-defined guardian actors that attempt to make the program more effective. This state of affairs tends to generate programs that are overly-concerned with defense-security, and profitability over the developmental and welfare needs of recipient countries. In short, superficial conflict among bureaucratic actors with often “mixed” guardian-spender roles tends to produce results that are often inconsistent with US foreign policy interests and the developmental needs of its intended beneficiaries. Operating in the larger context of American politics and US foreign aid policy-making, foreign aid functions as everyone’s “clay pigeon” for which it is alternately blamed for both successes and failures. Under these difficult conditions, foreign aid should either be discontinued as a waste of effort and resources, continued for other than economic or political development purposes, or drastically reformed to be an independent force for Third World development. Like many before me, I propose reform in the belief that foreign aid under appropriate circumstances can benefit both the long term interests of the US and the Third World. The author wishes to thank the following people who have contributed in both direct and unintended ways to my foreign aid and policy adventures, all of which have been in Latin America: Melvin Gurtov, Ronald Chilcote and Michael Reagan of the University of California, Riverside, Travis King, formerly with USAID in Costa Rica, Juan Bautista Schroeder, formerly with the Organization of American States (OAS) in Costa Rica, Robert

Preface xiii Grosse of the University of Miami, and James Rowles at Harvard University. Naturally they are all absolved from responsibility for the unfolding of this book, its mistakes, and conclusions. The author also wishes to acknowledge the support provided by a Faculty Development Grant from Georgia State University for interviews with USAID and State Department officials in Washington, D.C., in March and June, 1986. The author is also grateful for the reference assistance of Gayle Christian in the GSU Pullen Library.

Chapter One THE POLITICAL ECONOMY OF US FOREIGN AID: PAST AND PRESENT

Introduction As the US struggles for influence in the world power arena, its foreign assistance program follows along as a willing and important appendage. In a precarious context, characterized increasingly by desperation policy responses to coups, counter-coups, terrorism and religious fanaticism, the lines between US foreign policy, foreign aid and trade activities become blurred. That foreign aid lacks autonomy among American public policies is harmful in several ways to US interests abroad and to the needs of the developing world. First, foreign aid may be erroneously credited for foreign policy power successes though development projects fail and underdevelopment may be increasing. Second, and more commonly, foreign aid is unjustly blamed for foreign policy failures. The rantings of a Khadafy are viewed by many as another failed foreign aid attempt, the cut-off of which penalizes the Libyan people and not their officialdom. Consistent with this, some believe that foreign aid is not criticized enough. “Who could be against aid to the less fortunate?” ask Bauer and Yamey (Thompson, 1983:119). “When aid advocates talk of the disappointing record of aid, they mean not that aid has been ineffective or damaging but that the amounts of aid have been insufficient.” Foreign aid consists of five programs: (1) Economic Support Fund (ESF), (2) Development Assistance (DA), (3) Food Aid or PL 480, (4) Security Assistance, and (5) multilateral Development Banks. These interrelated programs have separate constituencies and are proposed for funding by bureaucratic actors that are, in turn, controlled for efficiency and effectiveness by Congress working in conjunction with them. But these “checks and balances” have produced paralysis instead of healthy competition. Task definition is imprecise and institutional distrust among key actors tend to inhibit the proper functioning of checks and balances. Further, that US foreign aid is a “microprogram” evaluated as a “macropolicy” creates false expectation and the likelihood of unfair judgment. Over the five major phases of foreign aid, the programs have been billed as dramatic macropolicies: spreading democracy, containing communism, and getting the poor ready for developmental “take off”. Tangled up with foreign policy events and the personal agendas of congressional “experts” and high-level amateur appointees in USAID, it can be stated without great risk that the bulk of foreign aid programs have been largely “unsuccessful” (or successful in a trivial sense). Perhaps the most successful program was also the shortest and most uncomplicated: the Marshall Plan. But since that program, the goals of foreign aid have broadened in almost inverse proportion to useful knowledge on the causes of poverty and underdevelopment. As the “fall guy” of foreign policy, the foreign aid program has been blamed by the left for DOI: 10.4324/9780203840184-1

2

The Politics of United States Foreign Aid

neo-imperialism and by the right for generating exaggerated expectations among the poor which destabilizes political systems. Foreign aid is a product of the American political system, a highly bureaucraticised network of actors that clash over resources and the authority (or turf) to influence policy. To the extent that we can analyze foreign aid through the Bureaucratic Politics lens, we may be able to point the way to a more autonomous policy that will be more likely to achieve realistic objectives. This book seeks to describe the evaluative dilemma of US foreign aid as part of foreign policy and to explain how and why the program is often ill-designed and poorly executed. The goal is constructive: to enhance the capacity of foreign aid to benefit the Third World which indirectly can enhance US influence in world affairs. The making and execution of foreign aid policy has been characterized by intense confusion over both objectives and evaluative criteria since its initiation in the early 1940s. Foreign aid is not really “foreign” policy or a “domestic” program; yet it is planned, executed and evaluated as if it were both. Hence, despite its marginal budgetary expense in US terms, it is nearly always a controversial policy (Montgomery, 1986:94) In FY 1985, the Presidential foreign aid request amounted to less than 2% of the budget ($15.2 billion out of $925 billion or 1.6%), In FY 1978, US aid amounted to only 0.23% of GNP or about the same level provided by Austria, Japan, Switzerland, or Germany (Congressional Budget Office, 1980:9). Nevertheless, Congressional Quarterly (1985:2688) suggests that foreign aid “one of the most unpopular issues that Congress faces each year..” Many have written of US foreign aid; many have written it off. But few have provided other than general frameworks for analysis. Critics of foreign aid tend to provide the more rigorous policy-oriented frameworks. Still, they tend to be simplistic, ignoring the real world of interest-driven bureaucratic policy-making which constrains both the US and its recipient countries. It is suggested that a Bureaucratic Politics model emphasizing “role conflict”, largely over budgetary resources, can be useful in explaining past failures and successes as well as providing a more solid foundation for foreign aid reform.

The Blurred Lines of Foreign Aid Policy Many definitions of foreign aid and its purposes have been provided. Let us examine some of them before describing the current US foreign aid program in detail. Many observers agree that US foreign aid is a feeble effort to transmit the contradictory values of American political culture abroad with largely uneven results. Operationally, foreign aid consists of loans and grants of funding or technical assistance for security, developmental and humanitarian objectives. Montgomery (1967:1) notes that American foreign aid is a disappointment to both those who think we ought to run the world and those who think we ought to let the world alone. He suggests that Americans want to tinker with the world and improve it, without really changing it much (1967:2). Foreign aid then, is one expression of national impulse by which the US tries to “buy time” in a fast-changing world (1967:5). That foreign aid is frequently pressed into service of American foreign policy for “diplomatic” (establish a presence by earthquake or food aid), “compensatory” (military base rights in exchange for aid) and “strategic” purposes (improve world order via economic and military aid) (1967:10–19), opens it to controversy on a variety of different levels. The distinctive purpose of foreign aid is to assist other countries attain conditions (political, economic,

The Political Economy of US Foreign Aid: Past and Present

3

social) that will serve world order and freedom (Montgomery, 1967:23). This purpose was reaffirmed in the 1983 Carlucci Commission Report which suggested that US assistance programs “make an indispensable contribution to achieving foreign policy objectives” (Schultz, 1984:1). This rather tall order means that a small expenditure of US aid can produce economic stability and political democracy under a variety of cultural conditions. Given the porousness of such objectives, it should be of little surprise that foreign aid often fails at one or more levels, or succeeds at one level and fails at another. At least four levels of results may be distinguished: (1) regional-international, (2) country program results, (3) project results, and (4) individual technician performance (Montgomery, 1967:74). For example, security assistance is often evaluated on a regional-international plane: deterrence, conflicts managed, alliances preserved, and access for US forces preserved (DOD, 1986:23). The success of US arms sales via Foreign Military Sales (FMS) is measured by its contribution to strengthening the professional military establishment, as well as providing jobs in the US and reducing unit costs for items purchased by the US military (DOD, 1986:14). The Economic Support Fund (ESF) is often evaluated by its capability to provide resources to prevent financial chaos and encourage economic reform (DOD, 1986:28). For Development Assistance projects (DA) of USAID, measures include, farmers trained, and land titles distributed. So, if a project for land titling succeeds in a country plagued by political instability that eventually turns into a totalitarian state, foreign aid shares the blame in the public and even official mind. More commonly, project results often produce unintended consequences for country programs. Montgomery notes the irrigation project that produced both increased agricultural production and malaria (1967:80). More recently, the Sudanese “Freedom From Thirst” projects financed by multilateral development banks (MDBs) and non-US donors, proposed 4000 wells per year to increase water for cattle. Failure to control surrounding land use (colonial governments had restricted the number of wells dug to limit pressure on the carrying capacity of the land) resulted in over-grazing around the wells with widening “circles of death”. Though the project attained its objective—more water, the wider effects have been environmental degradation and famine (Albright, 1985). Other writers have focused on the paradoxical domestic context of foreign aid policymaking. For example, O’Leary (1967:5) suggests that aid policy is supposed to please US public opinion, the host country, and the technicians who work in it. It must also please Congress and the President. At the same time, foreign aid policy is made in a “schizophrenic” fashion by one set of actors that favors bilateral military security (White House staff, State and DOD) another led by Treasury that seeks economic security through multilateral development banks (Rowen, 1986), and another led by USAID that seeks bilateral development. The larger, also paradoxical, objective of Congress seems to be to “micromanage” the program with many restrictions but to keep away from an unpopular issue. But aid policy itself is often a hodgepodge of grants, loans and trade favors. Where economic and trade policies end and foreign aid begin is often unclear in the minds of many decision-makers and this contributes to the continuing malaise about aid. O’Leary notes that while the level of public support and perceived understanding of foreign policy and defense postures are usually high, they are minimal for foreign aid (1967:125). Asher (1961:4) suggests that part of the problem may be semantic: “aid” connotes action “for” rather than “with” others; “recipient” or “host” are terms often used instead of “participating” countries. So when the US spends to achieve the grandiose and almost indefinable

4

The Politics of United States Foreign Aid

objectives of “strengthening” military defenses in the” free world”, and promotion of “political development”, predictable failure brings out the plaintive cry of wounded gratitude at best, sell-outs and conspiracies at worst. At the other end of the scale are strict constructionists, who exclude from their definitions of aid anything that results from donor selfinterest. For example, to McNeil (1981:45) a military base would not be aid; Montgomery, as noted, would view this as “compensatory” aid (1967:10–19). Since some suggest that donor self-interest is always the prevalent motive, by this view pure aid would not exist. The problem of defining an autonomous foreign aid policy is complicated by the continuing inability to define developmental objectives in other than subjective ways and the close relationship between defense-security programs and foreign aid. As noted by Morley (1961:1), foreign aid was supposed to be temporary after World War II and not a “continuing liability”. But foreign aid programs “have been continuously retooled to meet various military and diplomatic cold war crises”. Put another way, unlike domestic policies that are usually evaluated on their own terms, such as capital and operating subsidies from the federal government to improve local government transit performance (which have not been very successful but are not very controversial either), foreign aid is evaluated by the success of such vague foreign policy doctrines as “containment” or “frontiersmanship”. Conversely, when a President such as Ronald Reagan wants to make an impact on trade, defense and anti-terrorism, foreign aid benefits from a larger budget request and stronger political support. For example, the FY 1987 request of $16.3 billion for foreign assistance is about $1.8 billion higher than the current year level. According to Nowels (1986:3) “Much of the program is closely linked to US defense strategy and is viewed as a costeffective means of providing the maximum security benefits to the US without direct US involvement.” Many have noted the foreign aid-foreign policy linkage, but few have suggested practical means or even rationale for separation. For instance, anticipating current US foreign aid dilemmas in Central America and the Middle East and past ones in Southeast Asia, Morley laments that “.. in extending the policy of containment and the system of anti-communist military alliances around the world, Washington in effect undertook to support indefinitely the economies and military forces of certain less developed countries” (1961:1). So foreign aid remains obscurely defined, now a “permanent postwar innovation” but still “the most controversial aspect of foreign policy” (Montgomery, 1967:6). Whatever the deficiencies of a rationale for foreign aid, they pale in significance to field problems of implementation. The frequent hiatus between theory and practice in foreign aid is often explained by the political necessity of tying it to the vicissitudes of foreign policy. Expenditures unaccompanied by firm political support usually do not achieve intended results. In noting the lack of rhyme or reason in the pattern of US aid expenditures, Kaplan (1967:251) suggests that the most ‘festering criticism” of aid is that it is used by recipients for purposes inconsistent with US foreign policy and national security interests. He noted in 1967 what is still true, that allocation of defense support and military assistance is frequently unrelated to the eventual usefulness of recipient military forces (South Vietnam) and allocation of development aid doesn’t usually correspond to recipient growth rates (1967:252). Many technical reasons can be and have been advanced for the absence of a direct relationship between foreign aid and results. But the main problem is

The Political Economy of US Foreign Aid: Past and Present

5

that of translating broad, shifting and often contradictory goals into specific programs and projects to be executed in highly volatile political cultures. Morgenthau (Liska, 1960:vii) notes that even the notion of foreign aid as “instrument of foreign policy” is controversial since many believe that “foreign aid is an end in itself”. Controversy is often functional to public policy in that a scandalous political revelation, a good cost overrun, or a demoted whistle-blower can fasten public attention onto a problem, from which remedies can be applied and evaluated. So, in the intergovernmental transit grant program mentioned, perverse incentives have been generated in local transit agencies contrary to grantor intentions, such as overcapitalization and deferred maintenance. These conclusions became part of the public agenda and program revisions followed accordingly. The US domestic political system produces such programs; the transit program is a product of powerful conflicting interest groups such as labor unions, highways, local governments and mass transit advocates. Both qualitative and quantitative evidence on program results can be presented during annual appropriations hearings. Hence, the level of controversy is quite low (standards are fairly clear; debate is ongoing) except during periods of crisis (system breakdowns) and scandalous revelations (overruns and contractor fraud). This is the normal pattern of policy-making and revision in the US. But foreign aid is a hybrid of foreign-domestic policy, meaning a clamorous reauthorization process with little attention to practicality and substance in most fiscal years. The FY 86 authorization was the first time in three years that Congress passed a regular foreign aid authorization. But its continued unpopularity is evident from the fact that foreign aid appropriations are higher if the bill doesn’t reach the floor of Congress for debate, i.e. when foreign aid is folded into the “continuing appropriations resolutions” of the last several fiscal years. Legislative management and oversight of foreign assistance has been irregular because the program itself is a product of larger foreign policy themes such as “containment” of Communism and new initiatives, as well as the electoral need to “micromanage” a program viewed largely as porkbarrel and waste by Congress. For example, this year Congress is considering authorization (required before “budget authority”, and appropriation) of aid to Northern Ireland, Haiti, the Philippines, the Howard Baker plan to expand World Bank activities, and security assistance to trouble spots such as the Middle East. Relative shares of the foreign aid program budget (development, security, economic support) are determined, in other words, by the balance of power within the US policy-making process. Domestically, its political support is weak and diffuse except among special groups like expatriates of affected countries, and farmers. Where foreign aid ties into security, it is supported by the same interests (contractors, related defense industries, powerful congressional defense advocates) that contribute to growth of the defense budget. Logically then, the pattern of power politics within government reproduces itself for the US in the balance of power tradition of “realistic” international politics. The yardstick is power and foreign aid becomes a concomitant of legitimate power-seeking by the US. In contrast with the combined “liberal” 19th century British utilitarian notion of a world of nations, linked through trade and ruled by the power of enlightened public opinion (a visionary concept), and the moral-legalistic notion of Kantian categorical imperatives, realists such as Hans Morgenthau define “interest” in terms of power (Hoffman, 1985). Thus, foreign aid becomes part of the apparatus for advancing US world power objectives abroad.

6

The Politics of United States Foreign Aid

Whether or not this is true (many AID technicians would certainly deny this because they are evaluated on less abstract criteria in the field than contribution to US power), the foreign aid program is treated by Congress and the public as if it were. Controversy surrounds its ends (“develop” Central America), while its programmatic means often pass scrutiny unless they involve political monuments, such as roads or dams. Since foreign aid is attached to foreign policy, and lacks conceptual and methodological precision in generating “development”, this means that controversy tends to be endless. It cannot be resolved by a hearing, a scandal, a quick stroke from a committee, by reorganization, or even a revised expenditure pattern. Clearly improvements can be made. But in a previously unrealized sense, all foreign aid solutions under these conditions, lead to new problems. Before passing to a description of the current US program and criticism of it, let us examine several examples of the kind of negative controversy which surrounds foreign aid and tends to hinder positive results. For example, the US and Soviet Union compete for African loyalties by various means, including the provision of foreign aid. Most of the US aid is for agricultural rural development and nutrition ($148 million) and for shortterm economic support ($410.5 million) (USAID, FY 87 Congressional Presentation 1986:272). Most Soviet aid takes the form of capital-intensive factories and arms. Hence, one could expect the US to have an enormous advantage simply from targeting groups with the greatest need. Instead, US efforts, through the Food For Work (FFW) program have not been successful and instead have generated intense criticism from many directions. According to Shepherd (1985:44) “the primary beneficiaries seem to be those who possess land or who are in the food-aid bureaucracies.” Though intended in principle to satisfy basic needs of poor rural people by paying wages in food, the program employs people but at a social cost of “make-work” projects, and competition with local African agriculture for limited resources. Food For Work “undermines the price structure for locally-produced food” (1985:44) creating dependency on consumption of imported foods. Half the people of Lesotho now receive US food aid; fallow land increased 26% from 1970–1980 as farmers shifted out of production into FFW programs (1985:45). Put another way, the Soviets have actually gained more political capital from ignoring the difficult problems of poverty and famine than the US which often assumes the role of “pitiful helpless giant” or the fall guy of the international development business. When Congress appropriates money for domestic programs such as transit grants, the Department of Transportation which receives the money and spends it can be criticized for failing to generate increased ridership and ultimately contributing to the federal budget deficit. But even these criticisms can be countered effectively, by demonstrating that other variables prevented success, such as lower oil prices, or by demonstrating positive results in the total program. The controversy passes and more money is appropriated. Here, Congress need not concern itself with more exotic possibilities such as the transit system falling into Communist hands, the rise of local dictators, land ownership patterns or bizarre religious customs. Such factors would at least make program failure more interesting since they raise other issues besides the effectiveness of grant mechanisms in achieving often contradictory objectives. But they are not there; diversity of constituents tends to normalize policymaking in the US. The reverse is true in foreign aid. In the words of one Congressman “foreign aid has no constituency, but it sure does arouse passions” (Farnsworth, 1985).

The Political Economy of US Foreign Aid: Past and Present

7

Since project technical problems (build a rice mill or irrigation system) are intimately tied to local cultural and political conflicts, a domestically-conceived and internationallyevaluated program is almost beaten from the start. Recalling the unintended and contradictory consequences of the FFW program above, in 1984 Ethopia blamed the US for the extent of famine which it faced. USAID responded that the US was the largest food relief donor to Ethiopia ($109 million in 1985) and charged that the Ethiopian government had wasted between $100–$200 million on a 10th anniversary Marxist celebration instead of feeding its own people (Molotsky, 1984). Clearly both sides talked past each other at a non-programmatic level, and “famine politics” could be expected to continue. A major non-technical issue here was that some of the US food aid would likely reach rebels of the current Marxist government in northern Ethopia (Harden, 1985; Ottaway, 1985). So some challenge the program for disincentives to local producers; others broadside the politicalideological distribution of aid; still others want such extraneous conditions as population control attached to all emergency relief aid (Lamm, 1985). (The last criticism is absurd, of course, since falling death rates and not rising birth rates have been the main cause of population growth rates in poor nations; Eberstadt and Lewis,—1986:36). To condition food aid on population control would simply increase the death rate, a result nobody could possibly want! To be successful, foreign aid must therefore efficiently achieve economic and political development, maintain security, and provide welfare benefits to the needy. Failure along any of these fronts opens the entire program to passionate criticism, linked by the media and the public to foreign policy. For example, as evidence that the US-backed centrist liberal politics of President Duarte in El Salvador have improved human rights and “curbed the warlords of both right and left” (New York Times, 1985), others fear the implications of the US dream to find “the vital center in the developing world”. Some view security and political development aid as exclusive items that may even be tradeoffs. For, according to Tonelson (1985): “Today both sides portray their favorite bands of cuthroats as champions of freedom and often suggest that Washington is morally obligated to save them. Apparently they do not see that this is most likely to plunge us into another senseless and politically divisive war.” In a comment on the FY 85 foreign assistance bill for $12.6 billion, President Reagan stated that the measure contained too much economic assistance and not enough military aid (New York Times, 1985). Nevertheless, foreign security aid has been also controversial for producing spectacular failures for which the entire foreign aid program is challenged. For example, much has been written on the US role on the rise and fall of the Shah of Iran. Feinberg (1983:211) suggests that the “highly visible US presence in Iran ironically contributed to the Shah’s demise.” This was “ironic” because the large-scale arms transfers actually aroused the opposition as they became aware of the contradiction between the Shah’s purchase of used weapons at exorbitant prices while so many Iranians were still poverty-stricken. To the Iranians, the US was the main beneficiary of military aid via arms sales. This worked against any gains by the economic component of foreign aid, in particular “institution-building” for evolution to democracy. Emphasis on security at the expense of socio-economic institutional capacity meant stability depended upon developmental policy-making by the military, almost a contradiction in terms (the Peruvian military in

8

The Politics of United States Foreign Aid

the 1960s and the Brazilian military in the 1970s would be exceptions). Thus, it could be concluded with little surprise that: “.. the rise of Islamic fundamentalism has been provoked by the failure of developing countries in the Middle East to develop effective modern institutions. Government has been through an elite, corrupted by office or oil wealth (or both) and without any significant form of channeling popular opinion” (Graham, 1985). The penchant for military dominance of foreign aid cannot simply be attributed to hubris or imperialistic designs. The Foreign Military Sales program for FY 86 ($6.1 billion) alone amounted to 39.3% of all US foreign assistance appropriations (USAID, 1986:Table 2). As noted, foreign aid is akin to foreign policy but both are the uneven products of American interest group politics. While foreign aid is thin on Political Action Committee (PAC) group support, defense policy is not. Some suggest that congressmen see broader national security policy, unlike most domestic programs, as an issue set apart from constituency and interest group pressures, on which they vote their own “world view”. But weapons procurement policies are really a matter of distributing Defense Department (DOD) and armament industry goods and services to congressional districts (Woll, 1985:449–451). Hence, over the area of weapons procurement and sales, Congress exercises only marginal oversight and functions through pork-barrel instead of sound policy guidance. The influence of interest groups thus determines security assistance to a much greater extent than economic assistance. Defense interests are simply more powerful and sophisticated than non-defense interests at gaining budget authority (appropriations). So it should be no surprise that the US foreign aid program is punctuated by examples of “permanently dependent client states” (or “ward” states) spawned not by food aid or development projects, but rather by lavish provision of security-related assistance. Recipients dependent on US arms sales were repaying FMS loans with higher interest rates and this “has contributed in part to the Third World’s current difficulties in servicing total national debt” (DOD, 1986:59). To compensate, the US initiated concessional rates in some cases, and forgave FMS repayments in several countries (no interest or principal payments for Egypt or Israel). This is a major theme of this book, to be explored more fully in Chapters 4–6 on the regional functioning of US foreign aid. Shaw notes for example (Wall Street Journal, 1985) that spending for security-related assistance has doubled since 1981 ($6.7 billion or 41.5% of all foreign aid for FY 87, excluding the Economic Support Fund, which is also security-related) in current dollars and is up nearly 60% in constant dollars. This is a higher increase than the nearly 60% constant dollar increase in the Reagan defense program. If one includes “emergency economic aid”, Israel and Egypt consumed over 50% of the FY 85 security assistance program and more than 33% of the $15 billion foreign aid program (Shaw, 1985). Many such as Shaw attribute the current economic crisis in Israel to “overly generous US economic aid that has permitted Israel to put off facing up to the consequences of its economic folly. “But military aid to Israel, primarily by off-budget financing for foreign military sales (FMS) has taken on “political life of its own, unrelated to objective considerations of military need or sound budgeting. “US military aid has already equipped the Israeli military and now subsidizes “the growth of Israel’s expensive and inefficient defense industry” (Shaw, 1985). According to Kaiser (1984), “In Washington, reporters

The Political Economy of US Foreign Aid: Past and Present

9

and politicians share a cynical understanding that Israel and its American friends constitute probably the single most effective lobbying force in the country; they take its victories for granted.” Further, Egypt’s aid requests become more difficult to reject when Israeli aid is increased. After Egypt and Israel, the remaining $2.5 billion in security assistance goes to “base rights” countries such as the Philippines where policy has historically been made by executive agreements (US President to ex-President Marcos until 1986) apart from congressional considerations. Should this expensive security assistance operation fail to hold the line (against terrorism for example), the modest economic programs are deemed culpable by many. In a now familiar pattern, more US security assistance is pumped in to keep the ward regime afloat using ESF funds to buy time. This tends to work until that regime falls to political-religious extremists of the left or right. This pattern occurred in Iran before the fall of the Shah and was repeated in Nicaragua in 1978. That it did not recur in the Philippines in 1986 suggests that, despite all odds, learning may have taken place in the foreign aid arena.

Contemporary Foreign Aid Programs The theory behind the current foreign aid program is that it benefits less developed countries (LDCs) in two ways. First it increases the total resources available to them which increases their rate of investment. This sequence is a prerequisite to growth of production, employment and income. Second, foreign aid increases the foreign exchange resources of LDCs, enabling them to buy imports from abroad which can increase their rate of growth (Congressional Budget Office, 1980:4). The US foreign aid program is not simply bilateral and does not exclusively target low income countries. Nor is US foreign aid exclusively “developmental”. As noted, foreign aid consists of five related programs. Four are bilateral (Security Assistance, Development Assistance, Economic Support Fund, and Food Aid (PL 480) and one is multilateral (US contributions to multilateral banks or MDBs). Bilateral aid covers all forms of assistance provided by the US to LDCs; bilateral development aid refers only to loans and grants administered by USAID and technically excludes ESF and Food Aid. The title of this book is “US Foreign Aid”, meaning that all forms of aid will be examined, including military assistance, since all presume to “develop” the recipient. To restrict ourselves to Development Assistance (DA) would ignore the fact that after FY 1974, ESF replaced DA as the largest component of US foreign aid in constant dollars. For FY 86, budget authority requested for DA will be only 15% of all “foreign assistance”. The remaining 85% will be devoted to “International Security Assistance” which includes ESF and Foreign Military “Sales Credits” in the annual budget classification (Office of Management and Budget, 1986: Appendix). Some have argued that since military and economic aid both free LDC resources (tax revenues) to spend largely for their own purposes, the line between economic and military aid is arbitrary (Kaplan, 1967:283) Whether or not this is true, budgetary policy aspects of all related programs should be examined and the OMB classification aids in this task. The US foreign aid program is designed to reach different levels of LDCs (measured in per capita gross national product) for differing policy objectives. Asher (1961:65) notes that statistics reveal little about any principles of aid distribution. Virtually every

10

The Politics of United States Foreign Aid

nation and territory has received some assistance. But there is “no readily recognizable correlation between amounts obtained, population, area, per capita income or development potentialities…given the changing and multiple objectives of US government aid programs, no such correlation could be expected”. One reason for the lack of correlation is the effect of defense expenditures on the distribution of economic aid. Under conditions of increasing defense expenditures financed by aid, per capita income might increase as economic aid declined in relative shares. Bilateral aid is government-to-government and is, administered by USAID, “the largest bureaucratic actor in the international economic policy machinery” (Cohen, 1977:55). However, the annual aid budget request is pieced together by DOD, USAID, and State, with Congressional review determining the actual allocation of policy resources. The bulk of US aid is bilateral (91.3% in FY 1986). Actual FY 84 expenditures for the DA, ESF and PL 480 programs amounted to almost 50% of the total “Foreign Aid” request as indicated in Table 1.1. Allocation of foreign aid monies is measurable by budget authority requested, actual budget authority (appropriations) or actual obligations during the year by USAID and related agencies. Wherever possible, obligations should be used (instead of “outlays”) because they reflect the allocative judgments of the Executive as modified by Congressional action (Wolf, 1960:77). First, the DA program provides grants and “soft” or concessional (low interest) loans for functional development in areas such as health, education and agriculture. At 10% commercial rates, LDC government borrowers would pay around 1%–3%. The purpose is to provide funding for long-term developmental projects on a non-country-specific basis. DA expenditures conform more closely to the notion of “economic” assistance”. As of FY 1979, 81% of DA loans and grants went to 68 nations meeting World Bank poverty criteria of $625 per capita income (CBO, 1980:16). However, others such as Maynes (1983) claim that only 19% of US “development” aid goes to low income countries. The key fact needed to resolve such disputes is a definition of “low income”. The Overseas Development Council (Lewis and Kallab, 1983:210) uses $400 per capita gross national product as the criterion; World Bank uses $625 PCI. Both views are probably correct though the more basic question is what happens to the DA funds after they reach the recipient. A low income country might reloan the funds at higher interest rates to the middle and upper classes; a wealthier country might not do so and reach a greater proportion of the poor than the poor country. Many DA projects are “technical assistance” transfers on a grant basis, i.e. trainingsupervision by foreign “experts” (usually contractors) whose salaries, together with any capital equipment, are paid for by the USAID project grant. While in the 1960s many DA projects were infrastructure, such as dams, roads, and power plants, after 1973 (“New Directions” legislation which will be discussed below) projects stressed: water supplies, sanitation, family planning, agroforestry, energy and education. As noted, some DA projects are loan-financed which means “offsetting receipts” to the US Treasury (not to USAID) for repayments of principal and interest, usually in local currency. Since loan charges amounted to about 10% of what the LDC would have to pay (in 1978) for a commercial loan, the grant component of loans amounted to about 90% (CBO, 1980:17). Yet Congress prefers to continue the “loan” myth out of concern for the fiscal responsiblity of borrowers. For FY 86, DA amounts to about 40% of the budget sub-function of “Foreign Economic and Financial Assistance”

The Political Economy of US Foreign Aid: Past and Present

11

which, as noted, excludes both ESF and military credit sales (Table 1.1). Assessment of real economic aid emphasis depends upon the substance of budget classifications. David Obey (D-Wisconsin), Chair of the influential House Appropriations Subcommittee on Foreign Operations, argues that of the $14 billion foreign aid program for FY 84, only about $2 billion or 14% goes into long-term development programs (Congressional Quarterly, March 16, 1985:498). The second thrust of the foreign aid program is also the largest: ESF. In contrast with long-term DA which consists of either soft loans or grants of technical assistance that usually do not break even financially (sanitation, transportation), ESF funds are cash payments for recipient budgetary deficit and balance of payments problems in select countries deemed politically important to US foreign policy interests. This is the “gray area” of US foreign aid, lying between DA and Security assistance. ESF provides shortterm resources to enable the recipient to make long term policy adjustments (fiscal and monetary) as well as project choices (infrastructure). Much of its effectiveness in providing flexibility depends on timing and goals, which are often less a product of tested theory than practical experience. Accordingly, much of its disbursement and use is guided by shorter term power considerations. In the period 1973–1983, the share of politically motivated foreign aid (ESF) doubled to 41% of the total pie (Maynes, 1983). Table 1.1: Presidential Request For Fiscal Year 1986 Budget Authority For Foreign Aid NATIONAL NEED: CONDUCTING INTERNATIONAL RELATIONS (Functional code 150; in millions of dollars) Major missions and programs 1984 1985 1986 1987 1988 actual estimate estimate estimate estimate BUDGET AUTHORITY Foreign aid: International security assitance: Foreign military sales credit: On-budget under current law Off-budget under current law# Military assistance Economic support fund Other: Existing law Proposed legislation Offsetting receipts Subtotal, International security assistance Foreign economic and financial assistance: Multilateral development banks International organizations Agency for International Development Public Law 480 food aid

1,315 3,503 712 3,389

4,940 3,147 805 3,841

5,655 1,311 949 2,824

5,779 524 970 2,883

5,901 262 991 2,941

110

214

108

−86 8,943

−93 12,854

−99 10,748

110 145 −105 10,306

112 278 −161 10,324

1,324 315 2,013 1,377

1,548 359 2,286 1,540

1,348 196 2,113 1,307

1,348 200 2,133 1,296

375 204 2,171 1,286

12

The Politics of United States Foreign Aid

Peace Corps 117 128 125 128 132 Refugee assistance 336 350 338 340 334 Compact of Free Association (Micronesia) 299 146 148 Other: On-budget under current law 80 95 96 99 101 Offsetting receipts −493 −459 −479 −604 −660 Subtotal, Foreign economic and financial assistance 5,069 5,847 5,343 5,085 4,093 Subtotal, Foreign aid 14,012 18,701 16,091 15,391 14,418 Conduct of foreign affairs: Administration of foreign affairs 1,392 1,897 1,848 1,802 1,855 International organizations and 580 545 554 583 559 conferences Other 53 65 68 70 15 Subtotal, Conduct of foreign affairs 2,025 2,507 2,470 2,455 2,489 Foreign information and exchange 798 897 1,118 1,296 1,228 activities International financial programs: Export-Import Bank 829 3,940 Foreign military sales trust fund (net) −801 734 978 957 1,067 International monetary programs 7,774 Offsetting receipts −84 −85 −87 −89 −90 Subtotal, International financial 7,718 4,588 891 868 977 programs Total, budget authority 24,553 26,693 20,569 20,010 19,111 ADDENDUM International Affairs programs if the Foreign Military Sales Credit program were on-budget direct lending: Foreign aid: Foreign military sales credit: On-budget under current law 1,315 4,940 55,655 5,779 5,901 Off-budget under current law 4,401 Other foreign aid 9,194 10,614 9,125 9,088 8,254 Subtotal, foreign aid 14,910 15,554 14,780 14,867 14,155 Other international affairs programs 10,541 7,992 4,479 4,619 4,694 Total 25,451 23,546 19,259 19,436 18,849 Source: US Office of Management and Budget, BUDGET OF THE US GOVERNMENT, FY 1986 (OMB, Washington, D.C., 1985)

The rationale for balance of payments support is to provide foreign exchange at reduced cost to encourage expansion of the recipient’s productive base, such as cement plants and textile plant rehabilitation in Egypt. As noted, the ESF program falls in the porous area between “special security interests” of the US and military aid. The program is designed

The Political Economy of US Foreign Aid: Past and Present

13

to aid the former without engaging the latter. But, of course, this is a highly theoretical distinction. CBO notes (1980:19) that since “money is fungible, the effectiveness of restrictions is questionable.” The exclusion of ESF from military activities is especially suspect when it is considered that from the mid-1960s-1970s, ESF was focused on South Vietnam. While the strategically important Middle Eastern countries of Egypt, Israel, Jordan and Syria received 1% of ESF funds in 1971, by 1979 they received 91% (1980:5)! The ESF program is a descendant of the Marshall Plan’s “defense support” funds of the 1940s. In 1961, with enactment of the Foreign Assistance Act (FAA), this type of aid was formally designated as “Supporting Assistance.” In 1971, it was renamed “Security Supporting Assistance” and placed under the FAA title concerned with military rather than economic aid. In 1978, the International Security Assistance Act replaced Security Supporting Assistance with the Economic Support Fund and “Peacekeeping Operations” accounts (made famous more recently by their use in Beirut and Grenada). Significantly, as noted, ESF is now included in the annual budget under “International Security Assistance” rather than “Foreign Economic and Financial Assistance” (see Table 1.1). Due in part to increased Middle East tensions, the FY 87 ESF request rose from the estimated $2.8 billion to $4.09 billion (DOD, 1986:69). Thus, in practice ESF funds are closely related to military and security objectives. Though much of military aid is distributed and accounted for in secret, general patterns can be discerned through the improved DOD information, particularly the FY 87 “Congressional Presentation For Security Assistance Programs”. Since 1953, the US has provided about $100 billion in military aid to 113 countries, 60% of which has been in the form of grants (USAID, 1984:4). The military assistance effort consists of three major grant/credit programs: (1) Military Assistance Program or MAP grants which provided defense-related “surplus” services to Korea, Vietnam and the Philippines, and now provides “new” equipment and services amounting to $782 million actual FY 86 budget authority (USAID, 1986: Table 2), (2) Foreign Military Sales Credit (FMS) which finances recipient country purchases of US military equipment via loan guarantees or credits at both market and concessional rates. As indicated, the FMS program is the highest single foreign aid item, amounting to 43.1% of the FY 86 request or 39.3% of actual FY 86 budget authority. (Table 1.1 and USAID, 1986: Table 2), (3) International Military Education and Training Program (IMET) which grants instruction and training to foreign military personnel. The ESF program allocates assistance by country/region which generates greater political support against budget cuts. For FY 86, the ESF request ($2.82 billion) is about 20% of the total ISA request; the MAP program seeks $949.3 million or about 10% of ISA funds (OMB, 1986: Appendix, 1–06). The ISA component of all foreign assistance is roughly 50%. The geographic distribution of US military aid from 1946–1981 is indicated in Figure 1.1. From 1962–1984, the FMS program was “off-budget”, meaning that outlays were to some extent hidden from public scrutiny, and that they were not accounted for as part of the federal budget deficit. At least $31.2 billion (USAID, 1984:58) escaped notice under this loophole. Until FY 85, loan guarantees were not considered part of budget authority (BA) or outlays (0). According to Ippolito (1984:58), the FMS program worked like this: “..DOD issues loans to foreign governments to finance their purchases of US military equipment. The foreign government then takes the loan guarantee to the Federal Financing Bank, which purchases it for the full amount of the guaranteed loan principal. “The borrower thus

Figure 1.1: US Foreign Military Assistance, 1946–1981 Reprinted with permission of Overseas Development Council from Lewis and Kallab, 1983.

The Political Economy of US Foreign Aid: Past and Present

15

receives funds in the form of a direct loan from a federal source. But beginning in FY 85, FMS credits were moved from “off” to “on” budget status; in FY 86 all Federal Financing Bank activities may be “on budget” as part of the continuing quest for deficit accountability. FMS is now correctly called a “direct loan” program and requested BA for FY 86 is $5.7 billion (OMB, 1985:5–17). In the past, creative accounting made much of the FMS credit program invisible and reduced the apparent size of military assistance. On the other hand, it could be argued that the “real” contribution of military to total foreign assistance has been smaller than indicated. For instance, in the 1950s and 1960s, DOD charged the MAP program for the “full cost” of modern replacements and transferred goods at original acquisition costs (Asher, 1961:57). The final thrust of the bilateral program is Food Aid. This program has always been in the swirl of controversy and therefore an easy target for critics of US foreign aid. The Food For Peace program was established under the Agricultural Trade Development and Assistance Act of 1954 (P. L. 480). The goals of the program were: trade expansion, disposal of US surplus agricultural goods, agricultural price stability, support of US foreign policy, and foreign economic development. In the 1960s, the goal of surplus disposal was achieved as P.L. 480 shipments ranged between 20%– 29% of total US agricultural export value! (CBO, 1980:21). Benefits to US agricultural producers declined later in the wake of increasing LDC commodity production, partly the result of DA program successes in agricultural development. According to Easterbrook (1985:76), the enormous LDC increases in wheat, rice, and cotton production (everywhere except Africa and Japan) that have destabilized US farmer prices and decreased US exports “can be attributed partly to the spread of modern technology and farming methods—progress that the US has encouraged and in many cases funded and supervised.” P. L. 480 shipments declined to only 4% of US agricultural exports in FY 1978 (CBO, 1980:21). The P. L. 480 program has historically benefited US exporters and LDC importers, who are in effect paid by the US to buy and resell commodities locally. The criticism that the US accumulates vast sums of foreign currency from the program ignores the fact that the US paid for the currency via a grant designated as a “loan” to please the banking mentality of Congress. The theory was and is that a “counterpart” fund procedure, whereby recipients “paid” for food, would increase the efficacy of this foreign aid program (Asher, 1961:9). But in fact the main beneficiaries have been: recipient importers, recipient governments (in legitimacy gains) and US exporters. Famine and poverty have been largely secondary to foreign policy and stability objectives. That Food Aid is disbursed to many of the same countries that receive large amounts of ESF funds, Egypt, South Korea, Israel, and South Vietnam up to 1975, suggests a political distribution of funds. The more important question is whether Food Aid works as a disincentive to local producers. This effect has been documented by Shepherd (1985), as noted, in Africa, which actually increased dependence on commodity imports. But this finding turns on the local price effects of commodity imports. If the US imports add to the market, they have an “additionality” factor and decrease prices for local producers. But if prices are already reduced by local subsidies, or imports replace commodities that otherwise would have been imported, then the program cannot really be blamed for reducing local agricultural production. CBO, citing Sarris et. al. in Cline (1979) suggests that current research indicates only partial additionality (CBO, 1980:26). Many former Food Aid recipients, such as Brazil, Chile,

16

The Politics of United States Foreign Aid

Peru, Taiwan, Colombia, Japan and Spain are now commercial food buyers (1980:4). Finally, though far less than the congressionally-mandated 75% of concessional sales went to low-income countries (1968–73), the FY 80 allocations indicated that 80% went to low income countries (1980:24). For FY 86, $1.30 billion is requested for Food Aid, a 7.5% decrease from FY85 (Table 1.1). In addition to the four bilateral programs discussed, the US also contributes roughly one-third of the funds received by multilateral development banks (MDBs) such as World Bank and Inter-American Development Bank. US contributions to MDBs take two forms: (1) “paid-in” capital or funds authorized and appropriated by Congress which are disbursed to MDBs mainly for soft loans to LDCs, and (2) “callable” capital (which are like provisions in some US municipal bonds that give the issuer an option to redeem the bonds prior to maturity) or loan guarantees not disbursed to MDBs. The US Treasury indirectly guarantees the bonds MDBs issue in world financial markets to raise capital. Proceeds from bond sales are lent to LDCs on “hard” terms. The funds are called only if the MDB cannot meet interest and amortization payments on outstanding bonds. Though the Reagan Administration favors on-budget treatment for the FMS loan guarantee program as noted, here the Administration favors off-budget treatment. The rationale seems to be that although FMS “buyers” have defaulted in the past, appropriations for callable capital have never resulted in disbursements for MDB-recipient defaults. Further, MDBs tend to fear that even though off-budget treatment can accelerate passage of the annual foreign aid appropriations bill, it could also lead to less detailed issue analysis. MDBs also fear that off-budget treatment would diminish buyer confidence, forcing MDBs to pay higher interest rates on loans which would increase interest rates to LDCs. The real threat is that removal from the appropriations process could ultimately reduce loan volume for MDBs. Congress tends to fear the loss of funding control ostensibly available in bilateral aid when donating funds to MDBs. The FY 87 Reagan Administration strategy seems to be a “two-edged sword”, proffering increased MDB contributions in exchange for additional US programming leverage within the MDBs. The ploy is that MDB rejection would allow the US to decrease support. However, the effects of attempts to program aid via insiders are well-known to practitioners. Often institutional roles in MDBs are so narrowly defined that even the most radical newcomer soon turns quite passive and becomes an advocate of the bureau’s program. Shepherd (1985:46) suggested that “in a dramatic shift of American foreign policy, the Reagan Administration is cutting US funds to multinational donors. It will end all US contributions to the World Bank’s International Development Association (IDA). “The IDA is the bank’s soft loan arm. Nevertheless, the FY 86 Reagan budget states that “the US will honor existing commitments to MDBs including the IDA” (OMB, 1985:5– 18). For FY 87, the US requests $1.57 billion which is an increase of $169.4 million or 12% more than the FY 86 request of $1.4 billion indicated in Table 1.1. However, within the MDB line-item, it is clear that the Reagan Administration has cut back funding to IDA (Rowen, 1986). The contemporary foreign aid program (1981–1986) is founded upon earlier academic notions of “political development”, and it remains to be seen whether they will be any more successful. The Reagan approach stresses aid (and trade) to expand private sector activities in LDCs as a vehicle for stability, growth and freedom. Actual changes in USAID resource allocation for these purposes to date have been marginal. But Newfarmer (1983:118)

The Political Economy of US Foreign Aid: Past and Present

17

notes that “as was quite evident with the New Directions legislation of 1973, it takes a considerable time for the bureaucracy to design and implement new programs. “He notes further that most existing foreign aid programs already involve the private sector, such as agricultural projects with small farmers. “USAID mission directors in some instances ‘repackage’ existing programs to suit the new Washington rhetoric: farmers suddenly became rural entrepreneurs” (1983:118). However, AID has begun earmarking funds in a selected number of missions for the private sector. “If this becomes a common practice and excludes non-commercial agriculture (as it has in some cases), the actual shift in resources could be dramatic” (1983:118). The Reagan approach favors the private sector because it is critical of the size of LDC governments and their propensity to overregulate (1983:127). Though the presumed relationship between size and service is unsupported by most studies, i.e. greater development leads to greater expenditure on public goods and services (1983:127), it is clear that in practice, LDC governments do not effectively provide public goods and have not “developed” along with the rest of society. Trebat (1983) found that Brazilian state-owned firms are generally efficient, and this finding has been reconfirmed by other researchers. But most LDC public enterprises and general fund ministries are not very efficient and do not function as a stimuli or catalyst for industrial growth or even social development. They are rational in seeking to expand agency budgets (military and civilian) but are hamstrung by lack of managerial control, and the will or capacity to create appropriate conditions for development. That is, in most cases they are parasitic rather than entrepreneurial for the public good. For example, the Uruguayan GDP has been declining between 5%–10% per year since 1980. At the same time, yields from the value added tax (VAT) which produces 40% of total tax receipts, have been declining each year. To compensate, the state increases VAT and corporate tax rates as yields continue to decline. The seven major state monopolies (state enterprises) have all been running deficits of 1.5% of GDP or greater since 1982 (IMF, 1985:26). Some of these problems may be attributable to domestic adjustments from changes in the international economy. But it is clear that focus on the private sector could produce greater growth and efficiency than currently exists in many LDCs. Unfortunately, the linkages between private development strategies and political freedom are not precisely clear. First, the linkage between growth and freedom is indirect and may be just as likely to produce “bureaucratic authoritarian” (BA) regimes (or permanent military regimes seeking economic growth) or “new oligarchies” as in Chile, as political democracies (Guess, 1984; Collier, 1979). Nor does the US experience with slavery, abolition, and workers’ rights to collective bargaining suggest that political participation follows automatically from a free private sector (Newfarmer, 1983:134). It is also unclear whether the private sector is any more innovative or risk-taking than the public sector, especially for new projects. For example, a recent project to export textiles with new technology from Europe had to be financed by government subsidized loans instead of private banks (Atlanta Constitution, 1985). Next, the Reagan strategy would strengthen the LDC private sector by generating independent sources of intergovernmental power—a US federal system and pluralist bargaining model exported to LDCs. But historically, wealth and power have been concentrated in LDC elites such as landowners and their military lackeys. Hence strengthening the private sector means subsidizing groups hardly amenable to expanded

18

The Politics of United States Foreign Aid

political participation and democracy. To strengthen small farmer and business groups often requires “tunnelling under the ministry”, or if the military commands, around the ministry. But such strategies are dangerous and tend to be viewed by US friend and foe alike as interference with sovereignty, albeit the sovereign right to be exercised is usually oppression of local groups. Hence the dilemma for the new private sector foreign aid as well as the old—recipient expenditures are largely out of US control and to rectify this with tight conditions and controls invites criticism for neo-imperialism. Whether the new Reagan strategy succeeds, it at least has the virtue of simplicity. It will clash head on with LDC governments (Haiti, the Philippines, and South Korea in the future) and cleanly succeed or fail.

The Evolution of US Postwar Foreign Aid Like any policy, foreign aid is a product of its history. This implies lengthy conflict over means, ends and the measurement of both. The evolution of US foreign aid can be traced through five overlapping stages. These stages conform to the predominant purposes of the program as they changed after World War II to the present. They are: (1) the Post-War Relief period, 1943–1948, (2) “Containment” and “Frontiersmanship”, 1949–1960, (3) the “Alliance for Progress” period, 1961–1972, (4) New Directions legislation, 1973–1980, and (5) the Private Sector Initiative, 1981 to 1986. This is not the only possible classification. USAID, for example, distinguishes four periods: (1) Post-War Relief, 1946–48, (2) Marshall Plan, 1949–52, (3) Mutual Security Act Period, 1953–1961, and (4) Foreign Assistance Act Period, 1962-present. The foreign aid statutes highlighted by the AID classification will be discussed within the fivefold classification to be used here. Foreign aid policy evolved through these stages as a product of conflict over three general issues. While these issue determinants will be developed more fully in the third chapter under the “Bureaucratic Role Conflict Model”, it is important to introduce them at the outset of the discussion. First, US foreign aid responses have always been limited by cost and return considerations, or “profitability” objectives. The notion of financial profits from foreign aid is largely mythical but serves as a powerful political force in organizing the program each year. Profitability considerations are evident in the congressional preference for businesslike loans over grants, despite the fact they are usually soft and rarely repaid in full. Second, Foreign aid has always been used as the major “peaceful” means to respond to perceived communist security threats to political stability around the world. As noted, the coincidence of foreign aid and security objectives in places such as Vietnam results in an apparent failure for foreign aid despite the miniscule amounts actually spent on security projects there. Defense considerations have been more important than costs (or benefits); both defense and costs have been deemed more critical in the determination of foreign aid resource allocation than the third issue—humanitarian need. Poverty and hunger problems are rarely technical questions and to the extent that they are, the US tends to rely on support for MDBs as the appropriate remedy. Figure 1.2 indicates the coincidence of US foreign aid objectives with regional security issues, suggesting strongly that other considerations, like cost and welfare, are secondary. But it wasn’t always this way. Prior to the first phase of Post-War relief, cost considerations had kept the US out of the foreign aid business. For, there was ample cautionary precedent

The Political Economy of US Foreign Aid: Past and Present

19

from the effects of the dynamic European statecraft subsidies in the 17th and 18th centuries, and foreign loans in the liberal diplomacy of pre-World War I (Liska, 1960:36). To many observers, the politicization of French finances by its aid to Russia in a disastrous alliance attempt was extremely significant for developing a coherent US foreign aid rationale. That is, to prevent a Russian alliance with Germany in the 18th and 19th centuries, beginning with Louis XIV the French subsidized Russia to build an alliance and develop debtor resources. Only it didn’t work. France became identified with the reactionary Czars; the Russians didn’t use the aid as agreed, i.e. didn’t build a railroad to Poland despite the condition in the loan. Whereas early French loans presumed profitable returns, later ones merely hoped for political instead of financial returns. The tables turned ultimately when the recipient made political support of France contingent on additional loans! With fresh lessons of how difficult it could be to influence the policies and political behavior of lesser states via loans-grants and alliances, the US policy after WWI was designed to support only foreign “aid” projects that would bring handsome profits. For example, since wars and foreign aid gifts were viewed as losing propositions, President Hoover had denied reconstruction credit to Europe (Montgomery, 1967:30). President Harding had also tried to restrict the use of private loans to prevent European rearmament. In ignoring earlier Russian behavior in substituting aid resources (now called the “fungibility” problem), US policies both failed to prevent rearmament and encourage investments for desirable purposes in Italy, Germany, and later Japan (1967:31). It took another war, as Montgomery notes, to produce a policy of using “public capital for diplomatic purposes.” During WWII, the “Lend-Lease” program became the forerunner of the future foreign aid program in that it recognized that a “US contribution to the common cause of national survival could not be treated merely as a commercial transaction” (Morley, 1961:6). In “Dollar Diplomacy”, the US simply called grants “loans” to preserve the aura of businesslike conduct. If “bankable projects” were financed by hard loans, then projects with low break-even potential were financed by soft loans, or “gift-like loans”. The first phase of US foreign aid was premised largely on humanitarian relief. As noted by Montgomery, the US Congress had experience in this long before postwar foreign aid was conceived, such as funds sent to aid Venezuelan earthquake victims in 1812 (the shipment was ominously seized by pirates and distributed to counter-revolutionaries!) (1967:8). The US experience with foreign aid began with contributions to multilateral agencies for humanitarian purposes. From its inception in 1943 to its discontinuance in 1947, the US contributed 70% of the funds to the United Nations Relief and Rehabilitation Administration (UNRRA). The purposes of UNRRA were to meet the severe needs and prevent starvation in the former Axis-occupied countries of Southern Europe, China, Korea and the Philippines. UNRRA was a temporary agency that pre-dated the United Nations itself. In 1947, US support ceased as Congress increasingly perceived aid to be directed for “political purposes” in supporting “Soviet puppets” (Morley, 1961:13). Ironically, Congress had multilateralized aid to ensure cost control—preventing the drain of US gold reserves. But as Soviet domination of Eastern Europe became a fait accompli, Congress preferred political to financial responsibility and withdrew US support. US post-UNRRA relief continued in grant form under such programs as Government Relief in Occupied and Liberated Areas (GARIOA), which granted funds to Germany, Japan, Italy, Korea, China and the Philippines (totalling over $6 billion) (Morley, 1961:14).

Figure 1.2: Distribution of US Foreign Assistance By Region, 1946–1982 Reprinted with the permission of the Overseas Development Council from US Foreign Policy and the Third World: Agenda 1983, John P.Lewis and Valeriana Kallab, eds. (Praeger Publishers for the Overseas Development Council; New York, 1983).

The Political Economy of US Foreign Aid: Past and Present

21

There were, of course, strategic military and cost considerations that influenced the design of foreign aid in this period. But this was probably the only period of US history where “social justice” considerations dominated foreign policy. Accordingly, public support for the tool of foreign aid was high. Two problems led to the shift from relief to reconstruction and defense considerations during the second or “Containment-Frontiersmanship” period of 1948–60: (1) the failure of European economic recovery, and (2) Soviet-led penetration of Europe and elsewhere in the Third World. It was soon recognized that the conservative US “punitive peace” policies which provided relief and financing of US exportable surpluses alone would not succeed in restoring world production. Dismantling and decentralizing German and Japanese industry (the continuing fear of rearmament left over from WWI) was both hindering recovery and keeping German and Japanese paupers on the dole. The Morgenthau Plan had actually called for making Germany and Japan into low-level subsistence economies (Morley, 1961:14). Had this been implemented, early US foreign aid experience could have been credited with stimulating intentional underdevelopment! By contrast, the post-war relief plan was to end relief grants and stimulate long-term development via guaranteed loans to reduce trade barriers and increase private investment (Asher, 1961:47). But since Europe was in no position to obtain enough dollars to purchase commodities for economic development, war devastation and punitive policies still hindered recovery. Secondly, in response to Soviet expansion into Southeastern Europe and the Middle East (Greece, Turkey, and Iran), President Truman formulated a policy of communist “containment” in 1947. Though the Rio Pact of 1947 was an earlier deviation from postwar isolationism, the “Truman Doctrine” was boldly interventionist and marked the official end of US isolation from world affairs, indicated by earlier policies such as the 1823 Monroe Doctrine (“old world despots leave us alone”). The US, under the new doctrine, would insure a “balance of power” in world affairs primarily by providing economic assistance or foreign aid to Europe (Spanier, 1983:31). By providing aid in this fashion, the security threat of a collapsed European recovery would be avoided and the US would surround its enemies with allies and trading partners. Following British withdrawal of financial assistance to Greece and Turkey, the US decided to use foreign aid for the first time to counter Soviet indirect aggression via the Truman Doctrine (Morley, 1961:15). Supporters of US military assistance often refer back to the success of aid to these allies in preventing Soviet expansionism. However, the success of US aid to the Greek army in defeating guerrillas in 1949 has also been attributed to the Yugoslavian break with the Soviet bloc in mid-1948, resulting in a cut-off of important aid to the Greek guerrillas (1961:16). The major non-military vehicle for the Truman Doctrine of “containment” was the Marshall Plan. Since the Marshall Plan was the primordial foreign aid experience for the US and constituted its most dramatic success to date, a few words should be spent on the subject. While defense of Europe was an important consideration in containment and also in the Marshall Plan, it was also predicated on the calculus that a good investment now ($12 billion in annual appropriations for grants) would lead to large long-term benefits in the form of recovery. The calculus then was similar to that currently confronting international banks with large Latin American loans on the verge of default. Should the banks lighten their burden by “writing down” some of the $350 billion owed over the next five years? If

22

The Politics of United States Foreign Aid

they don’t, would not ‘write offs” from a global recession and failure to recover be greater than if they wrote them down now? (Rowen, 1985). The Economic Cooperation Act of 1948 provided for grant aid conditional on regional integration and cooperation. The Plan or European Recovery Program was to duplicate the US domestic mass market in Europe by eliminating tariffs, quotas and export licenses. Since skills were already at a premium throughout Europe, recovery was expected to follow from transfer of capital grants (Britain, France and West Germany received about $6 billion), replacing bombed but outmoded factories with more advanced American technology. The US supplied the commodities, technology and funds; the Europeans did most of the planning and execution. In contrast with most US foreign aid programs today, the effort was actually cooperative: between donor-recipient and between recipients. The recipients directed field operations with US capital and the results were spectacular. Further, Marshall Plan grants never amounted to more than a fraction of recipient GNPs (Morley, 1961:23). In contrast with some of the counter-productive dole relationships of today, e.g. Israel where US foreign aid in FY 85 amounted to 19% of the local budget (Meyer, 1985), the Marshall Plan provided mostly a psychological boost and stimulus for integration. The contrasting results from the previous “punitive-dole” policy were dramatic. According to Morley (1961:22) in four years industrial production was up 35% from pre-war levels, agricultural production up 10%, intra-European trade up 70%, inflation reduced, internal financial stability increased, and no communist takeovers. Despite such success, the Plan was viewed as a special case. The Truman “Point 4” inaugural address in 1949 and the subsequent Act for International Development of 1950 outlined the basis for future foreign aid via “technical assistance” in commodities, equipment, information and expertise to be financed by loans and grants (Asher, 1961:49). Implicitly, capital grants were reserved for conditions of exceptional local expertise (post-war Europe); future appropriations would target transfer of personnel and US goods through loans (and “tiedaid” grants) which, again, seemed more businesslike to Congress and the average American balance sheet mentality. One of the major lessons of the Marshall Plan was that its successes have not been transferable to the Third World. The US began in 1949–1950 to gather this in the Far East. With the collapse of Nationalist China and the establishment of a Communist government on the mainland in 1949, followed by the North Korean attack on South Korea in 1950, the balance of power in the Far East had shifted “gravely” against the US (Spanier, 1983:51). This was not totally unforeseen. The China Aid Act of 1948, providing $125 million in military aid and $275 million in economic aid was passed as part of the Marshall Plan legislation and grant military aid to Iran began with the Mutual Defense Assistance Act of 1949 (Morley, 1961:32, 37). The result was a blurring of different types of foreign aid and:” ‘defense’ became the umbrella for most forms of American aid. The Mutual Security Act of 1951 became the first of a series of annual measures covering, in a single act of Congress, the continuation on a world-wide basis of military aid, economic aid, and technical assistance” (1961:49). Whereas foreign aid to Europe was based on a jointly conceived economic recovery plan, foreign aid to the Third World proceeded largely on the basis of inter-locking anticommunist alliances, beginning with the North Atlantic Treaty Organization (NATO) in 1949. In the presidential campaign of 1952, the Republican candidate Dwight Eisenhower

The Political Economy of US Foreign Aid: Past and Present

23

fanned patriotic flames with his proposal for not mere “containment” or indefinite peaceful coexistence with the Sino-Soviet bloc, but now for rolling them back behind frontier lines, usually national borders, or “frontiersmanship”. “The walls would then come tumbling down, the enslaved peoples would be liberated and Soviet power would be forced to retreat. The world would once more be safe for democracy” (Spanier, 1983:75). Unfortunately, the Eisenhower response to Soviet repression of Hungarians (1956) and East German cities (1953) was limited to harsh rhetoric and food packages. Containment had merely been recycled in the form of new alliances in Asia (SEATO, 1954), and the Middle East (METO or Baghdad Pact, 1955). Clearly, the emphasis of US foreign aid had shifted from economic to military predominance in the mid-1950s. According to Morley (1961:29) the military assistance share of all foreign aid was 24% in FY 51, 38% in FY52, and about 70% in FY 56. The Rio Pact still served inter-American security interests, and during this period the US provided “political aid” in 1954 to the new Guatemalan regime after the “anti-communist coup” (Morley, 1961:46). It was soon evident, however, that the lack of indigenous regional nationalism in the Far East, Middle East and Latin America would make duplication of the European aid model difficult. Thus, foreign aid was dragged along an essentially defense-security plane from crisis to crisis, providing technical assistance for very uncertain gains often in the middle of civil wars and revolutions. When defense aid increased, economic aid also increased, almost as an afterthought. This pattern will become more evident when we examine specific country cases by region in Chapters 3–6. With the Suez invasion of 1956, foreign aid was again to be the instrument of foreign policy. Here, the “power vacuum” left by the British was filled by Nasser’s seizure of the canal. The US halted the regional skirmish but miscalculated both Egyptian and Soviet moves and thus suffered a severe propaganda setback. The immediate result was the formulation of the Eisenhower Doctrine of 1957. This authorized the President to give economic and military aid to any Middle Eastern nation deserving it and to use US armed forces to protect against communist aggressors (Morley, 1961:38). Despite their escalating costs, both foreign aid and foreign policy were unsuccessful in this period in both the Middle and Far East. Still, the use of Third World defense alliances for “leverage” through economic and military assistance continued. President Truman had used the term “defense support” to mean “economic aid as the necessary underpinning of militarization” (Morley, 1961:30). As noted, this conception was institutionalized via the Mutual Security Act of 1951. Despite the military dominance of foreign aid policy in the Containment-Frontiersmanship period, owing largely to Soviet expansion and US inability to deal with Sino-Soviet, Middle Eastern and Far Eastern actions diplomatically, foreign aid also expanded in the economic direction as well. Part of the reason was the Soviet initiation of its own technical assistance program to LDCs in 1953, beginning the East-West foreign aid rivalry (1961:43). Again, reminiscent of the volatile and self-defeating European “loan wars” of the 19th century, the US and USSR now employed foreign aid to gain the support and cooperation of “uncommitted countries”, to prevent them from lining up with the other side in the world power struggle (Morley, 1961:44). To counter the Soviet’s mostly soft loan program (for capital projects, repayable in borrower currency), the US developed such programs during this period as Food For Peace (1954), the Development Loan Fund (1957) and support of MDBs (1960). The antecedent of Food For Peace was the Mutual Security Act provision in 1953 that at

24

The Politics of United States Foreign Aid

least $100 million in foreign aid funds be used to buy agricultural surplus for sale abroad, payment in foreign currencies (Asher, 1961:12). P.L. 480 reduced the costs of storage for food surpluses since, according to Hudson (1971:90) the Commodity Credit Corporation (CCC) would have purchased them anyway. The Development Loan Fund was created to provide capital assistance loans, 75% repayable in borrower currency on soft terms. By 1960, the US was worried about the costs of foreign aid on US balance of payments deficits. Whether the real costs outweighed the benefits is debatable since many resources returned to the US. Hudson estimates benefits of food aid exceeded costs by $5.6 billion from 1955–1969 (1971:91). Nevertheless, the perception of unprofitability was there, derived in part from the colonial heritage of “primitive fiscal orthodoxy” (Kaplan, 1967). He suggests that by the mid-1950s, most US aid was programmed to meet local currency requirements rather than satisfy foreign exchange and balance of payments shortages. By contrast, the Marshall Plan had correctly focused on capital shortages and balance of payments instead of US repayment considerations. Post-Marshall Plan foreign aid policy refused to let local officials experiment with the notion that, for example, “properly managed budget deficits were permissible and might even be desirable”. The strict injunctions of colonial administration about balanced budgets, which assumed surpluses for the donor or “metropole” (Kaplan, 1967:289), was in practice was modified out of compromise and necessity. Irrational fear of gold drainages from the US together with increasingly competitive Soviet economic aid brought the US to the solution of “sharing the burden” (not simply humanitarian) for “developmental” aid. In practice, support for the International Development Association (IDA) of the World Bank (1960), the Inter-American Development Bank (1960) and the UN Fund For Africa (1960) meant expansion of soft loans from the West to LDCs (beyond that appropriated by Congress for the Development Loan Fund and PL 480 local currency repayment). Low-interest loans were thus available as a separate line from hard loans on commerical terms from the World Bank. Military aid continued on a grant basis virtually unquestioned in regard to profitability since it was more clearly a foreign policy instrument than foreign aid. Still the conflict persisted as Congressional “guardians” such as the Appropriations Committees sought to maintain US balance of payments surpluses and overall fiscal integrity by such means as financing foreign exchange costs only (as opposed to recipient costs and recurrent budgetary needs), “tied-aid” (1958), and the Cooley Amendment to PL 480 in 1957 (requiring that 25% of foreign currencies owned by the US be used for private US companies operating in LDCs) (Asher, 1961:14). On the other hand “advocates” for economic development and humanitarian concerns, including those who saw threats to US security interests in excessively hard loans, sought softer terms for LDCs. The conflict between purely fiscal and welfare perspectives persists today, most notably in the issue of Latin American debt repayment strategies. Hence, the second period of US foreign aid ended largely where it began—as a “succession of improvisions for the most part stimulated by Communist activities.” (Morley, 1961:51). Foreign aid remained “amorphous”, “provisional”, “casually operated” and completely dependent on annual authorizations by Congress. Liska (1960:53) suggests that US foreign aid in this period represented the French model or financial aid to economically risky and weak governments largely for political purposes, as well as the German model of sending

The Political Economy of US Foreign Aid: Past and Present

25

their own technicians to develop recipient productive resources, rather than the cautious and conservative British model. The third phase, Alliance For Progress (1961–1972) marks the first of three phases during which conscious efforts were made by key US policy-makers to forge coherence in foreign aid. This period stretches from President Kennedy’s centralization of foreign aid policy through integration of USAID within the Department of State in 1961, to the early Nixon years and the Pearson Report of 1970. Ironically, the emphasis of the previous phase on defense culminated in a shift to multilateralism and economic development about 1961. The Kennedy– Johnson years began exuberantly with a centralized administrative mechanism to export US bilateral concerns for economic development and political democracy via foreign aid. By the end of the Alliance phase in 1972, President Nixon had again shifted emphasis from bilateral to multilateral aid, not from fear of draining US gold reserves which had bothered fiscal conservatives in the Eisenhower and Truman administrations, but more pragmatically to reduce domestic political opposition to his programs by submerging them in multilateral anonymity (Hudson, 1971:198). The years 1961–1972 were characterized by contradictions and inconsistencies from the attempt to use foreign aid as the vehicle for export of American liberal political ideas. This was less a conflict between the explicit military-economic or finance-welfare approaches which characterized the Containment phase, but rather a dissensus on the meaning and application of the “liberal” ideas themselves. Nevertheless, the emphasis shifted early in the Alliance from defense to socioeconomic objectives. Total economic aid commitments increased from an annual average of $2.5 billion (1956–1960) to an average of over $4.0 billion per year (1961–1963). “In the fifties, military assistance and that part of economic aid which financed defense support or other security objectives had been roughly double the level of developmental aid; during the first half of the sixties, that ratio was reversed” (Nelson cited in Packenham, 1973:60). President Kennedy initiated the Alliance For Progress in March, 1961 as an “interAmerican effort for social and economic progress through democratic politics in the hemisphere. Democracy was a very salient goal for him from the beginning” (1973:70). The US foreign aid program was shifting away from its traditional security orientation toward developmental and welfare concerns. There were significant changes during this period in legislation, organization and concepts. First, the Foreign Assistance Act of 1961 replaced the Mutual Security Act as the “fundamental law for American economic and military aid” (1973:60). “Like its predecessor”, according to Fisher (1983:31), “it was reauthorized by Congress annually, because of its ‘size and uncertainty’. “This meant that Congress was attempting to control excessive presidential discretion in foreign affairs by this legislation, and provide themselves the flexibility to adjust foreign aid to rapidly changing conditions each year. Second, in 1961, USAID was established, absorbing the functions of the International Cooperation Administration (ICA) and the Development Loan Fund (DLF). It was organized along “geographic” lines by bureau, rather than purely functional lines such as agriculture (Packenham, 1973:61). However, competition for resources has increased along functional lines as well. By FY 86, “Functional Development Assistance” such as health and education amounted to 70% of the total DA request or $1.4 billion out of $2.0 billion (USAID, Congressional Presentation FY 87, 1986:6). The largest functional program is

26

The Politics of United States Foreign Aid

agriculture (43.6% of the DA funds) and Latin America is the largest regional recipient (29% of the DA funds) as of mid-1986. Third, several new concepts were introduced in this period. Behind the Alliance was the fundamental Lockean notion of liberal democracy. This translated into: self-help (loans again, not grants), social change (progressive tax reform, land reform, voluntary organizations, and greater political participation), the notion of democratic “evolution” (economic growth would push the political system toward democracy) (1973:63) and finally, the notion of “country programming”, or specifying sectoral and project relationships on a national scale. These, of course, were not totally new concepts. The evolutionary concept had been gleaned from the comparative politics literature long before and persists today in the repackaged form of “private sector”-led economic growth as the engine for democratic evolution. But their simultaneous application to foreign aid policy was new; nor had any US administration had the opportunity to react to recipients that bought some of the new aid package but rejected the rest! The notion of self-help is simply that aid will serve as catalyst for local self-sufficiency; and will contribute to its own elimination. Loans have rarely encouraged this, unless recipient export revenues can be made to balance debt payments. Unfortunately, foreign aid alone cannot usually generate sufficient productive capacity. Trade is usually required as the outlet for generated capacity, as indicated by the case of modern Brazil with both massive debts but massive capacity to repay and hold the IMF at bay (Boston Globe, 1985). In practice, loans often became grants, as noted, since if they were not “unbankable” in the first place (lacking economic justification or unlikely to financially break even) they would have been placed on harder terms. But hard terms are unaffordable by poor countries. Conversely, since many grants had strings attached, such as purchase of US goods, the real distinction between loan and grant was often unclear. The social change thrust also had mixed success. Often the implication of real social change, such as an election which produces anti-US leaders (Guess, Miami Herald, 1983) clashes with US foreign policy objectives and is often countered in the Cold War tradition by supplementary policies like “destabilization” in Chile, which raises public suspicion of the entire foreign aid program. Successful exportation of democracy via foreign aid, often ironically serves as the basis for criticism of the program, i.e. the evaluative dilemma noted earlier. Gurtov (1974:49 notes that “Kennedy had to overcome resistance within the AID and State Department bureaucracies in order to push through assistance programs to radical nationalist governments..” (in Guinea and Ghana). However, where social change turned against the US during this period, liberal principles were often jettisoned for Cold War reactions. Gurtov (1974:49) notes further that “Kennedy’s flexibility on relations with radical nationalist governments stands in marked contrast to his response to revolutions and civil wars, a response motivated by political expediency, considerations of economic advantage, and the imposition of a cold war framework around Third World political turmoil. “In the early 1960s, JFK feared that Cheddi Jagan was about to make another Cuba of British Guinea. Kennedy pressured the British, who changed the electoral system to proportional representation, effectively excluding Jagan’s party from the coalition government (Packenham, 1973:81). The Johnson years were even more ambivalent on “social change” and the direction of US foreign aid policy. On the one hand, LBJ followed the Cold War approach of economic

The Political Economy of US Foreign Aid: Past and Present

27

retribution and counter-insurgency to exclude subversives from participation. The Mann Doctrine of 1964 signified abandonment of JFK’s “export democracy” approach, and indicated that the US would no longer punish military juntas for overthrowing democratic regimes. For example, LBJ approved loans to the Duvalier clan by the Inter-American Development Bank ($2.5 million in 1964) and the new Brazilian junta within 12 hours after overthrow of President Goulart in 1964 (1973:96). LBJ’s “liberal” notions on Vietnam translated into a plan for a Tennessee Valley Authority (TVA)-Marshall Plan type organization centered around the Mekong River, to be funded by the new US-backed MDB: The Asian Development Bank (1966) (Packenham, 1973:91). On the other hand, in an apparent contradiction to the Cold War approach, the LBJ Administration reacted against the “buckshot effort at general development” whereby foreign aid translated into: highways, public administration, and agriculture in an unplanned way. Title IX was added to the Foreign Assistance Act in 1966. This directed USAID toward “political development” objectives. The political development mandate translated into “assuring maximum participation” in the process of economic development through encouragement of “democratic private and local institutions”, such as cooperatives, labor unions, trade and related associations and community action groups (1973:100). Title IX was LBJ’s attempt to demonstrate innovative capacity equal to the Marshall Plan, Point 4 and the Alliance by transferring his “War on Poverty” program abroad. A major premise of the Alliance was that stimulation of growth encourages political democracy. However, as noted by Packenham (1973) JFK officials tended to accept this relationship despite contrary evidence. It was soon clear that rhetoric did not easily translate into results. “Pluralism promoting” activities such as development of cooperatives had all occurred under the ICA before USAID. But recipient deviation from the US threshold definition of democracy would typically result in application of Cold War techniques, such as JFK’s cut-off of aid and suspension of the sugar quota with Peru after the 1962 coup. Specification of democracy-forcing policy measures was even less clear than the relationship between economic growth and democracy. Part of the problem is the inability to cite any examples of democracy constructed via foreign aid. For example, over the last 100 coups, deposing 70 African leaders in the last 25 years, “if there is a political system that has evolved as the norm, it is the one-party state, often based on one-man rule (Cowell, 1984). In Zimbabwe, the US has provided the most aid, $285 million in the last four years, and this resulted in the current on-party rule (Dickey, 1982). The problem of trying to find and support with aid the “tolerable spectrum” of force between the already polarized extremes remains for aid policy-makers. The final conceptual novelty of the Alliance period was “country programming” over the “project” approach. Since program versus project aid is still an important issue within USAID, a word of background is in order. Kaplan (1967:291) notes that the project approach to financing aid was firmly entrenched at the beginning of the Kennedy Administration and that strong resistance existed to changing this to the gimmickry of the new country programming approach. Nevertheless, JFK wanted an integrated aid effort comprised of technical, capital, and commodity assistance all “tailored to the particular circumstances of the individual country” (Packenham, 1973:62). The “project” approach was initially a reform intended to control “end use” abuses by recipients. Co-mingling US funds with other donors such as UNRRAor provision of “general

28

The Politics of United States Foreign Aid

program” support put recipient end use of funds out of US, particularly congressional, control. Montgomery (1967:32) noted that the “fungibility” problem of fund transfers to different uses than intended, created the strongest political complaints against any form of capital assistance. “Projects” were also a reaction to the disastrous international loans of the 1920s, made at high interest rates for ill-defined purposes, followed by defaults. In the post-war period, it was believed that more productive use of funds would be obtained if funds were tied to specific undertakings—railroads, powerplants, etc—and that this would also permit better appraisal and evaluation (Asher, 1961:67). The project rationale also extended to incremental policy-making derived largely from the US experience. Projects are smaller, hence changeable in mid-course. They establish decentralized growth points and their demonstration effects can be great. They also tie technical assistance to capital support for particular undertakings. Broader design and program issues are usually not central to their success. Growth is an erratic process, hence the need for a diverse system composed of small parts, not inter-dependent for success or failure. Plans tend to hinder foreign investment, encourage bureaucratic controls and government intervention in agriculture, industry and services (Asher, 1961:81). Further, LDC governments often use their powers to stifle “unanticipated growth”. Since so little is known about the growth, development and “take-off” process, the US should employ the project approach as the least damaging alternative. But the counter-arguments swayed the Kennedy Administration into country programming. The Marshall Plan worked because it had objectives (recovery) time limits (4 years) and a target price tag ($17 billion). US underwriting of sound LDC projects doesn’t prevent LDC governments from financing unsound projects in the wrong industries (1961:83). Planning and programming is therefore needed for basic resource projections (population, food, clothing, schools) to reveal inter-sectoral relationships and to anticipate the consequences of alternative decisions. In the US, foreign aid appropriations lack projections, it is argued, and this increases the unreality of that process. A recent General Accounting Office (GAO) report (1985:18) indicates that although the USAID Annual Budget Submission (ABS) process raises programming issues, “reviews do not routinely add, delete or modify project proposals”. The minuet of USAID project proposals for stability and progress, followed by congressional pruning which left the public “confused and disenchanted” 20 years ago, continues down to present day under Gramm-Rudman “sequestration” of budget authority across the board regardless of program results or country needs. The JFK approach tried to integrate project plans (especially with economic criteria) and budgeting into the larger scheme of the development program. Note that this approach was consistent with McNamara’s use of PPBS in the Defense Department at roughly the same time. But unlike defense issues that could be presented in “hard” quantitative terms that appear deceptively certain, foreign aid is conceptually hamstrung by the softness of its ends and means. Programming, required in theory, was difficult to do in practice. The entire issue was similar to the lengthy and probably irrelevant debate over loans versus grants. When does a project become a program in an area such as “community development”? Packenham (1973:62) noted that “in practice country programs to a large extent become packages wrapped around the things that would have been done under the project approach.” During this period, at each confrontation with painful reality, JFK and LBJ retreated to Cold War behavior, such as counter-insurgency programs and aid cutoffs. Liberal

The Political Economy of US Foreign Aid: Past and Present

29

democratic and populist at the outset with Latin America, by the end of the period the foreign aid program was mired in contradictions. For example, despite clear evidence of Latin American sensitivity to US interference in their affairs, JFK “tied” 95% of all new aid to US purchases to prevent the benefits of US aid from spilling over into Europe (Hudson, 1971:102). Similarly, the era of the most liberal and democratic foreign aid and foreign policy was also the era of the biggest catastrophe: Vietnam. The liberal response of JFK, and later LBJ, was to incrementally turn up both foreign aid valves—first increasing the number of military advisers directly (or indirectly by “Vietnamization” later), then increasing the funds for the “political” alternative to the military solution. This was the “strategic hamlets” program which isolated peasants into fortified villages and held elections by “secret ballot” (Packenham, 1973:83). The Vietnam policy was, of course, an example of extreme dissensus, like driving with the brake on and the accelerator to the floor. Since the military counter-insurgency effort was losing ground, the social or “pacification” program could not function either. As Popkin noted (1983), the ultimate success of pacification depended on the physical security of the villages. An unfortunate lesson of the Alliance period was that liberal foreign policy failures were also foreign aid failures, despite any measurable field successes. As Hudson (1971:98) noted, by 1966 the days of bilateral aid were numbered, and by the beginning of President Nixon’s reign, his goal was to divert the growing reaction against bilateral aid generated by the failures of past foreign aid, by militarizing US aid. The strategy for maximizing US “self-interest” via foreign aid in the 1970s was contained in the 1970 Peterson Report. The Report demonstrated concern for the three basic foreign aid issues of: (1) security, (2) costs and benefits, and (3) human need, and attempted to integrate past experience into its four major recommendations. First, it recommended a strategy to preempt congressional budget cuts in presidential requests by removing direct control of foreign aid from Congress. The report noted that in 1968 there was increased talk about “rat holes, dictatorships, corruption, balance of payments, threats of communism, brotherhood of man” (etc) and that “pent-up dismay over the Vietnam situation found an outlet in attacking foreign aid” (1971:101). The Report recommended use of a new development agency, the International Development Bank financed by bonds outside the appropriations process. Second, the Report recommended phase-out of military grants since LDCs were already significantly dependent upon US weapons systems via replacement demand. The premise was that since aid created trade, let LDCs take over their own military burdens—we provide the weapons at a price, they provide the manpower. Third, the Report recommended multilateral aid to involve European treasuries in the US loan program. The contradictory but long-standing preference for loans over grants again revealed itself in the notion that LDCs be kept on a “tight fiscal leash” to insure appropriate fiscal and monetary policies. Finally, the Report recommended staving off “social revolution” abroad by (1) free trade policies, (2) increasing local manpower for the armed forces, (3) introduction of modern capital technology, particularly in agriculture, and (4) mass population control to alleviate economic pressures on existing institutions (1971:104). Unfortunately, other than the recommendations for trade and population reduction, both of which would reduce the need for aid, the Report indicated little understanding of the issues confronting foreign aid or how to resolve them. Though attempting to export the 19th century British notions of government, foreign aid was

30

The Politics of United States Foreign Aid

simply part of the international power struggle. In a “realist” sense, US foreign aid in the Alliance period contradicted most of its liberal intentions. The Alliance began liberally and dramatically to develop democracy abroad but ended in the Vietnam quagmire with the vague sense that multilateralism might be politically safer after all. By 1973, however, pressures reintensified to get control of foreign aid policy again, this time for the purpose of helping the rural poor meet their basic human needs. The fourth period of US foreign aid “New Directions” extended from 1973–1980. The “New Directions” mandate contained in the Foreign Assistance Act of 1973 (Sections 102b, 107, 111, 113, and 281a) was fundamentally a reaction against the apparent failures of the “traditional approach”. That approach stressed provision of capital to increase local savings and incomes which would increase aggregate investment, leading to growth (impliedly capitalist and democratic), and take-off. Like foreign aid policy, the theory underlying it was smeared in the wake of foreign policy failures of the first three periods of foreign aid. Poor countries had not only refused to “take-off” and become democratic, they were still poor in many cases! Hence, despite little empirical evidence for or against the traditional theory, it was deemed politic to change the theory. It was argued that previous aid programs and projects (loans or grants) had benefitted mostly the middle and upper social sectors, and the poor had not advanced. Emphasis on capital projects/infrastructure and agricultural technology (the “green revolution”) had increased income inequality despite productivity increases. “New Directions” was a mandate to “target” aid to the poor by USAID and World Bank. Mickelwait et. al. (1979:2) suggests that emphasis on the distributional objective was even greater than in the US “War on Poverty” since that policy had merely added equity to the existing mandate for growth under the Employment Act of 1946. World Bank President Robert McNamara and Congress wanted an extreme change in emphasis: equity/distribution would “assume primary importance. “In most cases, the New Directions mandate challenged both the traditional notion of “trickle down” benefits derived from accepting recipient political and social structures, and “revolutionary” notions that class structures had to be eliminated before policies of any kind could benefit the poor, by suggesting that any growth with equity must be preceded by “asset redistribution” (such as land). Note the similarity with the social change thrust of the Alliance program. Many practitioners believed New Directions meant “progressive and incremental change” (Morss and Morss, 1982:29). “Incrementalists” used the examples of Morocco, Pakistan, the Philippines, and Mexico to demonstrate that rapid growth alone without relatively egalitarian distribution of assets would not lead to a reduction in poverty. Conversely, asset redistribution did lead to rapid reduction of rural poverty in: Taiwan, South Korea, China, Cuba, and Japan (Morss and Morss, 1982:29–30). However, the meaning of “New Directions”, like “social change” and “democracy” in the Alliance period, was unclear. This was particularly acute since the requirements of the mandate “conflict with other foreign policy objectives”, such as, “supporting friendly nations and maintaining continuity in the delivery and implementation of foreign aid” (Morss and Morss, 1982:28). Streeten (cited in Morss and Morss, 1982:28) suggested that “explosive interest in equality” could be another US “fashion cycle” (to be imposed on aid recipients) such as: (1) emphasis on import substitution to new primary export restrictions, (2) the switch from investment in physical capital to investment in formal education to

The Political Economy of US Foreign Aid: Past and Present

31

informal education and motivation, (3) from pessimism to euphoria about world food production, (4) from large scale to small scale production techniques, and (5) from growth to environmental concern. Morss and Morss (1982:66) also note that “The curtailment of population growth and the role of women in development are two further Western-inspired trend-setting concerns that are not totally compatible with the development philosophy of poorer nations.” Finally, the quest for control over foreign aid left over from the Alliance period, and the desire for a “big push” on behalf of the poor, led to an almost predictable organizational and budgetary contradiction that should have been evident by 1973. USAID decided to incorporate program budgeting (PPBS) in the 1970s as its vehicle for linkage of the programming and budgeting function. This led to the flurry of paperwork surrounding the new scheme of project development: Project Identification Documents (PIDs), Project Papers PPs), and the “logframe” (actually developed by Practical Concepts, Inc.), all of which continues today “with a vengeance” (Jickling, 1986). The misplaced zeal for PPBS was transferred to many LDCs as part of the US foreign aid contribution. Despite the fact that nobody really knows how to do program budgeting (Wildaysky, 1984:197) in the US, it can be found in LDC constitutions, such as Uruguay, as the law of the land! Similarly, the New Directions program produced fads like “logframes” (massive charts that sequence every conceivable activity over the life of the development project, apparently according to logic) that often ended up as “wallpaper” in local AID missions. Let us briefly examine the meaning (s) of “New Directions” in practice and suggest why it has been inconsistent with both empirical theory and LDC institutional reality. Most practitioners include the following approaches in their definition of “New Directions”: (1) Basic Needs, (2) Employment Generation, (3) Integrated Rural Development, (4) “Bottlenecks First” approach, and (5) Appropriate Technology. In contrast with traditional approaches, “Basic Needs” implies targeting aid first for physiological needs—air, food, water, then to follow a continuum, security, self-respect and self-actualization needs. The idea of “employment generation”, of course, is hardly new but with “increased productivity” as an additional bonus. The notion of “integrated rural development” has “become a guise for almost any effort directed at a rural area” (1982:34). Combining production, income-generating and social service activities “necessary” to improve the rural condition suggests the final obliteration of “program-project” differences since these are viewed as “processes” with large expenditures in the first few years. According to Morss and Morss (1982:34) “it is extremely difficult to specify in detail ex ante what is needed, how it should be timed, and what the implementation/organization approach should be. “The “bottlenecks first” approach assumes that development “bottlenecks” should be tackled serially. But this assumes that one can locate the right bottleneck and not mistake the means for the ends (1982:35). Morss and Morss note also that often consultants’ “institutional affiliation” can dictate selection of the obstacle, such as an airport construction recommendation to an airport-oriented construction firm (1982:35). Further, years ago Hirschman (1958:25–26) cast doubt that the removal of obstacles would cause development since they are themselves “reflections of contradictory drives and of the resulting confusion of the will”—such as distorted incentives in the policy-making process. How “obstacles” reappeared as a “new” approach in the New Directions mandate remains a mystery. Finally, the reaction against prior aid efforts to “diffuse Western technology” via tiedaid was most evident in “Appropriate Technology. “The idea was to avoid, for example,

32

The Politics of United States Foreign Aid

capital-intensive technology in high unemployment areas. Unfortunately, like most reactions, the tendency was to discard everything prior, good or bad. This meant that capital formation related to the poor, such as rural roads, training centers, and research stations were jettisoned as mere “infrastructure”. But frequently such Western social capital is required before New Directions can improve income and participation. For instance, IMF budget austerity requirements imposed on Latin America 4 years ago have produced drastic cuts in infrastructure expenditures which impair long-term recovery prospects. Capital expenditures for rural roads, transportation facilities, and maintenance have been slashed to balance budgets, but this is unlikely to increase debt repayment capacity in countries such as Mexico, because sustained development depends in large part on such investments (Witcher and Walsh, June 11, 1986). Similarly, “information systems” could be viewed as inappropriate Western technology when, in many cases, they are needed to link institutions and their intended beneficiaries for accountability gains. “Institutionbuilding” cannot really be accomplished without such prior investments (Morss and Morss, 1982:38). While it is hard to argue against “appropriate technology” any more than against “clean air”, many LDCs view this subprogram as a vehicle for withholding advanced (or what they consider “appropriate”) technology from them! (1982:65). The two major obstacles to New Directions success have been (1) lack of an applied policy theory basis., and (2) lack of practical relevance for real “needs”. First, despite the ongoing debate over development theory, in the last 25–30 years the Third World has progressed in health, education and welfare gains “beyond all expectation” (Morss and Morss, 1982:49). But there have been few empirical studies of Third World progress. Those that have been performed such as Morawetz (1977 World Bank, cited in Morss and Morss, 1982:51) suggest real growth in addition to growing income gaps between Third World rich and poor. The implication is that growth means expanding employment opportunities which improves income distribution. Therefore growth which expands employment reduces poverty by providing absolute income increases to the poor. Put another way, past growth efforts, without specific poverty policies, indirectly targeted employment and poverty. Therefore, “New Directions” has been largely an administrative fad. Little empirical evidence exists to support the New Directions repudiation of the traditional approach. As noted, since much of the New Directions consisted of recycled concepts from the Alliance and earlier, it was difficult to label it a “new” program. Second, in practice the Third World recipient probably wonders why it should pursue a New Directions strategy when the West didn’t develop from the “bottom up” by initially favoring the poor. In practice, LDCs view New Directions at best as the latest “expert flipflop” (sorry, we were wrong for the last 30 years) and at worst a new form of exploitation. For instance, environmental degradation is still viewed as a First World “luxury” issue by many, though forestry and energy are clearly more important issues now in poor countries given their effects on agricultural productivity. Morss and Morss (1982:60) imply that donors give recipients such programs as New Directions because “Foreign aid is rarely based on the needs of the recipients”. Instead of money for the poor, LDCs need (1) preferential trade agreements for their exports to donors, (2) untied program aid so that when food and fuel shortages occur, LDCs don’t have to buy Land Rovers or foreign technicians, and (3) aid for recurrent or operating expenditures such as salaries and maintenance. For, despite “New Directions”, the traditional theory persists that development occurs by capital

The Political Economy of US Foreign Aid: Past and Present

33

growth and that operating expenditures are not “bankable” (1982:60). Morss and Morss note, for example, the continual breakdown of water installations that inhibit achievement of water goals in Tanzania (1982:61). A better case could even be made for donor support of recurrent expenditures instead of capital items. Greater recipient support for the latter would indicate greater local commitment and perhaps fewer ill-designed and executed projects. (It should be noted that this policy has shifted in some US domestic areas such as urban mass transit. Frequently capital grant effectiveness was jeopardized by local lack of funds for operating and maintenance. Items that were “capital” such as engines and spare parts have been redefined as “associated capital maintenance” under Section 9 of the Surface Transportation Assistance Act of 1982). Hence, in theory the old methods of: sector loans, large infrastructural projects and macroplanning gave way to New Directions concepts of: participation, evolutionary process, grass roots, bottom up, local organizations, area development, integrated approach, decentralization, small projects, appropriate technology, sociocultural awareness, and small is beautiful—or “they do know how” (1982:92). In practice, Morss and Morss suggest that there have been” only a limited number of New Directions projects and only “some important changes in sectoral allocations since 1973” (1982:91). Unlike earlier policy shifts in the first three years of the Alliance, massive budgetary support did not follow the rhetoric of New Directions. Nevertheless, Montgomery suggests (1986:98–99) that by the early 1980s, nearly all of the US missions outside the Middle East had begun “concentrating on the reduction of poverty. And by 1986 there were very few programs linked to the politically attractive high-tech activities and massive interventions that had dominated the two previous decades.” With the arrival of Ronald Reagan in 1981, the new aid approach of the 5th phase was easily predictable. “Old-timers” in the aid business that had survived three or four Presidents were already trying on “supply side” rhetoric for foreign aid. USAID created the Office of Private Sector in 1981 with a view to harnessing profit incentives for development projects as noted above. Clearly, Morss and Morss are correct in calling this contemporary phase another fad like New Directions, and in blaming “excessive enthusiasm for superficial panaceas” (1982:103) on the fact that the top echelon of USAID is a “political dumping ground” for many administrators. However, it is also true that analysis of current LDC and donor incentive structures with a view to harnessing profitability motives is largely what Morss and Morss recommend (1982:104). For example, the failure to examine dysfunctional incentives in Latin American forestry bureaucracies has been a major impediment to the effectiveness of sectoral aid in the past (Guess, 1982). In this sense, the new approach deserves support. On the other hand, to the extent that our current “2-camp” view of the world is pushed on LDCs, the foreign aid program is also discredited further (Morss and Morss, 1982:105). That the Reagan Administration has given increased legitimacy to the “hard” or “security” school of development (law and order as a precondition for development) over the “soft” school (governmental legitimacy as a precondition for development) (Montgomery, 1986:67) is evident in the growth of security expenditures as a percentage of total foreign aid expenditures. Military aid rose from 39.9% of total foreign aid in FY 86 to an expected 41.5% in FY 87 (Nowels, 1986:12). Our brief review of US foreign aid indicates, among other features, that it cannot be examined in a historical policy vacuum. Despite rhetoric to one side, actual policy as

34

The Politics of United States Foreign Aid

expressed in expenditures and other actions tends to move in the opposite direction. The “humanitarian” Recovery era discovered the need for economic incentives for real welfare gains; the Marshall Plan discovered the need for security measures though not its original intention; the welfare-democracy orientation of the Alliance For Progress ended up in the confusion of Vietnam where neither goal was accomplished; the New Directions mandate produced few tangible changes but lots of new jargon; the current Private Sector Initiative is driven largely by security and private sector goals, yet some of the leading US-supported dictators have been eliminated and many LDC poor are reaping gains from worldwide economic recovery. Where one stands on foreign aid is often not where one sat initially. As noted, these aid policy shifts have been deliberate and often desperate responses to foreign policy crises. In each case, they have been billed as the ultimate panacea for our foreign policy problems and in each case they have been endlessly controversial but of questionable practical significance. In the next chapter, we examine the historically intense and growing criticism of US foreign aid more systematically as a prelude to the development of the Bureaucratic Role Conflict framework in Chapter 3.

Chapter Two CRITICAL PERSPECTIVES ON FOREIGN AID

Introduction Forty years of foreign aid experience have provided ample opportunity for both praise and blame. Aid recipients often dislike the PX-privileged Americans with their lavish spending habits and envy-provoking behavior in LDC cities. Even more suspect are American aid technicians in rural areas where their natural curiosity often turns up discrepancies and contradictions in government programs (Montgomery, 1962:178). As one might surmise, praise of the program is not extensive and usually limited to visible technical aspects, such as roads, dams and airports. Criticism, however, flows profusely from at least four ideological stances. First, many believe “it” simply cannot be done—“we don’t know how”, they don’t know how, nobody really does—hence foreign aid should be terminated. Second, the left criticizes US foreign aid as a tool of neo-imperialism bent on simple exploitation of the masses, while maintaining US predominance in the global capitalist system. Third, liberal determinists criticize US foreign aid for supporting governments and groups that have little interest in welfare, democracy or freedom. Finally, the right challenges aid as a waste of funds that alternately discourages honest labor while stirring up revolutionary expectations, all of which threatens the existing order. It is argued that while all contain grains of truth, they are flawed assertions that ignore the reality of foreign aid policymaking and execution. The innate simplicity of these critical positions is suggested by their common recommendation to terminate US aid. This chapter reviews the major lines of criticism and presents brief rebuttals to each. It then argues that criticism should be focused on the policy-making and implementation processes of US foreign aid, the reform of which could lead to improved foreign aid results. For this purpose, a budgetary policy perspective will be employed in the next chapter that will be applied to the three major regions of US foreign aid expenditure. The “Bureaucratic Role Conflict” model suggests that foreign aid failures have been largely due to the increasing presence of many “mixed” role institutional actors which compete over resources and policy direction. That is, in order for the Madisonian pluralist bargaining model to function properly, real advocates fighting for their service interests must clash with real guardians who are informed and seek to hold the line in the interests of profitability and welfare. In foreign aid, many interests have neither guardian nor spender roles, meaning they are not adequately represented. Foreign aid thus turns into a transfer of US bureaucratic problems abroad, with recipient benefits or losses as an incidental sideshow.

DOI: 10.4324/9780203840184-2

36

The Politics of United States Foreign Aid

Harsh Evaluations From The Political Spectrum As noted, foreign aid evaluations are often made at multiple levels—individual technician, project, country, and regional-international implications (Montgomery, 1967:73). Depending on who is doing the evaluation, ample latitude exists for praise or condemnation. Montgomery (1986:89), for example, notes the general tendency to smother misused military aid and aborted rescue missions in patriotism, while failures “or even marginal successes” in civilian ventures are often attributed to venality or at best naivete. Even assuming integrity, donors cannot compare dissimilar benefits, such, as reduction of crop pest damage versus road benefits. Since the comparison problem is “insoluble” and indirect proxy measures must ultimately be subjective, the measure often adopted is budgetary resources spent on the project (McNeil, 1981:12). Beyond the obvious methodological problems, the will to perform rigorous evaluations of aid projects tends to be an exceptional feature of the aid cycle. World Bank performance reports are unusual in their candor and criticism. The Bank is also unusual in its tendency to learn from past mistakes, which gives it its reputation among practitioners as the top development aid agency in the business. By contrast, USAID stresses how to do evaluations without provision of incentives to actually do them. The USAID “Design and Evaluation of AID-Assisted Projects” manual (1980), for example, spends a great deal of space on how to develop goals (reduce infant mortality by 1%), purposes (persuade couples to adopt family planning), outputs (establish 2000 family planning clinics within 3 years), and inputs (technical assistance resources), or the so-called “Logframe” method of designing and evaluating AID projects (1980:68). The AID technician as well as AID contractor next learns about Bar Charts, PERT Networks, and Network Time-Scaling in the grandest tradition of performance and productivity measurement used by US state and local governments. The manual cites Section 621A of the Foreign Assistance Act as requiring adoption of methods for comparison of actual results with anticipated results, and providing annual reports to Congress on the efficiency and effectiveness of the foreign aid program (1980:127). But one also learns that “regular” and “impact” evaluations are performed by USAID itself (1980:133). Only “special” evaluations are done by outside consultants. Predictably, in-house (often including consultants with close ongoing ties) “Project Evaluation Summaries” (TES) are general, with few fundamentally critical observations or recommendations for radical change of the project or program. For example, Montgomery notes (1986:92) that in attempting to develop a monitoring and evaluation system for the Mali Livestock II project in the Sahel that would inform planners whether their investments were being accepted or neglected by intended beneficiaries, AID “like most complex bureaucracies, was not concerned exclusively with the results of its operations… Washington did not want its field agents to change a project in mid-course after it has been approved.” Hence, much of the criticism comes from those who have worked on AID projects, and for some reason no longer need to feed at the trough of consultantdom. For, if criticism was based on offically-developed information, there would be little criticism of a systematic nature. It should be quickly noted that not all observers are critics. A National Defense University observer, for example, suggests that “Our 30 years of experience and success are unsurpassed among donors” (Hough, 1982:92). Ranis (cited in Hough, 1982:92) also notes the many innovative ideas generated by US bilateral aid, such as “flexibility on local

Critical Perspectives on Foreign Aid

37

cost financing”, and “initiation of rural poverty-oriented assistance projects. “Hough also indicates that “no other donor has been able to combine capital and technical assistance as effectively as we have in various assistance packages over the years” (1982:92). But he cites no concrete examples to substantiate the claim. The experience of many practitioners is that innovations usually flow from World Bank to USAID where they have often been copied or given lip-service. But evaluations of such projects as the Mali Livestock project, where drilling wells in grazing areas was the measurable output, found that the unintended results were actually widening “circles of death” around the wells (Albright, 1985). These results strongly suggest inflexibility in bilateral programming and execution, hardly innovation. The US can claim unparalleled success in one area—in absolute terms it has spent more money than any other donor: $5.7 billion in 1981, followed by France at $4.1, and Germany at $3. 1 billion (Lewis and Kallab, 1983:276). Additionally, the US provides more private voluntary assistance ($1.0 billion in 1981) than any other Development Assistance Committee (DAC) (created by OECD to coordinate member bilateral aid programs) country, and more than all the rest ($998 million) combined (Lewis and Kallab, 1983:277). But as indicated by McNeil (1981:12) “budget spent” is a compromise measure used because of the lack of appropriate criteria available to assess foreign aid performance. The first line of US foreign aid criticism stresses that nobody can really provide foreign aid properly and the best strategy is therefore to leave “them” alone. The view stems less from sociopolitical ideology than from an extensive projection of rationality onto recipients. It is thus difficult to classify on a left-right spectrum. Yet the view is powerfully held by practitioners and academics of diverse political views alike. In perhaps the best collection of anti-USAID anecdotes to date, Paddock and Paddock (1973) detail the ongoing bureaucratic tragi-comedy of the US foreign aid program. For example, they found two AID-funded agricultural research stations in Guatemala without communication lines between them. Nor was there any formal or informal communication between the Ministry of Agriculture and AID. Part of the problem, they found, is the constant change or elimination of personnel, meaning that discontinuous experts often initiate programs, followed by others who constantly “reinvent the wheel” (1973:10).. Elsewhere in Guatemala, AID tried to build a rice drying and storage plant as its “big contribution to the antiguerrilla campaign” (1973:19). The Paddocks indicate that the plants were easy to build; thousands of such facilities have been put up in the US. Additionally, AID should have had an easier time given its 10 years of technical assistance experience in the Guatemalan valley. Instead, consultants came and went, equipment arrived too late or too early for supervised installation, builders didn’t follow specifications, instructions were printed in English, etc. (1973:20). Despite its obvious unworkability, AID claimed the project was a success in that it deterred Communist guerrillas (1973:21). But the Paddocks rejected this claim as well, suggesting that the AID hierarchy was too inflexible to “pacify” the area of guerrillas, i.e. the project was a total disaster. Again, success seemed to lie in the eye of the beholder, not just for individual technician evaluations, but for project and country levels as well. Like the proverbial “pitiful helpless giant”, US “developers” bestow their gifts awkwardly but fervently. The Paddocks recount the excessive optimism of a group of “gringo” developers in a Guatemalan hotel, planning to bring American “know-how” to

38

The Politics of United States Foreign Aid

build silos, and grow strawberries and coconuts. They “know” that Guatemala and Oregon have the same conditions, and that all is needed is a “can do” attitude, helped by pulling a few strings among the big “politicos” (1973:62–65) : “Need to wake people up to the problem. If the silos work in one of the five places, they ought to work in the rest of the country.” “World’s going to starve to death in 1976, so we don’t have much time.” “How do they plant corn? Anyone here ever planted corn? (silence)” “Hell, the Department of Agriculture can tell us that. What we need to know is how to change the system here. It’s bound to be lousy..” (1973:62–65). Similar to the old saw of talking to cabdrivers and barbers for the real picture, the Paddocks find bush pilots useful because they have ferried around so many development experts on feasibility study assignments. One back-country sage, for example, informs them that “I’ve learned one thing. If a project succeeds, it is because it is right for the place, because the air and the land and the water are right” (1973:119). Impliedly, the US places its projects in the wrong places, pours money into them, and hopes to outweigh likely failure with positive publicity on other grounds (like deterring guerrillas). Land reform has also been a big area of not “knowing how”. The Paddocks suggest that many development experts feel confiscation of large holdings and their redistribution to the landless will bring the most rapid change to the rural community (1973:160). But the uncertainty of bank credits, differences in land quality, communal landowner skills (Mexican ejidos) and recipient governmental technological support, often mean failure in productivity and employment terms after redistribution of land. So the Paddocks suggest (1973:167) that land reform is “not a guarantee of either an increase in agricultural production or an improvement in the national economy. “Is USAID the only “fall guy” in the development business? Far from it. The Paddocks suggest that the Inter-American Development Bank is running a close race. “IADB has no more foresight when it comes to supporting successful programs than do present or past US foreign aid agencies.. For one thing, it has followed the stale policy of putting more funds into the combination of industrialization and electrical power than it has to the agricultural sector” (1973:264). The Paddocks therefore conclude that since we (the West) don’t know how to solve our own problems, such as subsistence farming and poverty in Appalachia, we should not be telling the Third World how to solve theirs (1973:303). The “know-nothing” critique is an important baseline from which to begin foreign aid analysis. The position needs to be taken and should be institutionalized into USAID, IADB and other US-fed MDBs as a healthy perspective for change and improvement. Anecdotes of dramatic failure in AID-financed projects are a constant. Without ability to relate a few (usually during the next mission assignment), one’s credentials are suspect among fellow practitioners. Everyone who has worked in the development business hears and relates them; most are largely true, though they tend to exaggerate with time and the conclusions one reaches from them are often questionable. Despite praiseworthy intentions, the big US donors can’t seem to get it together and one conclusion is that “they don’t know how” to encourage development. I will cite only four anecdotes that indicate problems which seem to impede US donor effectiveness.

Critical Perspectives on Foreign Aid

39

First, donor jargon often impedes public understanding of development work; it certainly constrains inter-donor communication. Paddock cites “deobligations”, “obligational year budget”, “reobligations”, “program loans”, and “sector loans” as examples of peculiar bureaucratic meanings. But the use of in-house language and concepts as instruments of power should not surprise anyone who has worked for large organizations and the use and understanding of federal budgetary and accounting terminology is standard for every agency and interested contractor. After working at USAID for about a year in Washington, however, I found that the language often took on a life of its own; even AID employees differed on their interpretations of terms and this made for lively staff meetings. During a joint USAID-IADB-World Bank project in the late 1970s, IADB staff were first surprised, then amused at AID language (“Nobody understood them”). The next thing they noted was that AID sent different people each month to staff meetings (“There was no continuity”). But development practitioners are supposed to be good at foreign languages. Hence this does not seem to be a major problem impeding US effectiveness. More serious is the problem of avoiding critical information. One can find this in two rather obvious areas (1) USAID cables to and from its Missions, and (2) Field missions for feasibility and evaluative studies. Much of AID work consists in sending and receiving cables to and from the Mission, such as Costa Rica, and Washington headquarters. Most contain lengthy and innocuous observations. Some are useful in recounting recent studies on particular subjects, such as forestry development. But any discussion of the implications of AID programs on real development problems—poverty, mass political movements, governmental legitimacy, or even the current political climate, earns the cable some level of “classified” status. Such observations are to be exceptional, for benefit of a specific organizational caste that has the apparent capability to deal with all the excitement. In many cases, the analytical elite turn out to be amateurs and political hacks, and this can severely damage foreign aid effectiveness. Donors like USAID also spend lots of resources on field “missions”. These are temporary teams of experts designated to gather information from the recipient country, often to determine whether a proposed project or program would be worth the cost. Missions can split up locally, fanning out to meet different ministries, groups, and leaders for information. More commonly, the mission is run rather stiffly by a leader seeking to advance his or her career through a solid performance. The leader might insist on large, impressive meetings with roomfuls of officials giving speeches, coffee and bocas to follow. Information might be exchanged each evening in the hotel lobby to assure that everyone is on the right track, i.e. no deviants saying harsh things about the local institutional structure. Under these conditions, few would persist in anything excessively critical, and the offical line is purchased often at a high price in terms of donor time and money. In an organization such as IADB, local (country member) representation on the mission is important. So, where the local returns to their own country on mission, it is a kind of “homecoming”, a chance to renew political ties. This, of course, can distort the information received by the mission by building in “sweetheart” relationships. Conversely, any dissent from the country member’s assessment of local doings is implied meddling in his or her country’s affairs by “outsiders” on the mission—a miniature attack and repulsion of imperialism, often in the hotel pub or lobby.

40

The Politics of United States Foreign Aid

In many cases, the donor and its permanent mission already know what they want, and attempt to use the visiting mission of experts as the vehicle for their position, i.e. the unbiased opinion of outside experts coincidentally recommends what we want. In 1980, I was sent to Ecuador as part of a two-man mission from AID/Washington to evaluate the feasibility of a forestry project (rewrite the “PID” in official terms). The other member, a forester, was charged with the technical work pertaining to small farmer forestry assessment, such as site analysis, species selection and projection of beneficiary income. After a month of interviews, reviews of relevant reports, and site visits, he concluded that forestry was not a viable option here because, among other factors, project area plots were too small (minifundia) and eucalyptus would be damaging to the already poor quality soils on steep slopes. I had concluded on efficiency grounds that technical assistance should not be given to the regular institutional machinery (the Ministry of Agriculture and several related agencies). Rather, I argued that smaller mixed-capital enterprises should be encouraged to expand their already successful work in small farmer forestry activities (Deely and Guess, 1980). The AID mission director wanted these conclusions changed to read: support of official institutions for small farmer forestry. We refused, left the next day for Washington, and several more “objective” consultants were brought in later to find the correct answers. This experience is not unique. McNeil (1981:60) describes the dilemma of consultants who found a Southeast Asian road project’s rate of return to be only 50% of that required for approval. But in that case they were rehired to assess the rate of return with additional “secondary benefits”! Agencies like USAID start small, with infinite missionary zeal and openness to innovative solutions. But they often deteriorate quickly into “bureaucratics” and the usual “hardening of the categories” that distinguishes the mature but top-heavy and inflexible organization. For example, despite a lengthy history of US intervention on behalf of US corporations to protect them from revolutions and nationalization, the Costa Rican AID mission was unable to develop a small farmer forestry outgrower scheme with Scott Paper Company. One AID staffer told me they were afraid of looking like supporters of the United Fruit Company in Guatemala (1954). Similarly, the AID mission would not support small farmer macadamia nut proposals in the 1970s. In both cases, private entrepreneurs, several of whom had bailed out of the Ministry of Agriculture, endowed themselves from plans intended originally for the poor! Part of the problem was change of leadership in the AID Rural Development Office in the middle of project development. The new director was simply fearful of making a bold move. One could also conclude that in many cases, AID simply doesn’t “know how” to move fast when opportunity presents itself. One AID/ Panama official told me that Cuba could supply technical assistance via teachers in one day, while “we would need a year of budget planning and paperwork to bring anybody in. The need is filled instantly by Castro and we lose the propaganda value. “Such is life in the tropics. Nor does AID really “know how” to use its own personnel to maximum benefit. While other nations such as Cuba gain propaganda from rapid-fire deployment of aid technicians, AID personnel tend to sit in Washington or recipient capital city offices, sending and receiving cables, in between the meetings and briefings that fill up their day. Coffin (1964:9–12) details a busy Mission Director’s day and believes that terms like “ugly American” have unjustly created a “cult of self-criticism” (1964:132) about

Critical Perspectives on Foreign Aid

41

US foreign aid. But much of the “work” is less related to recipient need career posturing and maneuvering for larger staffs. Those in AID who do not practice “goal displacement” tend to be isolated from the group. One frustrated AID technician, for example, became the first employee to be absent without official leave (AWOL). He spent months away from the AID mission in Manila distributing fast-growing leucaena seeds to Philippine small farmers, and now occupies a small, rather obscure office at Washington headquarters which features a leucaena tree growing next to his desk. The many that do go along with official policy often tend to contribute to failure by refusal to inform the hierarchy of the contradictions and complexities between field reality and the current program. Seitz (1980) implies that this kind of personnel wastage contributed to failure of US technical assistance policies in Iran. Fifteen years of US aid to Iranian ministries and the secret police (SAVAK) only contributed to the revolution and than instabilities of 1978–1979. He suggests that American advisers simply didn’t understand the complexities of Iranian political culture, i.e. they were naive (1980:409). Despite the surface plausibility of the “know-nothing” perspective, its weaknesses need to be brought into focus. First, as rational and wise as Third World peasants and small farmers may be, most are still poor and will remain so for their lives and that of their children. In countries such as Peru, the wealthiest 20% of the population have 32 times the income of the poorest 20%--the ratio is 4:1 for Finland, and 5:1 in the UK by comparison (Brown, 1984:202). Technical assistance and research efforts in the Third World, funded by donors such as USAID, have provided new crops and tree species that have raised Third World incomes and incomes of the poor in absolute terms. For example, USAID funds much of the work of the Consultative Group on International Agricultural Research (CGIAR). Through their work, the leucaena tree has staged a “one-species attack on hunger and soil erosion in the Philippines, India, Sumatra, and Africa” (Sinclair, 1985). Despite its comic bungling, the US foreign aid program has achieved successes for which little credit has been given. Much more could be accomplished. But it is far from true that “they don’t know how” in most cases. Second, often criticism of USAID turns out to be “sour grapes”. That is, “they” don’t know how because they wouldn’t listen to me. For instance, Carnoy (1972) modestly terms the frequent refusal of AID to listen to him on the housing investment guaranty program “the greatest failure of American foreign aid!” Giving funds to the Ecuadorian Forest Service in 1980 was, in my view, inappropriate. But organizational networks are notably unpredictable, and who is to say that additional legitimacy via AID funding might not have turned the Forest Service around? Similarly, the Paddocks (1973) claim that no one knows—not even they. But they repeatedly make recommendations that, if followed, would impliedly work. So, they recommend supportive credit and technology for land reform and better coordination for the rice plant project. Such suggestions are commonplace and simplisitic, and would likely make little practical difference unless their wisdom could dictate the right proportions and timing for each. Finally, AID is often blamed for inadequate planning and excessive planning, both of which are top-down. Much of this derives from institutional inflexibility common to most large organizations. But Third World governments also act hastily and are just as likely to make these errors without foreign aid. Aid can really be blamed only for the increased volume of funds (McNeil, 1981:71).

42

The Politics of United States Foreign Aid

The radical left also concludes that US foreign aid should end. But unlike the “knownothings”, they do claim to know how to develop the Third World. The Third World will not be developed by pursuit of “liberal” or “conservative” aid strategies. That is, just as the US foreign aid program represents a transfer of “liberal” political ideas (Packenham, 1973), such as political democracy, it also implies liberal economic ideas. The liberal economic idea is that an overall “international harmony of interests” exists that can be realized through such measures as liberalized foreign trade, foreign investment benefitting both lender and borrower, and diffusion of management technology. The liberal economic perspective presumes a “positive-sum game—not a North-South confrontation” (Meier, 1984:188). By contrast, the radical left views foreign aid as a strategy to promote the US “paramilitary self-interest” (Hudson, 1971:74). The program is simply one tool of US imperialism. US foreign aid and Third World development are then contradictory interests. By this perspective, foreign aid is a mechanism that perpetuates international inequality and widens the gap between rich and poor. Todaro (1977:55) notes that “disenchantment with the earlier ‘stages’ and ‘economic constraints approach’” generated increasing support among Third World intellectuals for the so-called “structuralist” model. This perspective “views Third World countries as being beset by institutional and structural economic rigidities and caught up in a ‘dependence’ and ‘dominance’ relationship to rich countries” (1977:55). Lack of development or modernization is not the result of “missing factors” such as capital or technology, but dependence of the economy on a larger and more powerful set of countries, actors, or policies. Elite activities combine to draw capital and expertise out of rural areas to produce a state of persistent “underdevelopment” or “boom and bust cycles”. Whether in the Third World or rural areas of the US, transfer of training and technical assistance to modernize rural areas only strengthens the system of dependent capitalism which can temporarily reform conditions but never restructure them (Guess, 1983:277–285). According to Wood (1980:2), “Foreign aid lessens the state’s dependence on its ability to appropriate domestic economic surplus by providing resources from outside the society. “Foreign aid facilitates” statist patterns of development” while increasing state dependence on private sources of accumulation, i.e. state capitalism. According to this position, foreign aid contributes to the exploitation of Third World masses by preventing Third World states from freely investing their own surpluses in poverty-oriented projects. Wood (1980:26) lists the channeling of aid through “development finance companies” (DFCs) as one way of excluding the state from “using foreign aid for state-controlled industrial projects because most DFCs are privately owned and only make loans to and investments in private enterprises. Similarly, USAID commodity financing (PL 480) passes through government hands but only on the condition that it be passed on to the private sector. According to Wood (1980:28) : “Apart from the possibilities for graft and power (not politically insignificant) to determine the allocation of import licenses and the like, the government is limited to an intermediary role and ends up mainly with its own currency. This local currency can be useful in keeping down inflation and in financing local currency costs of bloated bureaucracies, but it is useless for any kind of positive investment requiring foreign exchange.”

Critical Perspectives on Foreign Aid

43

Hence, aid cannot be translated into “state-run productive enterprises” and “aid introduces external resources for production without expanding the state role in production” (1980:28). Hudson makes the similar argument that by programs such as PL 480 which have indebted the Third World to the US and tied them to Treasury and State Department purse strings, foreign aid has “discouraged national planning for social and political independence” (1971:75). In short, the radical leftist argument is that foreign aid is ineffective because backward countries have failed to develop (Hudson, 1971:73). This means that use of neo-classical economic doctrines stressing growth rather than distribution of wealth is inappropriate. Needed are means of encouraging Third World “structural modernization” and “fundamental institutional breakthroughs” (1971:74). Here the left attributes success to US foreign aid that few critics would agree with. Hudson (1971:78) suggests that aid hasn’t “added to “Third World economies but rather to US balance of payments, industry, commerce and US long-term strategic goals. If this were true, foreign aid would be the most popular federal program and would undoubtedly have twice its current level of budget authority! The radical critique is that aid represents a “mortgage” upon future Third World balance of payments receipts and an “immense opportunity cost of not having acted earlier to guide their economies toward self-sustaining growth” (1971:78). Put another way, had Third World countries not accepted aid loans, their growth and self-sufficiency would have been greater (1971:79). At least four problems exist with the radical critique which seriously call into question its theoretical validity and policy relevance. First, they suggest that application of neoclassical economic theory to development problems results in a growth rather than distributive emphasis which actually intensifies poverty (Hudson, 1971:73). But this position is misleading. Some argue that no development models at all exist in US country aid programs because donors have faced too many surprises in trying to predict and control local politics (Montgomery, 1967:65). However, it is safe to assume that some ideology or developmental premise guides foreign aid decision-makers. For instance, many post-World War II development economists began to question the relevance of the market price system of neoclassical economics for the poor countries, as well as Keynesian analysis of income determination (Meier, 1984:133). Much of “development” theory applied in foreign aid by USAID and World Bank is not neoclassical at all. Rather, since the 1950s, the “unorthodox economics” of such writers as Prebisch, Myrdal and Singer have been the major guiding force of US foreign aid. The unorthodox theorists have produced such policy-relevant concepts as: dual economy, labor surplus, low-level equilibrium trap, unbalanced growth, vicious circles of poverty, big push industrialization, foreign exchange bottlenecks, unequal exchange, dependency, redistribution with growth and basic needs (La1, 1985:10). Many suggest that “the major thrust of much ‘development economics’ has been to justify massive government intervention through forms of direct control usually intended to supplant rather than improve the functioning of the price mechanism” (1985:10). In short, radicals criticize foreign aid for being market price-oriented while the guiding theory of foreign aid, in part, never implied this objective. Radicals seem to focus on the “old growth” economics premises that went out in the late 1940s, early 1950s. Naturally, many of the AID “lifers” (or old-boys) still employ traditional evolutionary growth premises of this period, and the current batch of Reagan appointees apply similar ideas. But

44

The Politics of United States Foreign Aid

from the mid-1950s on, most applied economists working in foreign aid used unorthodox theories for developmental problems. Hirschman (1979:84) also notes the beginning of the “reform” or “distributional” emphasis in the mid-1950s, backed largely by World Bank, after earlier concern with growth and industrialization in Latin America. In the 1950s, it was “planning” for economic development; by the 1960s, it was planning and industrialization to overcome dependence. Caiden and Wildaysky (1974:5) note that from such approaches as “take off” and “balanced growth”, “emerged directly what has been the most powerful concept in the development of poor countries for over two decades: rational economic planning. “The theories, in other words, were statist, unorthodox or “dirigiste” (Lal, 1985:10) to begin with. By these views, neo-classical theory was never really applied to Third World problems except for brief periods in the late 1940-early 1950s and the 1980s, when Third World government came to be viewed as a necessary evil in the process of accumulation, savings and investment. Second, radicals argue that foreign aid has inhibited real development policy-making by recipient governments. In fact, the US did begin technical assistance in the early 1940s to encourage Latin American production of strategic materials for the US in World War II (Montgomery, 1967:39) which is akin to “neo-colonialism”. Additionally, support for the proposition that aid “props-up” state capitalist governments can be found by noting the role of external capital in the recent Latin American debt crisis. According to Wiesner (1985:25) “Were it not for financing from these sources, which was growing rapidly and with evershorter maturities, an external debt crisis of the proportion experienced could never have developed.. What this financing did was to facilitate the postponement of measures that, in any event, would have to be taken to adjust the economies in question to the deterioration in their terms of trade..” On the other hand, it could be argued that for the first time in such magnitude, Latin American governments were free to make their own investments with private money not tied to official aid mechanisms. From 1975 onward, credits from international commercial banks increased much faster than financing by MDBs. By the end of 1981, liabilities to international banks accounted for 63% of the total external debt of the major borrowing countries (Wiesner, 1985:25). Also, 25% of these debts were short-term (1 year or less) and “an increasing proportion of the loans was no longer linked to the economic feasibility of investment projects. Most of the loans were being extended to the public sectors for the purpose of financing fiscal deficits and investment programs…” (1985:26). The banks assumed that “risk” was not an important factor and paid little attention to debtor economic policies. The result was predictable: “loan proceeds were not always well invested or used to generate foreign exchange or supplement domestic savings. At times, indeed, borrowing financed consumption rather than investment” (1985:26). Without being tied to foreign aid loan/grant criteria and with effectively untied program grants, nevertheless Latin American governments were unable to absorb funds and invest them wisely. External financing was therefore less important than absence of effective local public policies in creating the economic crises of the early 1980s. If the radical critique were correct, one could have expected either Latin American governments not to borrow short-term or to have invested wisely in their own development since use of funds was not tied to “neo-classical” project cost and return considerations.

Critical Perspectives on Foreign Aid

45

Third, throughout radical analyses of Third World problems one finds the presumption of state capacity to plan and execute development-oriented programs, over that of the private sector. If only the state was not dependent on “private accumulation” (Wood, 1980:5) and could “invest in direct industrial or agricultural production” (1980:4), the millennium would arrive. But such reasoning presumes an “advanced” state exists or can be developed in a backward society. According to Caiden and Wildaysky (1974:xvii) “Government is part of the environment of uncertainty and poverty. It also acts on that environment. But governments have not achieved mastery yet; that is part of what it means to be poor. To no one’s surprise, poor societies have poor governments. “To imagine that the commonly entrenched and bloated bureaucracies of the Third World are pent-up sources of human energy awaiting release is contrary to experience in most cases. Finally, the radical critique is penetrating and shrill but peters out just short of providing realisticalternativesforimprovement.Theradicalcritiquetendstoconcentrateonthemalicious, often imperialistic motives of the US in disbursing foreign aid and the “spurious” (Hudson, 1971:74) theories on which it is based. The impressive revolutionary recommendations of, for example, “structural modernization” or “fundamental institutional breakthroughs” (1971:73) usually boil down to either simple aid cut-offs or palpably trivial recommendations in the US rural development context such as organic farming or cutting off benefits to large farmers (punishing all farmers without weakening agribusiness intermediaries to whom much of the “structuralist” critique should be focused) (Guess, 1983:278). A recent example of the policy vacuity of many radical critiques is provided by former Venezuelan Communist Party member Teodoro Petkoff. Though he used to organize anti-multinational corporation efforts in the 1970s, he now vigorously supports them and urges their investment in Venezuela, because the left lacked any realistic policies to fill the capitalist vacuum. Specifically, the state invested massively in pine forests to reduce Venezuelan dependence on foreign pulp imports in 1964. But Venezuela lacked capital and technology to process the wood. The International Paper Company (US) offered to set up a pulp mill in exchange for access to the forests. The state balked on this and issues of foreign ownership. IPC left and Venezuela still imports $200 million/year in pulp! (Oppenheimer, 1985). Unlike the radical perspective which finds antagonistic US-Third World interests, the liberal critique is premised on a mutuality of interest. However, this perspective should be subdivided into (1) determinists and (2) reformers. The former group argues, akin to the radical perspective, that the US foreign aid program contradicts the principles of liberal democracy as exported to the Third World. Since security-dominated foreign policy ensures this result, foreign aid should be cut off for the benefit of the Third World. The latter group argues that foreign aid does all the nasty things liberal determinists say it does. But these tendencies can be reformed by thorough review of the determinants of foreign aid and an understanding of how current policy-making roles encourage behavior contrary to mutual US-Third World interests. Since the latter view approximates my own, it will be elaborated more thoroughly in the next chapter. Liberal determinists challenge the very concept of US foreign aid because it presumes to develop societies socioeconomically apart from existing power relationships. “US foreign assistance fails to help the poor because it is of necessity based on one fundamental fallacy: that aid can reach the powerless even though channeled through the powerful” (Lappe, et. al., 1980:10). The poor have been integrated into an exploitative system by foreign aid

46

The Politics of United States Foreign Aid

which ignores opposition to their attainment of power resources by multinational corporate lobbyists and Third World elites, who are the “real beneficiaries of foreign aid” (1980:13). Put another way, what the poor majority lacks is not technical know-how but economic and political power (1980:41). How else explain the amazing abilities of Peruvian and Colombian small farmers to efficiently grow, process and market cocaine without the benefit of US technical assistance? Among the most trenchant liberal complaints about foreign aid is that it does not flow to low-income countries. Only 21% of all DA, ESF and Food Aid ends up in poor countries such as Bangladesh. Most is concentrated, as noted in the last chapter, in countries of strategic importance like Israel, Turkey and Portugal. Liberals also complain that the top six recipients of US economic aid are also the top recipients of military assistance (1980:23). US aid supported dictators around the world such as: Somoza (Nicaragua), Zia Ul-Haq (Pakistan), Duvalier (Haiti), Marcos (Philippines) and Suharto (Indonesia) while they repressed civil liberties and prevented the poor from gaining access to social and political benefits. Why does US military aid support regimes that tend to negate economic aid benefits? They answer because only such regimes favor US corporate and anti-communist interests. Next, economic aid is often a catch-all category for projects that, in fact, do not directly benefit the rural poor. For example, “Food and Nutrition” project funds finance: rural roads, electrification, agricultural institutions, and satellite training and application. Rural electrification projects account for 40% of USAID’s food and nutrition category (1980:35). They suggest that $8. 6 million in “Food and Nutrition” funds financed Haitian highway maintenance (1980:39), indirectly propping up the Duvalier regime with funds ostensibly for the rural poor. Liberals also bemoan the fact that aid often flows to large farmers and agribusiness. This pattern of aid resource allocation indirectly takes land from poorer farmers and increases the number of landless sharecroppers or unemployed (1980:71). In other words, the major thrust of US foreign aid for rural development is to fund projects such as roads, and agribusiness that benefit larger farmers and later smaller ones on the “trickle down” theory. But since small farmers lack land and power, the benefits are absorbed by elites before heading downward. Or, as aptly put by McNeil (1981:69) “the recipient matter has proved unporous and the bottom layer has waited in vain for even a drop. “Emphasis is on increased productivity and a larger pie rather than on redistribution of wealth and income (Lappe, et. al., 1980:83). A frequent liberal claim, as noted, is that the Food for Peace program discourages local production. “Colombia’s wheat production dropped 69% while its imports increased 800%. By 1971, imports accounted for 90% of domestic consumption” (1980:96). In other words, US food imports are “additional”, confirming the criticism by Shepherd (1985), noted in the first chapter. The related charge is that Food for Peace revenues often finance unpopular dictatorial regimes. The bulk of food aid is “purchased” by foreign governments “that then may do with it what they please” (1980:93). The revenues in Bangladesh provide about 20% of the country’s operating budget, of which 27% goes to “defense, justice, and police” (1980:110). Liberals have also vehemently denounced the use of military assistance to “maintain hunger and underdevelopment” (1980:125). For example, the US provides military assistance to gross violators of international human rights such as: Zaire, Paraguay, Indonesia, Phillipines, Guatemala, Thailand, El Salvador, Bangladesh, and South Korea

Critical Perspectives on Foreign Aid

47

(1980:132). “MAP” grants of arms and equipment amount to four times that of DA project aid to the Third World. The “Excess Defense Articles” (EDA) program grants surplus arms and equipment (at “acquisition value” now, not replacement value anymore!). From 1950– 1979, $11 billion in EDA went to General Stroessner’s Paraguay (1980:133). The US also promotes military sales under “FMS” (to such allies as the Shah of Iran) to partly offset rising trade deficits (1980:33). Finally, many of the Private Voluntary Organizations (PVOs) that received impetus from the “liberal” New Directions legislation of 1973 “have become heavily dependent on US government monies, making them increasingly indistinguishable from agents of the US government” (1980:138). Therefore, most liberal determinists would halt US economic and military assistance to countries “where a genuine redistribution of control over productive resources is not under way” (1980:122). Liberal refomers also find US foreign aid to be misallocated to a poor majority that culturally cannot conform to rapid-fire project schedules of donors like USAID. They are skeptical of “trickle-down” theories, poorly planned and executed projects, and provision of funds to repressive dictatorships. For example, McNeil (1981:52) notes that the UN spends 41% of its regular budget in New York City, 30% in Geneva, 11% in Vienna, and 18% at 60 LDC “duty stations”. Based purely on expenditures, one could conclude that about 18% of UN funds “trickle down”. On the other hand, according to Kaplan (1967:297), “90% of USAID personnel are in Washington engaged in spending 10% of its appropriated funds. “This means that 10% are in the field distributing 90% of the funds. At least in the US case, the funds seem to be in the right place. In many ways, according to some liberal reformers, the present confusion about aid means and ends, tends to permit an interplay of overbroad interests in Congress and narrower interests in the executive branch, which permanently derails the foreign aid program. But Maynes (1983) suggests both liberals and conservatives are misguided on the issue. US conservatives are callous, resentful and superior toward the Third World, while liberals are “cloying in their mood of meekness and guilt toward developing countries. “But Tendler (1975:38) suggests that AID organizational confusion reigns in relative isolation from outside criticism which has now been “institutionalized” and that few have focused on the “non-macroeconomic” or “organizational” (1975:92) implications of these factors related to the kind and amounts of aid produced for Third World development. However, the liberal critique is not immune from weaknesses and contradictions of its own. We have noted that liberal notions of direct “distributional” programs seem to have less ultimate effect on improved incomes than policies which attack growth problems first (Morawetz cited in Morss, 1982:51). Hence, little empirical support exists for liberaloriented programs such as New Directions. This implies that traditional capital projects with trickle down premises may work just as well in the long run. Additionally, liberals challenge US foreign aid support of repressive dictators. But they also oppose intervention against dictators (Liska, 1960:147). The contradiction is illustrated nicely by the current liberal rift over $27 million in “non-lethal” aid to Nicaraguan “contras”. Some liberals oppose aid to them as immoral. That is, with the elimination of more democratic leaders such as Eden Pastora, the remaining “contras” consist largely of ex-Somoza loyalists of various shades that are unlikely to encourage an open and participatory political system in Nicaragua. But other liberals now support them as “freedom fighters” against the increasingly repressive Sandinista regime (McManus, 1986).

48

The Politics of United States Foreign Aid

Liberals, as noted, also have a way of pushing the latest US policy fad on the Third World, such as equal opportunity, environmental protection, and women’s rights, in the name of development. But just as easily as a small farmer loan for rural development, this could be (and has been) termed “intervention”. The core problem seems to be that liberals view Third World cultures to be pristine and rational (efficient) in their natural state, but not so natural that each liberal idea cannot improve the culture marginally. On the one hand, the US should not diminish the already rational culture by “diffusion” of technology and other alienating practices from the First World. On the other hand, “we” want “them” to meet our ideal of the “New Man-Woman”, meaning that such aspects of foreign aid as New Directions are intrinsically good, while military aid is evil because of its presumed anti-liberal bias. This perspective is quite naive in that it ignores the real possibility that whatever cultural practices exist could be detrimental in the long run. According to Fallows (1985:52) “Through the years, certain cultures have rewarded behavior that eventually proved ruinous to the society as a whole—the British upper class’s desire to be free of the taint of commerce is the most famous example. “Similar behavior of Third World political cultures may impede development by any definition (except development as survival). Wolfson (1979:47) notes that: “There is a strong circular causality between cultural values, the pattern of education, and economic development. Traditionalist society breeds its own patterns of educational values. Mandarin Chinese culture, classical Arabic education and West African bush schools have in common that they are not meant to be vehicles of social change.. Traditionalist intellectual pursuits in humanities, particularly literature and law, fail not only to fit the needs of a more technocratic society but also tend to perpetuate an undue emphasis on form over substance, on discussion instead of action, on endless litigation rather than decision, and on politics for the sake of politics.” In short, even a badly-conceived “interventionist” foreign aid program may be better than nothing, if for no other reason, that organizational skills imparted by the program’s bureaucraticized emphasis, may be useful later in stimulating political change. The liberal dilemma is that local cultural practices exercise a brake on political assertion and eventual development. The final group of foreign aid critics are termed “conservative” in the sense that they favor a narrow financially break-even program oriented toward the private sector. J.S.Mill once quipped that “Not all conservatives are stupid. But all stupid people are conservative. “In the debate on foreign aid, however, conservative opinion has rarely been stupid, and it is often more sober and consistent than liberal views on the subject. For example, many have erroneously attributed “public choice” theory to conservative thinking, since it was an outgrowth of neo-classical economic theory appied to US metropolitan area problems. But its major contribution is policy analytic realism, realized through the application of a “self-interest” assumption to rural people and bureaucrats instead of “some romantic or quasi-religious ‘communistic’ view of peasantry” (Russell and Nicholson, 1981:8). This kind of framework is conservative only in the sense that “it can provide a powerful basis for indentifying schemes that will not work.” (1981:8). We have come a long way from the debates over Vietnam and War on Poverty when liberals knew what was right (though not

Critical Perspectives on Foreign Aid

49

how to achieve it) and conservatives offered appalling Social Darwinist and “hard school” solutions to most social problems. Now it is the liberal turn to recycle murky concepts and worn-out solutions, or simply waffle (“justice” is the problem in El Salvador, noted Senator Ted Kennedy recently), while the conservatives are doing all the analysis and most of the original thinking. This is especially true in foreign aid. Conservatives generally believe in reduced governmental presence in society and emphasize efficient transactions between people and institutions that maximize each other’s self-interest. Early critics of foreign aid stressed the lack of businesslike efficiency which drained our gold reserves and added to balance of payments deficits (Morley, 1961:2). One might expect a renewal of this 1940s argument against foreign aid as the US is now a net debtor nation for the first time since World War I. Conservatives have always viewed foreign aid as a “something for nothing” do-gooder giveaway to mostly undeserving paupers. Some suggested making foreign aid more businesslike through loans instead of grants, tied aid and shipping materials in US “bottoms” (ships). But others noted that since 80% of the foreign aid contribution to balance of payments deficits was due to military grants, the economic component was more efficient and perhaps even contributed to favorable balances by stimulating purchases of US exports. Also, by 1962, 20%–33% of all outputs from US locomotive, fertilizer, and iron-steel industries were sold to the foreign aid program (Montgomery, 1967:88). But conservatives challenged the very notion of Third World development by foreign aid. The radical suggestion that foreign aid is inefficient because backward countries failed to develop (Hudson, 1971:73) is similar to the conservative view that “External donations have never been necessary for the development of any country, anywhere” (Bauer and Yamey, 1983:121). Conservatives believe that foreign aid is wasted on governments that are indifferent to development and that such funds diminish LDC productivity. They even assert that foreign aid “created” the Third World or the “fiction” of a “uniform collectivity with common interests” which has evolved into the “reality” of the “North South dialogue, confrontation or conflict” (1983:119). Consistent with their ideological views, conservatives (1) define foreign aid in narrow fiscal terms, (2) argue that recipient governments are indifferent to local development and Western alliances, and (3) claim that foreign aid contributes to US balance of payment deficits. First, they define foreign aid’s maximum contribution as “avoided costs of borrowing, i.e. interest and amortization” (Bauer and Yamey, 1983:121). That is, conservatives believe poor countries “that can use capital productively may borrow at home and abroad” (1983:121). But the poor countries that use capital productively do borrow from banks, at home and abroad. Wealthier LDCs owe most of their debts not to the US bilateral aid program but to commercial banks. The catch is that many such countries reached this enviable position of capital productivity through the effects of prior foreign aid. Poorer countries, as noted, must rely on softer loans for projects not acceptable to private sources because they involve such nonfinancial items as health, education and welfare. To say that Western countries relied on trade historically for raising capital is to ignore substantial differences in world cultures and economic systems. Cost avoidance is simply too narrow a definition for any aid program (US domestic or foreign) that includes security, social development and food aid as its goals.

50

The Politics of United States Foreign Aid

Second, bilateral US aid goes to governments whose policies tend to impede development. This point is raised by nearly all critics of US foreign aid. Bauer and Yamey (1983:125) note that such recipients of foreign aid strengthen their grip over the rest of society and adopt “anomalous” programs such as: expelling productive members, e.g. the Chinese from Vietnam and Indochina; traders from Zaire by Mobutu, and destroying food production by large-scale collectivization (Nyerere of Tanzania). Foreign aid also helps governments build lavish new capitals like Brasilia or establish new state airlines (such as PLUNA in Uruguay which at one point, according to V.S.Naipaul had over 1000 employees and only one functioning airplane!) which have little positive effect on either poverty or development. Moreover, Wolfson (1979:16) notes that the “consumer sovereignty” model (on which the US foreign aid program rests since it presumes evolution to a capitalist democracy) is contravened by the imposition of preferences in the Third World by both “selfpromoting bureaucrats” and a “kleptocratic autocracy” or “kleptocracy” (an autocracy that steals resources such as foreign aid for its own use). As noted, the Marshall Plan was successful precisely because the recipient governments wanted to develop. They exercised their skills in planning and executing their own recovery with only marginal external aid (Morley, 1961:23). Aid can contribute to the build-up of improper governments, and also to the “pauperization” and “de-skilling” of formerly productive people by handouts, such as in Micronesia (Bauer and Yamey, 1983:127). Conservatives often seek the easy way out by recommending aid via voluntary organizations (PVOs) “notably to nonpoliticized charities” (1983:134). There is much to the “inappropriate regime” thesis since governments are either receptive (will but absence of capacity) or not (capacity but absence of will), in which case they will resent US infringement on their right to adopt “anomalous” policies. However, in the real world of global power competition, bilateral aid packages are going to be designed to change local systems for donor benefit. Further, there are few nonpoliticized systems in the world, and the conservative critique often seems like a naive quest for neutral (but democratic!) systems where production is the primary value. But productive democracies are precisely the result of institutionalized politics favorable to those values. Improved political perspectives on foreign aid can therefore improve the result more realistically than the faith that elimination of politics can save the world. Finally, it is argued that US foreign aid is a drain on US balance of payments. “A shopkeeper does not prosper by giving away his cash to people, some of whom may later spend part of it in his shop” (Bauer and Yamey, 1983:129). This may be true of shopkeepers. But the global shop (or the international economy) requires stimulation of competitors for the benefit of consumers and producers and foreign aid is a necessary part of this process.. US foreign aid creates marginal economic losses—textiles and shoemakers for example, but contributes to larger US system benefits in trade and investment. It should also be noted that the bulk of countries that compete with the US are now “capitalist” (some even democratic) producers. This was not necessarily pre-ordained by the simplistic balance sheet perspective of fiscal conservatism after World War II. The overall US-aligned system has gained through marginal injections of foreign aid. According to CBO (1980:34) earlier work on the issue of whether bilateral aid assists US exports indicated that “more than 90% of US development loan monies returned in the form of additional exports. If US aid funds are spent instead in a third country, there will, depending on the economy of that country,

Critical Perspectives on Foreign Aid

51

still be some positive impact on US exports. Estimates range from 8%-30% of the initial loss of the untied aid funds. “In any case, based on the fact that 41.5% of all US foreign aid is military aid (Nowels, 1986:12) and that several FMS loan defaults have occurred, any negative foreign aid contributions to balance of payments deficits would likely be due to military aid (a curious favorite of many conservatives) which could and should be clarified and reduced. In sum, US foreign aid is an amalgam of programs harshly criticized or erroneously praised for major Third World developments or setbacks. This misplaced praise or blame is due in part to a failure to appreciate the limits of what “aid by itself can do” (Cassen 1986:12). US foreign aid is only a small percentage of the US federal budget distributed for a variety of often conflicting economic and military purposes. By itself, aid can neither reduce poverty, eliminate security problems, or create political development. It is a technical, fiscal, and psychological catalyst available to recipients predisposed to change and development. Nevertheless, US foreign aid occasionally increases instability, harms the poor (1986:11) and retards political development. Where these results are due to planning and execution failures, they can be remedied by improved decision-making incentive structures. Given its close association with foreign policy, particularly the security dimension, foreign aid is often dragged into major foreign policy crises, in some cases simultaneously as cause, effect and potential remedy! To some extent, foreign aid has institutionalized outside criticism and insulated itself from the swirl of domestic politics. But criticism still flows forcefully from conservatives, liberals, radicals, and know-nothings. Since much of this criticism is “from the hip” by people who influence the course of the program, it no doubt contributes in some immeasurable way to the many irrationalities and complexities that make foreign aid less effective than it might be. The criticism will continue. US foreign policy will also continue to blunder along in an increasingly unstable world that makes tough schoolboy language by US leaders and their Third World adversaries of the moment, acceptable if not necessary. The foreign aid program requires autonomy and distance from these policy level events and processes. In the short run, by any means, not much can be done about recipient country poverty and underdevelopment. But much can be done about improving the consistency, flexibility, and creativity of US foreign aid policy-making and execution. Much can be done to improve USAID activities in Washington and in its 60 world-wide missions. For this reason, analysis of the institutional factors which create and perpetuate foreign aid programs is in order. To this we now turn.

Chapter Three THE BUREAUCRATIC ROLE CONFLICT MODEL

Introduction To the extent that the criticism voiced in the last chapter is valid, the fault lies largely with the donor because it “controls the rules of the game” (McNeil, 1981:102). But few institutional systems reform themselves without sustained external pressure. In the case of foreign aid, as noted, pressure for improvement is hindered by the inherent political unpopularity of the issue among legislators and the lack of sustained interest group pressure. Under these conditions, foreign aid policy continues to be the result of a mixedmotive bargaining process which produces complex programs often removed from the needs of Third World recipients. The purpose of this chapter is to fit the complexity of this process into a simplifying framework that can point the way to reform. While most US public policies have been given this cleansing treatment on a regular basis by scholars and practitioners alike, and despite the vast extant literature on US foreign aid, few theorybuilding efforts have been attempted.

The Need For Policy Models It is hardly necessary to state that the “facts” do not simply present themselves for analysis. Models or paradigms are useful in providing frameworks that can tell us what to look for and analyze. By testing propositions related to this targeted phenomena, we can build theory and advance the notion of a policy science. In foreign aid, the relationship between fact, process and result is complex and some would even say, unsystematic. Nevertheless, certain models seem to fit what we do know better than others. Foreign aid programs tend to produce results that are: (1) unintended and full of surprises, (2) result in recipientdonor dependency or ward status, and (3) favor large and complex projects. Since these well-documented results suggest an originating mixture of politics and bureaucratics, the question arises whether a “bureaucratic politics” model can explain them? Allison (1971), for example, included the “Bureaucratic Politics” model as one of three conceptual lenses through which one could explain the causes and resolution of the Cuban missile crisis. His model focused on organizational “routines and repertoires” to describe what happened and suggest means for preventing inapproprirate responses to similar future crises. Impliedly, since politics and other societal variables filter through complex organizations on the way to becoming policy, it is most appropriate to examine the decision processes of relevant organizations. Others, such as Victor Thompson (1977:152) suggest that bureaucratic problems may really be “informal subsystems” out of control. For instance, he notes that where personal behavior patterns “exaggerate the characteristic qualities of bureaucratic organization,” DOI: 10.4324/9780203840184-3

The Bureaucratic Role Conflict Model 53 they are “pathological because they do not advance organizational goals” (1977:153). That is, bureaucratic problems may relate less to the existence of organizational routines and repertoires than to “self-serving behavior patterns”, such as “ritualistic attachment to the routines and procedures” themselves. Hence, the Allison “routines and repertoires” approach may be too simplistic in that it ignores the possibility that such procedures could be cause and/or effect of the organizational problems that lead to inadequate policy results. On the other hand, the Thompson emphasis on “bureaupathology” leaves us with the problem that organizations are normally rational but for the vicissitudes of individuals and informal subsystems. Thus control of organizations becomes an infinite regression problem of smaller unit coordination for multiple policy objectives. The framework doesn’t explain what caused the individual problems. One could easily discern that they may have been caused by the organization itself, bringing the framework in full circle! Put another way, many “bureaucratic politics” models are simply “black boxes”, inviting the writer to fill in his or her favorite variables. But theory-building requires development of a systematically related set of generalizations, the successful testing of which should explain the phenomena and confirm all or part of the theory. The above examples tend to be static frameworks composed of loosely connected generalizations. That is, must efficient organizations be composed of psychologically “normal” individuals? Does elimination of “bureaupathology” decrease or increase efficiency? Given the “impossibility theorem” of Arrow (1951) that where preferences are disparate, rational action may be impossible, one may doubt that rational inputs lead to similar kinds of organizational outputs. A model is needed which explains policy-making problems under realistic conditions of conflicts over resource allocation among multiple, specialized hierarchies. Personal and political agendas would enter this picture as byproducts of roles directed toward resource accumulation and allocation. With this simplifying assumption, greater explanatory power may be obtained. Many scholars have used resource allocation and role conflict models for diverse policy purposes. For example, Kanter (1979:5) indicates that the defense policy process is “encompassed by the budget cycle.” Participants recognize that “The budget cycle not only drives much of the policy process, but the formal budgeting system which defines that cycle also distributes bargaining resources unequally among the participants and produces difference in ability to influence outcomes” (1979:5). He notes how the rules for budget preparation affect both the “services”, i.e. Army-Navy, which view these rules as “obstacles and incentives which intervene between their objectives and fulfillment” (1979:6), and the administration, or the Presidency, which views them as a “fundamental mechanism by which to impose its view of national security on the direction of spending” (1979:6). Since military assistance is the dominant component of US foreign aid, and the conflict between services-administration over budget resources determines “substantive policy outcomes,” this perspective should be extremely useful in explaining foreign aid policy-making and results. In other words, it is more realistic to examine foreign aid as a product of budgetary policy conflict than as the simple but overgeneralized outcome of the American political process. As noted by Kaplan (1967:257), those interest groups with a stake in the process (farmers, banking, consultants) tend to have high access. But they are “passive” toward the parts of foreign aid on which they have no stake; no group pushes effectively and consistently for foreign aid as a policy, like agriculture or transportation. Similarly, McNeil (1981:51)

54

The Politics of United States Foreign Aid

notes that while individual investors and politicians try and influence particular foreign aid decisions in Britain, there is no effective pressure group operating to modify foreign aid patterns. In the US, the “loose coalition” of foreign aid groups (Morss and Morss, 1982:78) and a relatively “irresponsible” Congress (Kaplan, 1967:255) permit the executive branch to dominate the process. Hence, we emphasize bureaucratic policy-making and budgetary conflict as an explanation for foreign aid results. One weakness of resource conflict models is they tend to explain everything and therefore possibly nothing. If all outcomes are the product of budgetary conflict, we know where to look for policy determination but have little knowledge of the dynamics other than that everyone wants to maximize their personal or offical resources and will do so to the limit of their political support. Several writers therefore have added a teleological dimension based on degree of communication and trust among participants in the conflict over resources. Hirschman, for example, (1981:133) postulates that the development process in Latin America can be viewed as the “sequential unfolding of the entrepreneurial and reform functions..” Historically and presently, entrepreneurs have attempted to gain hegemony and depending upon their ideological support, timing and group identity, have tried to increase accumulation and profitability, or in our terms, “guard” national and personal resources. Conversely, the reform function begins after entrepreneurial activities as efforts “to improve the welfare and position of groups that have been neglected or squeezed” (1981:124). So, for example, in the wake of export-led economic development by local entrepreneurs in Uruguay (wool and meat), a determined effort at redistribution and reform took place. The tension and stages of conflict between these two functions explains the development process and implies that imbalanced growth leads to more substantive gains from both functions. Foreign aid programming often tends to ignore its potential contribution to the relative balance between these functions in the recipient country. Such dynamic approaches presume that the balance of power in functional or role conflict determines the likelihood of appropriate policy results. Severe power imbalances may persist, impeding communication and trust. But to the extent that the balance shifts, alternating between reform and entrepreneurship in the above example, trust between actors is likely to improve. This creates a climate in which improved mutual problem-solving is possible. To the extent that policy problems are amenable to solution by improved information and better decision-making, this should improve policy results. McNeil (1981:97) is correct in asserting that it is hard to say how trust emerges between donors and recipients (or among donors and among recipient actors), but its absence clearly impedes the effectiveness of foreign aid. Consistent with this assertion, Guess (1985) found that clarification of roles between donors (guardians) and recipients (spenders) in a US intergovernmental capital grant project for rail transit increased trust and improved policy results as indicated by reduced contract delays and cost overruns. In short, our “Bureaucratic Role Conflict” model is a “rational choice” model in which constrained actors trying to achieve different objectives assume roles attached to the positions they occupy in their respective institutions. For example, some try to guard the treasury against those advocating increased expenditures; others seek to expand the economy over the opinion of they who would conserve resources for financial stability (Caiden and Wildaysky, 1974:xii). Such a model is preferred over more familiar approaches that combine agencyenvironmental relations into a model for explaining bureaucracy (Warwick, 1975:9). For, to

The Bureaucratic Role Conflict Model 55 suggest that opportunities (appropriations) and constraints (internal routines) determine results in the State Department (1975:156), tends to encourage static analysis, i. e. a list of factors encouraging change or resistance to it. The role conflict perspective, by contrast, presumes that conflict over resource allocation produces different incentives which changes both perceptions of constraints and goals. To date, the major application of such an approach to US foreign aid has been by Tendler (1975). She describes the efforts of actors such as Treasury, Congress, Office of Management and Budget (OMB), and General Accounting Office (GAO) to control USAID policy-making activities, and the counter efforts of a decentralized agency with few domestic supporters and many distant beneficiaries. At the same time, the substance of the policy-making conflict remained vague because developmental goals are “not easily translatable into problem-solving tasks” (1975:25). As noted, this state of affairs tends to shift the balance of power away from Congress to the executive branch, diminishing the intensity of conflict over real issues. On the other hand, Tendler (1975:56–57) notes that the existence of such open-ended goals and procedures surrounding them, encourages AID to define output in terms of “resources transferred” or moving the money (1975:56, 88). Both donor and recipient contribute to the maintenance of this system, which biases resource allocation to large, foreign-exchange cost dominant projects. The important point is that both spenders and guardians know little about attainment of the real output (“development”) or how to control expenditures to attain that objective. Under such circumstances, conflict is more formal than substantive and the real beneficiaries tend to be donor and recipient development organizations. Some actors have guardian roles which are weak and ill-defined (or mixed), while other actors have spender roles which are clear, powerful, and often in conflict among themselves. Specifically, absence of an appropriate guardian for foreign aid, such as the authorizing (standing) committees, Appropriations sub-committees, or the State Department, delegates responsibility for policymaking and execution to mixed spender-guardian actors, such as the Defense Department, OMB, and USAID. Resource allocation and program definition is the result of an annual power play, not between shifting alliances of guardians and spenders but mostly between spenders themselves. This distorts foreign aid policy-making in favor of the most powerful spenders, such as DOD and the Presidency, precisely where the power also lies in foreign policy-making. The Bureaucratic Role Conflict (BRC) model presented here seeks to build on the work of Tendler, Hirschman and others who recognize the need for dynamic resource allocation analyses of public policies. The BRC model contains three independent variables, some combination of which should explain foreign aid results. As indicated in Figure 3.1, the major results of US foreign aid (Y) seem to be (1) generation of recipient dependency on donor military and economic assistance funds, with frequent degeneration to ward status, (2) projects of immense bureaucratic scale and complexity, and (3) consequences beyond donor intent and control. Much of this can be explained by conflict between ill-defined and weak guardian roles (OMB, Appropriations Committees, State, Treasury) and clear spenders with widely different power endowments (USAID, DOD, recipient institutions, US contractors and single interest groups) (X3). On the other hand, much of this conflict is structured by the existing rules and repertoires of the foreign aid budget process. Hence, the major institutional constraint to changing priorities and practices consists of the many Congressional (Foreign Assistance Act) and

56

The Politics of United States Foreign Aid

administrative rules (OMB, AID, State. Treasury) to which plans, budget requests and field activities must conform, such as foreign exchange cost financing only, violations of human rights, environmental protection, and utilization of US contractors (X2). Finally, resource allocation conflict occurs over differing institutional perceptions of foreign aid goals (X1). The predominant goal, historically and presently, has been defense-security. A secondary goal, often related to the first in practice, has been profitability or the technical requirement that benefits narrowly defined, meet or exceed costs. The least important budgetary goal has been development—health, education and welfare. In sum, the BRC model suggests that X1–X3 interact mutually to produce Y. Where perceived international crisis produces natural or human-initiated violence, such as earthquake or war, the security objective dominates the “goal” variable (X1) as well as “rule constraints” (X2) and “actor role conflict” (X3) as the key determinant of results (Y). Since perceived threats are almost normal, defense-security objectives have largely determined foreign aid results, except for limited periods as Postwar Reconstruction and a few years during the Alliance for Progress. Conversely, where developmental needs are predominant (absence of crisis and political support for welfare objectives), the results are still limited by rule constraints (X2) and role conflict (X3) dominated by powerful security-oriented spenders such as DOD. The BRC model will be applied to results of the US foreign aid program in Latin America, Asia and the Middle East in Chapters 4–6. Specifically, it will be used to forge preliminary explanations on apparent failures in Vietnam and Nicaragua, and successes in places such as Brazil, El Salvador, Taiwan, and more recently, in the Philippines.

Figure 3.1: The Bureaucratic Role Conflict Model

Results of Foreign Aid As noted in the last chapter, conservative and radical critics charge US foreign aid with preventing appropriate use of local resources by weakening host government incentives (Hudson 1971; Bauer and Yamey, 1983). Though the formal dependency-imperialist thesis

The Bureaucratic Role Conflict Model 57 overstates foreign aid costs, and events have discredited the “mercantilist” description (exports are good but import-competing industry is not—so the “parent” directly subsidizes exports to the “home” country), foreign aid does tend to generate host dependency on complex projects with often unintended consequences for both donor and recipient. The problem is how to measure these results. For example, the budgetary “obligations” used by USAID in its “Green Book” (1984) may be flawed in that they do not necessarily translate into total appropriations or budget authority for the fiscal year. Beyond this, total appropriations (including supplemental appropriations during the fiscal year) simply measure money moved and provide no insights into results obtained. Nevertheless, it will be argued that foreign aid results can be largely indicated by the three output measures proffered here. First, US foreign aid has encouraged recipient reliance on US imports of goods and services (spare parts, commodities) and the US has gained access to recipient resources (geopolitical security via bases, natural resources such as oil and strategic minerals). The basis for all US foreign aid is a long series of military pacts (Asher, 1961:65). The use of military and economic aid to secure and maintain base rights has been a form of “compensatory” aid (Montgomery, 1967:16) which often confuses program purposes. Montgomery also notes that this is “the most dangerous form of strategic foreign aid” (1967:23) in that it increases at least the “appearance of dependence” (1967:22). Military aid may, of course, decrease LDC dependence on the US for defense. However, in that such aid strengthens military groups contending for power in the recipient country, the likely result is a long-term increase (dependence) in military aid as the military perspective on social reform gains a foothold. For example, Montgomery noted the resistance of the US military to pushing the Vietnamese government on reforms—they viewed defections from the Diem regime as a “military problem” (1967:55). Economic aid also creates dependence on more US economic aid. Food aid, as noted, often decreases local incentives to produce where it is not an “add-on” to the local market. Though the “mercantilist” quality of the food aid relationship has given way to interdependence as LDC agricultural exports increase, generally “tied-aid” provisions have increased LDC dependence and weakened their balance of payments positions (Asher, 1961:41; Hudson, 1971:122), requiring more aid to support trade and budget deficits. Overly complex projects are often a cause and consequence of aid dependency. Directly or indirectly, most of USAID activities relate to “sector” programs which are simply a new way of grouping individual projects “for presentation purposes” (Tendler, 1975:6). Though some local-cost financing occurs in particular functional areas such as Agriculture, Rural Development and Nutrition, most project financing is limited to import costs. According to Tendler, 80% of total AID funds continue to be spent in the US for this reason (1975:73). The emphasis on foreign-exchange financing encourages large-scale projects since they require “less staff time per dollar transferred than a smaller one” (1975:56). Capital aid and technical assistance to specific regimes becomes “administrative intensive”, reinforcing the centralization pf power that already exists in most Third World societies (Montgomery, 1980:423). This, in turn, encourages oligopolistic local groups to gain power, results in capital over recurrent expenses, with all this implies for maintenance and operating social programs, and keeps “unsavory” regimes in power. The rationale for the foreign exchange financing rule is that aid is temporary, and that capital, not consumption, will lead to societal self-reliance (McNeil, 1981:73). But the rule distorts foreign aid and leads to overly

58

The Politics of United States Foreign Aid

complex, large-scale projects. In Ecuador, for example, though studies indicated that more efficient management could increase the volume of goods handled without building more warehouses and quays, the solution was not adopted largely because capital and foreign exchange costs of the project were insufficient (McNeil, 1981:74). Finally, foreign aid projects often produce unintended consequences. A rural road project opens up new areas for settlement and encourages severe deforestation in the process. This permits soil erosion which silts up a nearby dam, rendering it inoperable. Or suppose that without deforestation the dam works as intended but prevents enough siltage from passing through, depleting downstream soil of nutrients and requiring expensive fertilizer (McNeil, 1981:70). The former happened in Colombia (Eckholm, 1975), the latter in Egypt at the Aswan Dam. More recently, the Cahora Bassa hydroelectric project in Mozambique was supposed to produce up to 2075 megawatts for a cost of $2 billion. Instead, it produces 10 megawatts, loses $60 million annually, and is sabotaged repeatedly by rebels seeking to overthrow Mozambique’s socialist government (Frankel, 1986). These are dramatic but hardly unique examples of the unexpected “Rube Goldberg” results of some foreign aid projects. The purpose of reciting such examples is not to blame or ridicule but to shift attention to the project planning stage, where the course of US foreign aid is determined by powerful institutional roles. As will be noted, in addition to the incentive for having project plans approved, the policy process encourages inattention to implementation. This permits additional unintended consequences via recipient fund shifts. Occasionally, recipients could reprogram funds to increase project efficiency but are hindered by AID rules. Conversely, AID rules may be ignored or circumvented as funds are shifted from economic to defense uses or from economic development to the president’s re-election campaign. Kaplan (1967:136) notes that Food For Peace funds financed UAR and Indonesian military adventures in the 1960s— against Yeman-Congo and Malaysia respectively. He notes that “one of the least understood aspects of the aid program is the receiving country’s ability to substitute aid resources for other goods and services” (1967:135). Despite donor doubts about recipient “absorptive capacity”, the US is often prevented from enforcement of accountability provisions by charges of neo-imperialism. This happened in Vietnam; it occurred more recently in El Salvador. Many suggest that the US has spent over $l billion on the pretext of aiding land reform and economic development, to build a “mirage”. The unanswered questions remain whether such a small country can absorb vast amounts of money in such a short time and, more importantly, where the money is actually going? Nelson suggested (1983) that the “substitutability” or “fungibility” issue here is not simply a matter of efficiency but corruption. To find out where the money went, he recommended sending in teams of accountants to the field to check the books. Since several accountants subsequently “disappeared” in El Salvador on such missions, Nelson seems to have touched a sensitive nerve! Both kinds of unintended consequences continue to occur; both are related to how USAID plans and executes its projects in the larger US policy-making context. Despite their connotation as dysfunctional results, dependency and complexity may be useful to development under certain circumstances. Unintended consequences have long been the major result of human activity, individual or organized, and may also serve as a useful means of flexibility. USAID now includes “success stories” as part of its annual Congressional Presentation. For FY 87, AID notes many successes that could

The Bureaucratic Role Conflict Model 59 also be interpreted as increasing dependency, complexity and unintended consequences. The implication is that they be required for developmental success. For example, the $4.0 million in US commodity imports to Mauritius resulted in an investible surplus of $3.5 million for rural industrial buildings as the importer deposited the local-currency equivalent to the dollar purchase in a local US bank. (USAID, 1986:615). An AID loan for $5 million to General Motors Egypt resulted in 1400 local jobs at managerial, technical and blue-collar levels (USAID, 1986:621). Finally, an AID-funded complex waste heat gas turbine generation plant in Pakistan (Gudu) will increase capacity by 25% and reduce the number of power shortages due to priority irrigation requirements six months of the year (1986:624). Similar successes are documented for road, education and health projects by region.

Decision Premises or Goals It will be argued that policy-makers can be divided into spending and guardian roles over the crucial issue of resource allocation. However, the motives for preferring advocacy or guardianship will vary and with Downs (1967:103), we argue that motive is often strongly related to time spent on one’s specialty. Specialists are responsible for spending money but not raising it. Hence specialists are often in spending roles. Conversely, politicians (legislators) are “just as responsible for raising money as for spending it” (1967:104). This is the “guardian” role, the exercise of which transmits pressure to economize downward to each bureau. However, politicians are increasingly responsible for spending more than they raise, giving spenders the edge in policy-making. Recent Gramm-Rudman-Hollings legislation was designed to counter this tendency by building in an “automatic guardian” to restore balance to the budgetary process. In particular, guardian-spender conflict determines the relative importance of each goal. Schultz (1984:1) suggests that the foreign aid program serves four US interests: (1) growth of the world economy, (2) security of US vital interests abroad, (3) building democracy, and (4) alleviation of suffering by the world’s poor. Given the relative strength of security advocates in the foreign aid policy process (aided by the fact that the National Security Advisor has daily access to the President but the Secretary of State is limited in this capacity by foreign travel and the duty to oversee the State Department bureaucracy), the goals can be compressed into three in order of importance: (1) security, (2) profitability, and (3) developmental need or welfare. These goals reflect decision premises which, in turn, suggest the relative power of their holders in the policy process. The question then is how important is each goal to foreign aid? An answer presumes: (1) a clear path between foreign aid activities and goal attainment, and (2) an aid agency pursuing activities which are known to be related to developmental results. Neither of these conditions exist, allowing evaluation to occur on multiple levels. Under conditions of immense organizational slack then, USAID responds rationally by developing its own incentive structures that may or may not lead to “development”. These incentives are designed to attain in-house goals, adjusting for needs and objectives of institutions with which AID must share power, such as DOD and State. The importance of each foreign aid goal to AID will depend on the strength of they who specialize in that particular area. In that policy-making strength is a function of access to power (budgetary resources and

60

The Politics of United States Foreign Aid

Presidential support), it can be argued that the defense-security goal is predominant, developmental need least important, while profitability serves as the benchmark or standard unless overridden by crises involving the other two goals. Given the very real prospects for failure along some scale of performance and the risks of being held accountable, the foreign aid decision-maker can rationally avoid risks by: (1) rigid rule-adherence, and (2) stabilizing the recipient’s society. Foreign aid rules are a product of both USAID, DOD, and State Department actions, as well as Congressionallyimposed concerns attached to annual Foreign Assistance Act reauthorizations. AID strictly adheres to the rules because their net effect is often vague and contradictory. Strict interpretations serve as defense mechanisms for harrassed policy-makers who rightfully suspect they otherwise lack support for their actions. Stabilization of the recipient’s society is obviously a bigger order, but can be disposed of in the short run by dispensations of security assistance. Jacoby (1966:8) suggests that “foreign economic assistance should be regarded as an instrument of US national security.. “In this sense, many would agree with Hough (1982:x) that the US needs a “more assertive and long-term security-oriented US foreign policy in the Third World..” or a “national security rationale for US economic assistance” (1982:127). The US provides aid to “minimize the causes of world tension” (Montgomery, 1967:23) and stable democracies and economies are viewed as means to this end. But policy-makers are familiar with unintended consequences and know that, for example, economic and educational aid could lead eventually to unrest and security problems for which military aid might be required. So, foreign aid for sociopolitical stability can ironically cause a security problem. But security assistance to directly confront unrest often creates a morass of problems for US foreign policy. First, foreign aid strengthens the military hand of the recipient government and assures non-democratic rule. This polarizes society and the US loses any credit that might be obtained from military aid. For instance, US support of Chilean strong man Augusto Pinochet after the 1973 coup provided him with needed legitimacy to gain control of an increasingly anarchic political situation. But after 13 years of repression, the Communist Party is stronger than in 1970 (an estimated 20% of the popular vote as opposed to only 16% before) and they are considered the second best organized party in Chile after the Christian Democrats. Today, the military and Communists feed off each other in mutual dependence. “To justify repression, Pinochet always mentions the outburst of terrorism. To justify confrontation, the Communist Party always cite increased repression as proof that the military government intends to perpetuate itself (Graham, 1986). Naturally the military is the most efficient institution in most LDCs and military aid can produce useful infrastructure projects. It can also release local budgetary funds for economic and welfare spending. But most military governments spend the aid on themselves and frequently dispense its benefits on local populations in the name of communist (i. e. opponent) threats. Moreover, Morss and Morss, 1982:63) notes the ability of countries in search of foreign aid to extract “bribe money” to prevent them from going to war among themselves, viz. Egypt and Israel. Nevertheless, the security goal will remain predominant for the foreseeable future, and for this reason US aid will not be replaced by trade and investment as the principal means of LDC influence. As noted in Table 1.1, 66.8% of the

The Bureaucratic Role Conflict Model 61 FY 86 foreign aid budget request was for “International Security Assistance”, suggesting the ongoing predominance of this goal. The second “decision premise” or goal of US foreign aid is that financial benefits shall exceed costs. Such a premise is virtually ubiquitous in formal and informal aid criteria. For example, Wilford (1979:267) complains that the AID “New Directions” strategy replaced the “basic criteria of economic efficiency” with “some nebulous equity criterion for an equally nebulous target group.” His point is that “Foreign aid is about the business of economic development” and that 30 years of development literature have taught us that resource allocation decisions “must be made on the basis of objective economic criteria, whether it be cost-benefit analysis, cost-output analysis, social marginal product, or internal rate of return” (1979:267). The profitability premise applies even to soft loans for projects that do not meet normal break-even criteria. The presumption is that health projects will profitably benefit recipients, agriculture projects will produce goods for market, highway projects will stimulate commercial income growth, land reform projects will stimulate productivity and employment benefits from the security of title, etc. But the unforeseeability of many intervening variables for long-term projects (wars, droughts, coups) renders formal profitability analysis suspect. Costs and benefits can be defined narrowly or broadly, and the final balance or ratio often depends on who is doing the calculation, such as consultants and donors ready to “move the money”. According to McNeil (1981:15) “There is a view that the answer to many of the problems of foreign aid lies simply in increasing sophistication: more and better planning, project preparation and appraisal. Factors other than technical and economic are understood as important-but the solution in all cases is seen as more experts and, more studies. “The technical evaluation myth that only “beneficial” or “profitable” projects shall be funded went hand in hand with the “traditional development approach of increasing aggregate income through largescale infrastructure investments and the importation of Western technologies” (Morss and Morss, 1982:25). While this goal did not apply to military aid (grants and concessional loans finance weapons and military assistance through FMSC and MAP programs because hard loans would otherwise “impose a severe economic burden” Office of Management and Budget, FY 86 Budget, 5–17), “profitability” worked itself into USAID routines and repertoires. A high proportion of US foreign aid funds are for capital projects, mainly in the Middle East. Given the “security interest” of such projects, and the prohibition against local cost financing in most cases, the profitability approach is taken seriously. Congressional and AID practices require that benefits exceed costs and, on paper, this is possible to demonstrate for most proposed foreign aid expenditures. But DA projects are usually not capital intensive and some local costs can be financed. The more creative thinking in foreign aid has been in the area of design and execution of: small-scale energy (hydro, windmills), agro-forestry and social service delivery projects. As noted in Chapter 1, the New Directions mandate of 1973 was a reaction to the traditional capital growth-security approach to development which pervades foreign aid. But for DA projects, AID policy-makers have always been aware of the limits of this theory and had consciously followed alternative models offered by FAO, UNDP, World Bank, and other bilateral donors such as Germany, Canada, and Switzerland. Project development and appraisal follows the AID “logframe” (USAID, 1980:59–106) which is employed to

62

The Politics of United States Foreign Aid

predict workable relationships in advance, such as the ecological and economic effects of highways. Though the process is institutionalized, it is applied as a supplement to decision-making and not a technical, “numbers-crunching” replacement for the decision process. Project proposals are subject to multi-disciplinary critiques over a lengthy period of analysis. Nevertheless, as the foreign aid budget request is aggregated and moved upward in the policy process, profitability considerations tend to gain hold. For this reason, the developmental needs goal itself is the weakest contender of the three, despite its highest direct value to US long-term interests in the Third World. In AID, like any large organization, policy-makers have mixed-motives despite their specializations and the composition of agency motives tends to change over time with larger political shifts. Current thinking tends to be dominated by security considerations as the primary means of assuring developmental need, i. e. the “hard” school means to the “soft” school result. Confirmation is provided by the fact that for FY 86, the request for “Functional Development Assistance” within USAID will be only about 13.1% of all budget authority requested for foreign aid ($2.1 billion out of $16.0 requested) (OMB, 1985: Appendix).

Constraints: Rules and Repertoires Whatever the mix of forces that determines foreign aid goals, their ultimate appearance as policy depends upon institutional actors, i. e. the bureaucracy. As in many policy areas, the administrative apparatus tends to take on a life of its own, determining policy acccording to its own incentive structures and customary practices. Given the porousness of foreign aid ends and means, subgroups within the policy process can employ rules and procedures to further their own ends, such as increased power, status, and budgetary resources. This phenomena is called “goal displacement” (Merton, 1952) or “suboptimization” and leads directly to excess formalism or rigid adherence to the rules over their substantive purposes. Subgroup (bureaus and departments within AID, State, DOD, etc.) exercise of parochial values means that the part becomes more important than the whole. The culture of bureaucracy is extremely important to foreign aid since, as noted, this is actually five programs with separate and overlapping constituencies. Foreign aid is a domestic program subject to domestic rules and repertoires whose ultimate results are abroad. Tasks necessary to accomplish goals (development, security, welfare) are not clear, diminishing the incentive for useful feedback. Like most policy areas, some rules are more important than others to final results. Which are important here? It can be argued that foreign aid results are a product of rules which affect how resources are: programmed, allocated, implemented and evaluated. This is a more bureaucratic formulation of the standard systems-feedback loop often used for policy analysis. In particular, it suggests that goal achievement is a function of corrective actions taken after external information is used to develop programs-plans, formulate budgets, monitor execution, and evaluate results (Anthony and Young, 1984:9–12). Let us place the relevant foreign aid routines into this framework to observe the degree of constraint on goal achievement. Planning and programming each aid program is performed separately, for example by USAID and DOD, and later integrated by State into a unified request for the Office of Management and Budget (OMB) that is presented to Congress. The process for USAID

The Bureaucratic Role Conflict Model 63 is described schematically in Figure 3.2. The relevant actors for programming are: State Department (which sets long-range annual resource levels, and planning levels with and for USAID) and USAID missions (69 countries in FY 86), USAID regional bureaus (Latin America and Caribbean, Asia and Near East, Africa) and USAID/Washington (Centrally Funded programs). USAID programming consists of three interrelated steps: (1) preparing and approving Country Development Strategy Statements (CDSSs), (2) formulating and reviewing Annual Budget Submissions (ABSs), and (3) developing, reviewing, and approving Project Identification Documents (PIDs) and Project Papers (PPs). The CDSS is a planning document introduced in 1978 by AID Circular A–384. It provides: the “conceptual framework” to AID missions for planning; the “basic reference document” by AID/Washington for “country program reviews” and is the “major reference point” for AID/W project reviewers; and it “sets forth what AID expects to achieve in a country and how it intends to do it” (GAO, 1985:3). Prepared by AID missions and approved by AID regional bureaus, CDSSs cover up to five years. A “basic premise” of A–384, according to GAO (1985:3) is that “AID must have an acceptable and current strategy of assistance for each country before budgets are submitted or projects approved.” The ABS justifies annual funding needs to carry out CDSS objectives. ABS levels are established by State-AID consultation based on perceived security-developmental need mixes. The missions submit their ABSs based on prior guidance to AID/W by the end of May. Regional bureaus consolidate this data into an overall AID ABS which, in turn, is “integrated” into State Department’s overall foreign aid budget. The State Department’s “Science and Technology Undersecretary” integrates and coordinates the foreign assistance requests of DOD, Treasury, USDA and USAID. According to USAID (Interview, January, 1986) emphasis is now placed on “integration of security and development” objectives. The integrated budget is then submitted to OMB in September for review and approval. Finally, the foreign assistance budget is incorporated into the President’s Budget and AID’s CP is submitted to Congress in January or February. After initial hearings and markups, the foreign assistance budget request receives authorization by the key standing committees (House Foreign Affairs, Senate Foreign Relations). Congress then attempts to develop the “nonbinding” first budget resolution, “crosswalking” (302b exercise) each program to the relevant agency from the appropriate committee (s). The Appropriations Committees use the results of the first resolution process as ceilings. After the appropriations process produces “budget authority”, Congress “reconciles” appropriations with changed revenue and expenditure estimates in the second resolution. In recent years, the Foreign Aid bill has been hidden within the omnibus “continuing resolution” that has resulted from inability to agree on appropriations. This has helped foreign aid by shielding it, as noted, from public wrath. But because of changes in schedule for appropriations measures made by Gramm-Rudman-Hollings legislation for FY 87, the House may take up foreign assistance appropriations as a separate bill for the first time since 1981. “Many believe that this procedural change will expose the foreign aid budget to further reduction amendments” (Nowels, 1986:ix).

Figure 3.2: USAID’s Planning and Budget Process, Fiscal Year 1986 Source: US General Accounting Office, “AID Recognizes Need To Improve The Foreign Economic Assistance Planning and Programming Process” (Washington, D. C., 1985), GAO/NSIAD–85–110, p. 2.

The Bureaucratic Role Conflict Model 65 The third programming phase involves development and review of projects. A great deal of AID work consists of developing and reviewing PIDs and PPs at all three levels of the organization (Washington, Regional Bureaus and Missions). Since proposals can be submitted and approved at any time during the fiscal year, project-related meetings seem always to be on one’s calendar. Missions initiate PIDs, but they are often rewritten by consultants and by AID/W with other “experts”. PPs are usually reviewd and approved by AID/W, though increasingly most missions can now approve PPs below $20 million (which decentralizes AID even further). Finally, the “project authorization” (PIO/T) is signed and the mission then negotiates a “project agreement” with the recipient government and US consultants that actually execute AID projects, such as Development Alternatives, Inc. or Westinghouse Corporation. These routines raise several issues that, together with the problems they create for AID, tend to operate as a constraint on the achievement of foreign aid goals. First, little is known about the precise relationships between programming foreign aid resources and their global effects on economic development, change, or stabilization. Foreign aid remains more of an “art” since the effects may be mutually interactive, i. e. change orientation interfering with economic productivity (Montgomery, 1967:55). Though the lack of an integrated foreign aid theory on which to base programming is important, clearly the way to build that theory is via actual field experience with programmed funds. But the high risks (personal career, and agency loss of funding) associated with unknown actions encourages large public organizations such as AID to interpret its rules narrowly. And if AID doesn’t, its overseers do—OMB, Congress, and State. To reduce uncertainty and avoid taking large risks, organizations emphasize routine, well-defined and measurable tasks. For example, at the encouragement of AID/W, missions spend enormous amounts of resources in annual rewrites and updates of CDSSs. The Liberia mission spent 5 staff years and nearly total resources for over 6 months prior to submission to AID/W. It was rejected anyway and had to be revised two more times (GAO, 1985:10). Despite the substantial resources devoted to CDSS revision, missions are still unclear about what should go in CDSSs! Guidance from AID/W is “all encompassing” and the CDSS is becoming a “catch-all for congressional and AID concerns” (1985:14). Since missions generally believe they should be general and flexible and AID/W reviewers want specific objectives and measurable benchmarks for future accountability, lack of consensus over the scope and purpose of the CDSS exercise persists (1985:13–14). A basic premise of AID programming (A–384) is that CDSSs must be approved before ABSs are submitted (to ensure that budgets and project proposals are consistent with country strategies). Since this rarely happens, either project proposals are delayed, such as in the Dominican Republic in 1983, or budgets are reviewed without approved CDSSs. The programming process encourages a numbers game which means emphasis on quantifiable objectives. These objectives are often “means” such as schools constructed, seeds distributed, and sites visited, rather than more specific changes in client welfare or behavior. Since debate about the latter would (and does) raise serious questions about many ongoing AID programs (meaning endless revision of ABSs, CDSSs, PPs, PIDs, etc.), nearly every institutional incentive within AID encourages business as usual. That is, commonsense deviations from past plans are viewed as threatening even during the programming phase when new information is supposed to be at premium value! Others

66

The Politics of United States Foreign Aid

have noted that most AID incentives are geared to preparation and planning and not project execution. “Like most federal agencies, AID offers few incentives for good project and program implementation. Instead, rewards come primarily from getting funds committed and projects approved and into the implementation stage. The best staff in Washington are concerned with issuing directives on how projects must look before being approved, in the field the staff focuses on obligation of funds” (Morss and Morss, 1982:85). Consequently, approved CDSSs and ABSs are not easy to change during execution. Montgomery notes the “relentless search for certainty” by AID/W which inhibits goal achievement by preventing field agents from changing a project in mid-course after it had gone through the byzantine approval process. The problem is that the project must be compatible with AID’S programming strategy “which called for a detailed blueprint for each project so it could be approved in Washington” (1986:92), Ironically, this institutional quest for certainty isolates AID/W from field problems, leading to the very unintended consequences which all hope to avoid. Further, AID programming routines encourage aid packages that may not be in the LDC recipient’s best interest. Project financing is mostly limited to foreign exchange costs (import costs), as noted, which encourages both donor and recipient planners to find visible, capital projects of use to both. This results often in excessive complexity and cost which are nonetheless safe and certain for both parties (donor and recipient). Such mutuality of purpose becomes a liability for larger US and recipient interests when security issues develop. In such cases, AID is locked in to the recipient government in a “ward” relationship, and this is reinforced by frequent DOD and DOS pressures to supply arms and security-related “services”. This financing rule tends to discourage smaller projects and recurrent cost items (wages, operations and maintenance) that can actually produce a higher benefit/investment ratio than the larger capital projects. Obviously, maintenance deficiencies can destroy a capital project, and schools without sufficient numbers of trained teachers can visibly retard development. The bias against recurrent cost financing also results in the tendency for AID to “ventriloquize” project ideas “top down” (McNeil, 1981:68) to keep the money in the pipeline. This is not a product of AID indifference to the poor but from institutional incentive structures that favor the same result. Finally, AID programming routines may discourage creativity in foreign aid policymaking. As noted, AID is but one actor in foreign aid programming. DOD also programs funds for MAP and FMSC activities which, in turn, affect the efficacy of DA and ESF funding. Kaplan noted (1967:285) that “administrative separation” within the US government of military and economic aid “imposes serious limitations on the effectiveness of aid programming. “By this argument, military aid can import social and technical skills and advance stable democratic evolution. Separation of military and economic aid programming therefore reduces the usefulness of military aid for non-military purposes. He also suggests that given its propensity to become excessive in both foreign aid plans and Third World government “wants” (not needs), “military programs should be subject to the same review and competition for funds as any other governmental activity” (1967:286). But, DOD suggests that actually AID fuels recipient demands for DA by developing its own projects, while DOD merely responds to recipient demands for security assistance (Interview, June, 1986, DSAA).

The Bureaucratic Role Conflict Model 67 Realistically, defense contractor and interest group pressure also generates internal demand to sell weapons rather than respond indifferently to recipient pressures. It was noted that the State Undersecretary for Science and Technology now coordinates defenseeconomic programming, so that the process is no longer totally separate. Further, Section 620 (s) of the Foreign Assistance Act requires that AID take into account the amount of foreign exchange or other resources a country has spent on military equipment (GAO, 1983:81). But most of this occurs after AID missions have submitted ABSs compatible with CDSSs, Hence the “arbitrary inflexibility” remains and tends to inhibit “maximization” of foreign aid “priority purposes” (Kaplan, 1967:285). The second constraint on goal achievement is budgeting or resource allocation. Budgeting refers to revenue-raising and expenditure for achievement of security, profitability and welfare objectives. How AID and its related foreign aid colleagues acquire funds and allocate them to programs and projects is extremely important to an understanding of foreign aid policy. In foreign aid, as in most public policy issues, the “bottom line” is the budget since it both reflects and directs decisions and power relationships. Although the relation between revenue and cost of production are not direct or linear in public organizations, in foreign aid the connection is even more distant. USAID is a general fund agency within the State Department. The latter effectively establishes resource levels and regional goals for USAID. Hence, both revenue and expenditure purposes are largely determined before the operating phase even begins. While Congress provides budget authority in the form of annual appropriations or continuing resolutions, little substantive communication takes place between Congress and the Executive branch on this issue. As noted, the domestic political climate encourages Congress to build up its budget from the the domestic side first, with foreign aid as an afterthought. Conversely, the President works down to domestic issues from foreign policy which includes foreign aid. Congress is largely isolated from the executive policy process and intervenes sporadically to change the larger directions of foreign aid, usually to the chagrin of AID line people. But, as will be noted below, at least in an aggregate sense, this relationship seems to be improving. One result of historic institutional distrust is that AID budgeting becomes a routinized, mechanistic process that tends to ignore field level needs or contingencies. Similar constraints work on budgeting that worked on programming. Though the budget process should be the final analytic filter before actual disbursements, several factors work against any kind of systematic assessment of project proposals. First, the foreign aid program must be reauthorized annually. The rationale for this lies largely in congressional interest in control (Fisher, 1983). But some components of the foreign aid program receive multi-year budget authority, such as FMS credits for arms sales. Budget authority for DA is mostly annual, meaning that enormous amounts of institutional resources are wasted in rationalizing DA before a Congress interested less in project management than global abstractions. Resources spent by AID to annually update CDSSs do not facilitate budgeting plans into action. Perhaps ABS time would be more justifiable if it resulted in analysis that produced rational shifts of resources for recipient needs. But GAO (1985:18) notes that the ABS does not play a critical role in “the programming process” and over time has become more “financially oriented”. Consistent with budget guidelines for FY 86 issued in Apri1, 1984, the typical AID mission submits budget requests in “largely tabular form” with “minimum narrative”. GAO also notes that

68

The Politics of United States Foreign Aid

“although programming issues are raised, reviews do not routinely add, delete, or modify project proposals” (1985:18). No “systematic assessment” of project proposals takes place and missions do not always receive routine feedback from the reviews. Imagine the kind of detailed analysis occurring in a process in which AID Africa Bureau could review ABSs from Cape Verde, Chad, Gambia, Mauritania, and Upper Volta in one two-hour session! (GAO, 1985:22). If the ABS does not focus on substantive issues of developmental performance in costs and benefits, on what does it focus? According to GAO, the annual country budget submissions are primarily concerned with “ensuring that established funding levels are met” (1985:17). Even the “issue papers” are primarily “budget oriented”, in a narrow “controloriented”, line-item sense. These papers identify such topics for discussions as: “loan/ grant ratios, operating expenses, staffing levels, and appropriate mixes of appropriation accounts” (GAO, 1985:18). Beyond this emphasis on formality over substance, missions are not always notified of ABS review results, i. e. they don’t know the results of their efforts. For example, as of May, 1984, the Kenyan, Liberian, and Rwandan missions did not know the results of the FY 85 ABS review conducted during the summer of 1983 (GAO, 1985:19). Results are not final since they may be changed by subsequent DOS and OMB decisions (r985:19). For instance, missions are instructed by AID/W “not to submit projects that exceed or fail to meet predetermined assistance levels” (1985:21). Given the fact that these funding levels are often “sacred” because of State Department” pressure” (1985:21), the AID budget process tends to become irrelevant in that the case is decided before the facts are in (as in any “fixed-ceiling” budget system). But the State Department often does not have field level knowledge of development needs or how to obtain them. As one State official told me in March, 1986, “we deal in words, AID people get their hands dirty in the field”. While this may be an exaggeration, it does suggest that State tends to be an isolated “guardian” of the foreign aid purse. At the other end of the scale, GAO (1985:22) notes that AID “mission representatives do not routinely attend ABS reviews to defend, clarify, or expand budget proposals!” In short, the AID budget process is less a learning experience in a development laboratory than a fund control exercise for bureaucratic consistency. In addition to the dysfunctional incentives of the AID ABS on domestic resource allocation, the budget process also works as a pressuring device on recipient countries. The drive for programmatic certitude by AID also suits the budgetary needs of recipients. For example, Montgomery (1986:64) notes that AID macroeconomic programming procedures allowed Americans the illusion that they were running Vietnam in the 1960s even though they weren’t. While administrative chaos and dissension within the recipient government may impede project implementation, the recipient rationally seeks to get AID money into its budget. Members of local line ministries (agriculture, transportation, etc.) know that while AID may complain of fund shortages, it often has trouble finding enough projects for the year. Line ministries are therefore more eager for fundable projects than staff ministries (planning, treasury, etc.) which are likely to be more discriminating (McNeil, 1981:22). Hence, a project for which foreign aid funds are assured “has an extremely good chance of being included in the budget against possibly better projects which are locally funded” (McNeil, 1981:20). The scale, scope and inflexibility of the AID budget process, together

The Bureaucratic Role Conflict Model 69 with the contradictory constraints placed on its actions by Congress, contributes to this perverse but entirely rational result. Intimately related to budgeting and dysfunctional foreign aid results is the budget format used for preparation and approval of annual funds. As noted, the format is largely line-item and tabular with minimum narrative. Funding is for personnel, schools, or clinics rather than results to be achieved. What this means is that AID emphasis on project evaluation is largely pre-budgetary. While AID indicates that the narrow financial emphasis of its ABS process is due to congressional budget earmarks, such as that 12% of its development and disaster assistance funding be given to PVOs, the more plausible explanation is that this format serves AID purposes in creating an aura of certainty amidst volatile congressional and Third World environments. Once funds are budgeted, the execution phase of foreign aid begins during which obligations (commitment of funds through agreements, grants, contracts, and purchase orders) must be tracked for consistency with appropriation purposes and to ensure that outlays do not exceed budget authority. The third set of routines that constrain foreign aid results are operational and accounting rules. Much of this area related to enforcement of legal aid obligations between the US and recipient via the “project agreement”. This document sets out goals, objectives, financing requirements (dollars and local currency) and obligates US dollars for the first stage of the project. It also authorizes the AID mission to release the first installment of “counterpart” funds to the host government. But the “agreement” only binds the US in practice. Absence of real sanctions (withdrawl of funds after a negative post-audit does not work because country dollar amounts have already been fixed by the AID programming-budgeting process and by congressional approval) means the host isn’t really bound to the bilateral agreement! Put another way, “efficiency” is often a trivial concept in Third World contexts, and financial-technical standards of performance are often incomparable (Montgomery, 1962:106–7). The bureaucratic control orientation noted during AID programming and budgeting carried over into the execution phase. It is here that the most blatant contradictions between plan and reality occur because (1) accounting procedures stress narrow tracking of budgetary obligations, and (2) recipient “fungibility” or inter-program fund transfers are largely out of US control. First, auditors and accountants in OMB, GAO, and AID stress the legalities of obligations (petty cash mentalities are evident here) rather than their relationship to funding a multi-year project in an uncertain environment. Narrow emphases on financial issues and cash balances tend to exclude physical performance considerations. Hence, valuable feedback to the programming phase is eliminated by fiscal year tracking procedures. Montgomery (1967:76) noted how a technically successful program establishing special organizations known as “servicios” was terminated because financial auditors suspected them of loose accounting practices. This suggests a great deal of distrust between AID line activities and financial overseers. In one sense the cash flow and control emphasis is necessary. For FY 86, new AID budget authority was $5.8 billion. But total program funding includes: unobligated balances carried forward ($1.5 billion), transfers between accounts ($10.3 million), recoveries of prior year funds (deobligations and reobligations) and reimbursements (USAID, 1986:4). Failure to control fund movements carefully could seriously impair foreign aid purposes.

70

The Politics of United States Foreign Aid

Further, failure to account carefully could open the foreign aid program to reverse exploitation by the recipient. Practically any form of aid is fungible or convertible to other than intended uses, e. g. commodity grants release local resources for expenditures on luxuries or weapons; soft agricultural loans are reloaned under hard terms to middle class beneficiaries; economic assistance ends up in military budgets (especially if the government regime is military, as in most cases). Montgomery notes that “any form of capital aid may conceivably contribute to uses of which the US disapproves” (1967:33). Often, it may be unclear which donor funded which project where a recipient receives multiple donations. McNeil (1981:66) notes that “a recipient ministry may be receiving assistance from 10 different donor agencies, each with accounting procedures that are not only alien and more demanding than those used by the ministry for its own funds, but also different from one another.” But the institutional control emphasis tends to replace substantive goals. The “paradox of procedures” is often that any organizational rules can ensure freedom and prevent official whim—but they can also hamstring results. As Tendler indicates (1975:41) “the technique of watchdogging a development program is, in a certain sense, just as underdeveloped as the knowledge about how to help countries develop. “For example, distrust runs high between OMB and USAID over attempts by the latter to recover prior year funds. AID does this by “deobligating” funds (withdrawal of obligated but unexpended funds where a determination has been made that the project is “nonperforming”) and “reobligating” them (new obligation and effectively a saving of funds) to higher priority projects. OMB requires that AID receive clearance for any reobligation despite the fact that any other federal agency can do so up to $500, 000 without clearance. This kind of accounting hamstrings foreign aid results and is directly the product of distrust and failure to communicate between key bureaucratic actors. The AID offical responsible for this activity indicated (Interview, June, 1986) that he has to show up at OMB unannounced to receive clearance because his phone calls are rarely returned! Finally, it follows from the above that evaluation and feedback routines also constrain achievement of foreign aid goals. In addition to the underdeveloped state of task-results linkages, if one assumes lack of integrity on the part of the donor, recipient (or both), the possibilities for destroying any constructive feedback are almost infinite. From what has been said, incentives in the foreign aid programming, budgeting, and execution phases are to design safe, large capital projects that will help AID move the money to receptive line ministries which gain correspondingly in power, and prestige. Under such conditions, evaluations are unlikely to be critical of existing operations, since all parties could loose in the short run if practices were changed (except the intended beneficiaries). It should come as no surprise that AID Project Evaluation Summaries are quite general and conciliatory in tone. By contrast, World Bank Performance Evaluation Reports are extremely critical and written for the constructive purpose of providing feedback for improved future projects. This also “controls” cash flow more efficiently than narrow financial audits by saving expenditures and increasing outputs. The general refusal of AID to take the lead in criticizing its own work (programmatic self-flagellation before Congress) opens it to all kinds of useless, crank-type criticism from all shades of the political spectrum, as noted in Chapter 2. Tendler (1975:43) notes that, deprived of domestic constituent support, the agency is also deprived of an important source

The Bureaucratic Role Conflict Model 71 of criticism. She notes that “the number of possible mistakes in a program like AID’s will be much greater than in the State Department simply because the latter does not have to build projects and rely on outsiders for essential inputs” (1975:42). To make matters worse, where AID tries to conduct thorough project investigations it may be hampered by other foreign aid actors. GAO notes (1982:3) that even though AID operates in a “high risk” environment of corruption and waste abroad, the ratio of Inspector General staff to program (1. 5) is much smaller than the average for the rest of the government (4. 9). Further, AID efforts to locate Inspector General staff near recipient projects has been consistently resisted by the State Department (1982:9), apparently the result of an ongoing bureaucratic turf war. So we return to the initial theme of Chapter 1, that foreign aid often takes the rap for larger US foreign/defense policy failures. We have seen that, given the immense uncertainty of the task environment, AID has designed procedures to shield it from unpredictable winds. But this incentive structure has its own costs and tends to produce projects which are: overlycomplex, security-oriented, consistent with recipient ward status, and likely to produce spectacular unintended results. The larger effect of AID rules is to “clog the pipeline” and tax the “absorptive capacity” of the recipient. Standard operations of the foreign aid process then, tends to overwhelm recipient institutional machinery and resources, which diminishes the likelihood of planning and executing successful foreign aid projects (Tendler, 1975:86). The tragedy is that, despite its marginal contribution to Third World development efforts, the US could do much more without spending more money. AID procedures tend to shield these possibilities from view.

Opportunities: Role Conflict We have discussed some of the rules and repertoires which limit US attainment of foreign aid objectives. Let us now turn to the actors at all levels of the foreign aid process and suggest (1) how they tend to drive goals and routines to produce inappropriate results, and (2) how their conflict can ultimately lead to improved foreign aid results. To this point, the model has been static—relatively constant goals and rules. The “role confict” variable is viewed as less of a constraint than as a dynamic, evolving opportunity. In that the character of conflict between major policy actors drives the rest of the system, it is the most dynamic variable. Changing the character of role conflict, therefore, can have greater effects on foreign aid outcomes than changing either goals or rules. Role conflict and resultant trust or distrust affects the translation of goals into policy. Advocates of security, profitability and developmental goals clash among themselves and within each area, making for a highly fluid policy arena, and impeding the effort to derive political behavior patterns beyond ad hoc case analysis. Part of the problem is that it is easier to transfer a weapon or build a road than create a middle class democracy or productive small farmer, given local cultural and political limitations. Straight transfers of resources to build democracy often produce blackmail by groups standing to lose influence, and unintended results. Beyond these problems, which can be and have been overcome with experience in comparative contexts, lies the essential fact that the US policy apparatus allocates resources in ways that are unlikely to produce much security, development, or even long-run profitability. That the goals and rules which reinforce them are skewed in the direction of either technical quick-fixes, professionalization of the military as the

72

The Politics of United States Foreign Aid

appropriate form of government, or long-term “dole” operations is important too. But realistic change depends on the character of the exchange between the institutions which make, execute and monitor foreign aid policies. The role conflict perspective suggests that rival trust will result from realistic conflict between competing guardian and spender roles. Where exchanges are inhibited by lack of role clarity, missing roles, or lack of daily work contact and communication, distrust persists and the conflict is dysfunctional from a policy perspective (Guess, 1984, 1985). Many, such as O’Leary (1967:134) have noted the “challenge and response within the political system” over foreign aid. But conflicts surrounding foreign aid have not produced “a clearer vision of where we are going, or indeed, where we now stand with regard to foreign aid” (1967:134). Conversely, where roles are defined and communication provides mutual understanding, the program is likely to be successful as in the case of the Republic of China (Jacoby, 1966:132; Montgomery, 1962:Chapter 2). Like most US public policies, foreign aid is a mixed-motive bargaining game between institutions that compete and compromise, producing results that are both satisfactory and unsatisfactory but rarely brilliant or disastrous. But owing to the lengthy history of foreign policy security and democratization adventures around the world, and lack of task clarity in the development business, the policy process tends to revolve around security considerations (International Security Assistance including ESF) with rules and projects designed accordingly. Guardians are “cost-cutting” for some combination of domestic political gain and foreign economic reasons. Spenders represent different foreign aid programs and seek to maximize them within guardian limitations. Guardians want to keep the lid on competitors for power, turf, and resources. Victory is often temporary and based on coalitional power. This is due in part to the fact that the interests advocated or guarded may be multiple or contradictory. Congressional defense “guardians” (House and Senate Armed Services Committees) usually “advocate” larger defense and security assistance budgets; USAID and congressional Democratic members of House Foreign Affairs and Senate Foreign Relations Committees have historically advocated more for development aid to “guard” the political center and to stabilize the recipient society. Despite the semantic problem with these terms, there are clear advocates of expanded influence among donors and recipients that are countered to varying degrees by cost-cutters. This general conflict, evident in regional foreign aid cases, can remain superficial to the detriment of foreign aid or produce rival trust and better results. To the extent that roles are ill-defined or missing, superficial confict will permit more powerful actors to receive most of the funding without necessary relationship to recipient need. For example, the Reagan Administration recently shifted policy and will now provide sophisticated Stinger (heat-seeking anti-aircraft) missiles to Third World insurgent forces. The shift occurred, according to Ottaway (Washington Post, March 30, 1986) after “activists in the Pentagon and CIA backed by conservatives in the Senate and elsewhere, overcame opposition by officials in the State Department as well as some in the CIA”! While many adversary relationships exist in foreign aid, no one adversary process exists to critically examine claims of guardians and spenders. As noted, the President and Congress approach the issue from opposite political ends. Foreign aid is a political process largely driven by State-DOD and congressional foreign policy interests. The general

The Bureaucratic Role Conflict Model 73 features of this role conflict process can be viewed in Figure 3.3 with actual changes in conflict over budgetary resources in Table 3.1. Excessive guardian influence in the form of control-oriented rules and insistence on profitability and security tends to produce countercontrol efforts by line personnel and many unintended consequences. Excessive spender influence results in the establishment of bureaucratic wards most of whom favor large capital projects and significant weapons transfers. In that weakened role conflict tends to permit goals and rules to persist, changing goals or rules would be be unlikely to change the balance of power among foreign aid actors. Realistic change depends upon evolution of role conflict between clearly defined positions over substantive issues, leading to rival trust. Such a condition would make possible a foreign aid program that considers “total needs” as indicated on Figure 3.3. Where rival trust exists, actors can engage in realistic problem definition and solution. That the Bureaucratic Role Conflict model can provide general explanatory utility for foreign aid policy analysis, is demonstrated by the behavior of aggregate guardian and spender roles for the: Development Assistance, Foreign Military Sales Credit, Economic Support Fund, and Military Assistance programs. As evident from Table 3.1, lack of communication and distrust, indicated in part by greater percentage differences between budget requests and appropriations, relates to uncertainty among both US donors and Third World recipients. Where trust and communication increase, both the process and results seem to be clearer if not more appropriate. By this theory, over the last 11 years (1975–86), trust and communication seem to have improved. From 1975–1977, differences between executive (DOD, AID, OMB, State, Treasury) requests and Congressional appropriations were large, ranging from –2% to –68%. This may have been due to executive branch in-fighting and the general unpopularity of the issue itself. Between 1977–80, although budget authority declined for foreign aid as a whole, differences between President Carter’s requests and Congress’ diminished substantially, owing in part to Democratic party control in those years. Between 1981–83, the early Reagan years, Congress increased the President’s request substantially, suggesting growing trust among the two branches and role clusters. This also coincided with perceived security threats in the Middle East and Central America. Executive requests grew substantially from 1984–86, bringing them closer into line with greater congressional support for foreign aid in practice. Consensus between guardians and spenders seemed to be increasing that foreign aid programs could contribute to improved Third World stability (agricultural productivity, health, education) and counter the false romantic appeal of terrorism as a means of legitimate opposition. Between 1984–86, differences between requests and appropriations were marginal. This may signal a new era of trust and open the possibility of using foreign aid as a more autonomous policy tool for foreign affairs. In this final section on role conflict, let us briefly examine the occupants of guardian and spender roles in order to clarify how policy is made and how it might be changed. As indicated, foreign aid policy is made by an extremely fragmented system of institutional actors characterized by power plays and shifting alliances. Under such conditions, coherent policy-making and effective results become difficult to attain with any regularity. Since the initial balance of guardian-spender conflict has always been established by Congress, through such laws as the Mutual Security Acts and Foreign Assistance Acts that authorize the program, let us begin by discussing the “guardians” and their leading guard, Congress.

Figure 3.3: Guardian and Spender Role Conflict in US Foreign Aid

1174.6

1158.8

1347.6

1646.3

1757.5

1821.9

1311.5

1389.0

1445.0

2109.7

2104.6

1976

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

2066.4

2226.5

1866.4

1790.2

1847.0

1683.8

1582.5

1531.6

1261.3

1128.9

975.6

626.5

–1.8

+5.5

+29.1

+28.8

+40.8

–7.5

–9.9

–6.9

–6.4

–2.5

–16.9

–35. 9

5655.0

5100.0

1000.0

950.0

1481.8

734.0

658.8

1042.5

707.7

2179.6

2430.2

555.0

% Request Change

5190.0

4939.5

1315.0

1175.0

800.0

500.0

645.0

1024.5

675.8

740.0

1205.0

300.0

FMSC BA

–8.2

–3.1

+31.5

+23.6

–45.9

–31.8

–1.9

–1.7

–4.5

–66.0

–50.4

–45. 9

% Change

4024.0

3438.1

2949.0

2886.0

2598.5

2030.5

2007.0

1854.4

1921.3

1836.5

1908.3

1699.8

Request

3706.0

3826.0

2903.3

2661.0

2926.0

2024.5

1951.0

1882.0

2219.3

1757.7

1739.9

1200.0

ESF BA

–7.9

+11.2

–1.5

–7.7

+12.6

–0.2

–2.7

+1.5

+15.5

–4.3

–8.8

–29.3

% Change

949.3

924.5

747.0

557.0

134.4

104. 4

160.2

133. 5

230.0

279.0

790.0

985. 0

Request

–12.8 –17.5

805.1 782.0

–4.8

–31.2

383.3 711.7

+30.5

+5.7

–31.2

–37.5

–4.3

–5.3

–68.1

–51.7

% Change

171.4

110.2

110.0

83.3

220.0

264.5

252.0

475.0

MAP BA

Source: USAID, 1986; Defense Security Assistance Agency (DOD), “Congressio Presentation For Security Assistance Programs, “Volume 1, FY 87, pp. 89

977.4

Request

1975

FY

DA BA

Table 3.1: Budget Authority Requests and Appropriations For US Foreign Aid: 1975–1986 ($000)

76

The Politics of United States Foreign Aid

Congress reauthorizes (usually on an annual basis but biannually in FY 86) the foreign aid program and provides it budget authority (appropriations) based largely on its own preferences mixed with the clamor of single issue groups that lobby for and against the different aid programs. Congressional aid preferences are usually broad, such as foreign assistance to the Philippines for US bases, Liberia because freed slaves established it, and Costa Rica because of its good democratic image. Occasionally preferences are narrower, such as Senator Mark Hatfield’s (D-Oregon) recent vote to provide poplar trees to Nepal from one of his constituent’s nurseries. Congress enacts foreign aid legislation amidst a barrage of special interest groups advocating: CARE, PVOs, and specific countries such as Israel (AIPAC), Greece and Cyprus. But, as noted, in contrast with other US domestic policies, no one group advocates the foreign aid program itself. Of the various positions taken by Congress, those predominant are: “national interest”, “fence-sitters”, and “opponents” of foreign aid. Those least taken are: humanitarian and moral responsibility, as well as new approaches (Morss and Morss, 1982:79–83). With increasing terrorism and attacks on US personnel abroad, congressional support for a security definition of aid has been increasing. Part of this may be overall impatience or “donor fatigue” with the pace of Third World development. For example, despite two years of economic austerity and recession (1982–3), Mexico is again on the brink of a debt crisis ($103.7 billion, the debt service of which is 54% of exports!) (Rowe, June 8, 1986). Other countries in the Middle East and Central America have also been slow to develop for a variety of reasons. On the other hand, conflicts between the authorizing and appropriating committees in Congress over setting foreign aid policy, tend to reflect the new activist spirit that the US can change the world, such as by providing aid to “freedom fighters” in Afghanistan, Nicaragua and Cambodia, and depose dictators in Haiti and the Philippines by aid cutoffs. Balance between military and economic aid is finally worked out in markup sessions of the House Appropriations Subcommittee on Foreign Operations. But much of the current balance is the product of conservative views of “guardianship”, meaning US power to change the world by military aid (Congressional Quarterly, December 21, 1985:2688). Even “fence-sitters”, who know or care little about development issues, and those who oppose aid “giveaways”, can be roused by the clarion call of “needed” defense and security aid to prevent a leftist takeover. Given this overall orientation, it is no surprise that the many farming, civic, humanitarian, and business groups often have little influence on foreign aid policy-making and execution. Their absence from the process tends to perpetuate lack of clarity in both authorizations and appropriations bills. “These groups are rarely organized and there is no general public constituency for foreign aid. And since the president is more concerned about having a foreign aid bill than the specific form it takes, foreign aid legislation is a fragmented ‘grab-bag’ of initiatives that are often contradictory in intent and intended result” (Morss and Morss, 1982:78). Such a fragmented planning and approval process can lead to such anomalies as projects to keep people in rural areas to increase agricultural productivity, while others lure them into urban areas to work in small industries (1982:78). Kaplan also noted (1967:257) that interest groups gain access on issues relating to commodities, technical assistance and loans. But they are passive on programs on which they have no stake, leaving most of foreign aid to be designed by a largely uninformed Congress or security-dominated executive branch in many cases.

The Bureaucratic Role Conflict Model 77 For Congress, “guardianship” means largely narrow cash flow analysis of usually inflated executive branch figures to arrive at marginal cuts each year. This lack of intelligent oversight means that Congress alternates between acquiescence in executive branch definition of priorities, and interference to satisfy its narrow preferences via earmarks and conditions on aid. Neither stance is well-informed by analysis or information. Congress votes budget authority for aid by broad functional uses such as health, and in some cases as Latin America, by region. This generalized authority means that Congress feels little responsibility for actual apportionment of funds, in other than a narrow cash-flow sense. Funding decisions are often based on “visceral reactions” and “not a realistic weighing of alternatives” (Kaplan, 1967:255). Throughout the history of the foreign aid program, Congress has occupied the rather consistent role of “carping critic and reluctant accomplice” (1967:256). Moving from more general observations on congressional guardianship to policy specifics, it is clear that, like AID itself, Congress makes policy in an arena of decentralized fiefdoms, each vying for resources and decisions. Weaker leadership and the unpopularity of the issue prevented the authorizing committees (Foreign Affairs and Foreign Relations) from passing authorizations bills from FY 1982 to FY 1985. Much of the legislative and budgetary “guardian” function therefore combined in the appropriations process in those years and currently (despite the required authorization before appropriation). Prior to FY 1982, annual foreign aid authorizations and appropriations were the principal vehicle for exercise of congressional influence on foreign aid. Authorization statutes set policy guidelines and funding ceilings, while appropriations bills emphasize fiscal management. In 1982, the Reagan Administration found it could achieve its foreign aid objectives (El Salvador and Caribbean Basin Initiative) via supplemental appropriations without the need for going through the authorizations process. By 1984, 80% of foreign aid appropriations were enacted without authorizations (House Committee on Foreign Affairs, 1985:75). Hence, little executive pressure existed for authorizations and Congress followed suit with its traditional lack of interest in a politically unpopular issue. As noted, from FY 82–86, there were no foreign aid authorizations. This meant foreign aid was funded by continuing resolution and supplemental appropriation. While the aggregate pattern indicated by Table 3.1 seemed stable and improving, distrust within foreign aid actors ran high. For example, distrust among authorizations and appropriations committees was extremely high as the latter began to make legislative policy. The most important consequence of the breakdown of the authorizations process has been reduced debate on foreign aid issues in both House and Senate. In 1982, during 37 hours of debate on the Continuing Resolution, the Senate spent about 2 minutes on foreign aid (1985:72–78). The predominance of appropriations committees has also changed the character of foreign aid priorities. Strong leadership in the House Foreign Operations Subcommittee, for example, (David Obey, D-Wisconsin) as well as the House Appropriations Committee (Jack Kemp, R-NY) and Senate Appropriations Committee (Daniel Inoye, D-Hawaii, and Robert Kasten, R-Wisconsin), has reinforced the AID and State preference for long-term aid (DA) over straight cash (ESF), favored for its flexibility by OMB. Priorities are thus being largely determined by the Appropriations Committees instead of in the traditional constitutional pattern by Authorizing Committees first. Congress “guards” foreign aid in several ways. First, it “earmarks” funds to specific uses, usually to keep control of foreign aid and to express its own preferences. In FY 85, for

78

The Politics of United States Foreign Aid

example, the Senate Appropriations Committee recommended an earmark of $1.2 billion for Israel, $815 million for Egypt, $15 million for Cyprus, and $6 million for judicial reform in El Salvador out of a total ESF request of $3.8 billion (Senate Appropriations Committee, June 24, 1985:71). In FY 84, the Continuing Resolution for foreign aid contained 31 earmarks that set specific aid levels for certain countries and 55 other provisions restricting or prohibiting Presidential action. By contrast, in FY 81, the numbers were 15 and 23 respectively (House Committee on Foreign Affairs, 1985:20). But, as noted, Congress also earmarked $15 million to Uruguay. AID had initially planned to “break” the earmark (via the Section 614 “waiver” procedure) but decided agaist it. Particular committees attach earmarks to foreign aid appropriations to try and force the Administration to do what they want, such as aiding Egypt, Israel, Turkey, the Philippines and the “Child Survival” program. Since about 67% of ESF aid is earmarked, this reduces AID ability to effectively manage its activities. The extent of many congressional appropriations hearings, according to one AID official (Interview, March, 1986) is to find out merely if AID will go along with proposed congressional earmarks. The use of the remaining unearmarked funds (33% of ESF) by AID must also be approved by State Department. Second, Congress often provides AID with more funds than it requested. In the FY 83–86 period, Population Planning, and Health functions received substantially more than requested, indicating congressional preferences and leadership in these areas. Such an appropriation (actually continuing resolution) pattern clashes with AID priorities that emphasize agriculture and selected development activities, such as small farmer forestry and alternative energy production for LDCs. As indicated in Table 3.1, this happened in FYs 83–84 in both DA and FMSC, suggesting that in addition to the increased trust explanation, it is also possible that Congress was stepping on both AID and DSAA to achieve its respective preferences in development and arms sales. Congress then plays the “guardian” role over executive aid requests and activities. Much of the role conflict is adversarial. However, it should not be forgotten that AID, State and DOD counterpart personnel are almost in daily contact. This pattern tends to work in the opposite direction, meliorating conflict and pushing the relationship from adversarial to symbiotic and non-adversarial in many cases. As in the analysis of any public policy issue, the particular balance of power depends on the foreign aid issue at hand, administrative and congressional leadership, available power resources, and the existence of an election year. This does not mean that one must engage in “political particle physics” to discern what is occurring and why. For this purpose, the budgetary/bureaucratic role conflict (BRC) perspective will serve as a preliminary if not adequate tool. The BRC model is based on the proposition that rival trust develops from realistic conflict. By definition, conflict is confused and crisis produces distorted information which tends to reinforce distrust among policy actors. But, over time, adversaries gain knowledge of each other’s behavior and the results of joint activities tend to improve. Congress distrusts the executive branch because of its technical expertise and capability to justify inflated claims. Interest groups and policy expertise networks provide the framework for exercising influence over policy in many cases. As noted, foreign aid lacks over-arching interest groups and this works against comprehensive and coherent policy-making. Congressional distrust of executive foreign aid policy-making is evident in the tendency to “earmark” funds, particularly ESF, and to attach binding restrictions, prohibitions on

The Bureaucratic Role Conflict Model 79 administrative policy, reporting requirements, and non-binding “sense of Congress” controls on foreign aid funds. The example of Uruguay is instructive. AID would like to use these funds ($15 million) for Africa or other needy areas. Congress (representing the particular interests of several members who visited there in 1985) would like the money to go to Uruguay. As indicated in Table 3.2, after all the earmarks, only $1.32 billion is left over for “unearmarked” (AID preference) countries after Congress finishes its purely political allocations. AID advocacy is also based on internal agency politics, but largely on the results of its process of programming and budgeting aid funds. However flawed that process may be, it is an attempt to inject “rationality” into a highly politicized arena. Nevertheless, the distrust between AID advocates of particular uses of DA, PL 480 and ESF funds and Congressional ESF earmarkers runs high. Congress prevails in most cases because its consensus-building activities are more sophisticated than AID’s, which must run a rear guard operation against not only Congress but also State as well. For instance, PL 480 has an excellent congressional constituency; ESF which is administered by AID does not. Only where all foreign aid is under attack, such as in FY 87 by Gramm-Rudman and mid-term congressional elections, can AID hold the line by unified executive branch support and the rather hyperbolic argument that aid is a “national security” matter. To the extent that AID advocacy can attract more powerful allies, its conflicts with Congress can produce greater trust and improved aid results. On the other hand, distrust among foreign aid advocates has also been historically high. In the past Republicans simply favored military and ESF funding while Democrats favored DA and PL 480; Democrats favored soft loans and Republicans more profitable aid vehicles as trade or hard loans. DOD favored military aid; AID favored DA and State favored political uses of economic and military aid. Aid requests reflected these differences in perspective and the entire aid budget policy process was fragmented and incoherent. Each agency bureau would individually present requests to the Secretary of State, similar to the federal budget process without the “executive budget” (before 1921) where each agency would seek individual funding from Congress. In FY 82, Secretary Haig tried to control this with initiation of the “Integrated Foreign Aid Budget” request which put individual agency biases together in one process to produce one packaged request. Instead of many aid requests from DOD, Treasury, AID and so on, OMB in FY 86 has a “single point of contact” according to one State Department official (Interview, March, 1986) in the Undersecretary Table 3.2: Congressional Earmarks For Economic Support Fund FY 86 ($000) Country

FY 1986 Request

FY 1986 CR

FY 1986 G-R

Israel

1,200.0

1,200.0

1,148.4

Egypt

815.0

815.0

780.0

Cyprus

3.0

15.0

14.4

Portugal

80.0

80.0

76.6

Tied aid



50.0

47.9

So. Afr. Reg.

30.0

30.0

28.7

Tunisia

22.5

20.0

19.1

80

The Politics of United States Foreign Aid

Philippines

95.0

125.0

119.6



15.0

14.4

Uruguay



15.0

14.4

Ecuador

15.0

15.0

14.4

TOTAL Earmarks

2,260.5

2,380.0

2,277.9

Total request/approp.

4,024.0

3,706.0

3,546.6

Amount remaining for other countries

1,763.5

1,326.0

1,268.7

Afghanistan

Source: US Agency For International Development, 1986.

for Security Assistance, Science and Technology. In May, coordinated by the State Department’s Politico-Military Affairs Bureau, the Secretary’s Assistant Policy Working Group (SAPRWG) coordinates policy requests by the International Security Assistance Bureau of DOD (DSAA), the regional bureaus of State and AID as well as Treasury and OMB. In July of each year, the State Undersecretary of State For Science and Techology (“T”) oversees development of the Integrated Request. “T” reviews country (13 countries receive 80% of US aid) and program requests by AID (DA), US Department of Agriculture (PL 480), Treasury (MDBs) and the SAPRWG. Science and Technology tries to balance military and economic needs by country and assess congressional preferences on each. An informal appeals process is available for dissenters on the final balance and direction of aid, and the harder issues go to the Undersecretary and Secretary of State by September (before the FY begins on October 1). Unlike the stormy relationship between congressional guardians who earmark “pet” projects and concerns that tie the hands of AID technicians in the field, over the last several years the level of distrust between executive militaryeconomic aid advocates has decreased and roles seemed to have sharpened. Table 3.1 budget data is supplemented by additional data here noting that in FY 83, 13 major issues ranging between $10–$100 million were appealed to the Secretary of State. In FY 87, only one issue was appealed, suggesting greater consensus, trust, healthy conflict and probable improved results. The opposite interpretation, of course, is that major actors have been co-opted in a process of apparent pluralistic tolerance, narrowing the range of debate and conflict, with worse results likely for the world’s poor. Consistent with the BRC perspective however, this new consensus is based on a lengthy history of severe conflict suggesting that it has a realistic basis. No contemporary discussion of the congressional role in foreign aid would be complete without an examination of the effects of the new budget-balancing act known as GrammRudman-Hollings or (GR-H) on programming and budgeting. So far, this new rule acts as both a constraint and opportunity and hence will be discussed under the “role conflict” section. The Balanced Budget and Emergency Deficit Act of 1985 established annual deficit targets and, if not met, provides a formula for reducing the level of deficit annually until it reaches zero in 1991. The formula, calculated by CBO, OMB and GAO is used to produce a uniform percentage of “outlay savings” across “programs, projects and activities” (PPAs) through requirement of a predetermined “sequestration” order. G-R-H is designed to “restrict Presidential discretion in shaping budget reductions and reshaping funding priorities to meet presecibed deficit targets” (Maroni, Nowels, and Woldman, 1986:35).

The Bureaucratic Role Conflict Model 81 Hence. Congress affects foreign aid scope and purpose via earmarks, regular appropriations and now G-R-H. For FY 87, PPA outlays must be reduced by 4.3%. The problem with G-R-H is its uniformity, i. e. cuts are at this flat rate. In foreign aid, savings as a percentage of budgetary resources really reflect “first year spend-out rates” and these vary by program (Maroni, Nowels, and Woldman, 1986:12). For instance, nearly all MAP funds appropriated are spent in the first year, while DA funds are spent much more slowly. Whether this cutback formula is efficient, fair or useful depends upon particular circumstances. But more likely than not, the G-R-H effects will be to reward programs that can move money quickly (security aid) and create an incentive for DA to do the same, which given the volatile contexts of development projects, is likely to increase the number of unintended or unprogrammed consequences. Aid policy-makers can always reallocate reductions required by G-R-H. They can do this through existing special authority which enables them to modify congressional earmarks and change country and program allocations during the FY. For instance, AID can still reprogram and transfer funds between program accounts after “sequestration” has taken effect (Maroni, Nowels, and Woldman, 1986:20). Unfortunately, this opportuity to reallocate resources to achieve outlay savings takes time, resources and trust. Recall the difficulty AID currently has on gaining clearance from OMB on many smaller deobligations and reobligations to achieve similar resource savings. Given existing AID, DOD and State programming practices, it is likely that more efforts will be spent on planning and programming and even less on improved implementaion of projects. In this sense, G-R-H will be a serious constraint to careful and sensible foreign aid policy-making. To balance the budget, more money will effectively be wasted! Congressional relations with the relevant executive aid actors has improved. But this may not be the same thing as improved or “intelligent” guardianship. Despite new programming and budgeting procedures in the executive, Congress still reviews the annual aid request in piecemeal and politically fragmented fashion. If realistic or intelligent guardianship were its purpose, Congress could, for instance, authorize and appropriate funds by country. In response to energetic country lobbying then, Congress would have to beef up its foreign aid review process and explain how its decisions conform to US-recipient interests. Congressional appropriation hearings by country (not just current hotspots such as Nicaragua) could also reduce executive control and flexibility since much of foreign aid is predicated on treaties and foreign relations reserved for Presidential control (Kaplan, 1967:256). But the current practice of congressional funding by function or region from a continuing resolution (since FY 1981), permits free play of partisan squabbles that clutters up the aid bill with riders and restrictions. This is unrealistic conflict. Though well-intentioned, the many small lobbies for housing, environment and human rights, as well as the more powerful groups such as AIPAC, often hamstring execution of projects. Such broad congressional oversight also effectively delegates authority to the executive branch (as in foreign policy) where DOD and the security perspective on foreign aid control resources. As noted, AID spends much of its time rewriting CDSSs while, paradoxically, many of later budgetary reviews exclude programming issues. Much of this is due to annual reauthorizations and the knowledge that any congressional questions are unlikely to relate to country specifics. The rest is due to the generally fragmented and questionable information Congress uses on

82

The Politics of United States Foreign Aid

which to base aid authorizations and appropriations votes. The fact that the congressional process alone includes: Senate Foreign Relations, House Foreign Affairs, House Money and Banking, Senate Finance Committees, House and Senate Armed Services, Appropriations, and Government Operations, means a “pluralism of responsibility and effort” (Montgomery, 1962:270) which impedes development of a national foreign aid consensus. Hence, the executive can afford to avoid the substantive issues since Congress will too. Changing the role of Congress to aggressive guardian could be accomplished by requiring country appropriations and providing multi-year appropriations for DA (most DA is annual while ESF is usually multi-year). This would give Congress time for substantive review of particular countries over staggered periods of time, and not in response to country “crises”. It could also reduce the traditional bureaucratic incentive to obligate all appropriated funds by the end of the FY to avoid “unliquidated obligations” and reversion to the Treasury of unused funds. With a more active and intelligent role, Congress could give aid based on balancing of the total recipient behavior pattern (“Total Needs Funding”, Figure 3.3). The recipient would be rewarded on the basis of responsiveness to broader US developmental concerns, not simply maintenance of stability. With a reoriented executive budget process that focused on field issues as well as internal programmatic considerations, Congress could be forced to consider economic aid in conjunction with strong constituency pressures for defense procurement and home district benefits (Woll, 1985:450). Current activism by leaders of the appropriations subcommittees on foreign operations suggests what an aggressive Congress could do for foreign aid. This should be institutionalized by requiring rolling examinations of all countries for budget authority over a 3–5 year period. Though Congress is neither “intelligent” guardian nor spender at present, USAID occupies the mixed role of both guardian and spender. AID/W is guardian of its 60 field missions, which also are both guardians and spenders! This mixed or dual role has been noted in implementation of “New Directions” (Mickelwait, et. al., 1979) and by GAO (1985) in the CDSS approval process. AID/W seeks program performance benchmarks from its missions in an attempt to prevent loss of control at best, “going native” at worst. Missions, on the other hand, want flexibility. Overwhelmed by guidances for preparation (GAO, 1985:14), missions know that they are unlikely to be approved on the first cut and despite the A-384 mandate, know that their budgets are not totally tied to CDSS results anyway. So, they often engage in “counter-control” tactics, tangling up the AID/W-Field Mission relationship with red tape and formalism. This kind of goal displacement behavior is functional in that it prevents excessive AID/W intrusion. The only potential losers are the Third World recipients. But, as one AID technician told me, “Since when has an Ecuadorian peasant resulted in a Mission add-on?” As noted, even AID/W is a spender at annual budget time. Several other federal agencies have a strong voice in US foreign aid policy-making (besides DOD). Section 204 of the 1961 Foreign Assistance Act required AID to consult with the Treasury, Commerce, Eximbank, and State Departments before approving any project, through meetings of what was then called the Interagency Development Loan Committee (Tendler, 1975:44). Based on its responsibility for US balance of payments, Treasury then “had veto power over all AID project proposals” (1975:44). As noted previously, State also intrudes by setting “long-range annual resource levels” (GAO, 1985:3) based on US security interests and developmental needs of recipients. By congressional fiat, USAID thus delegates advocacy in functional development to other

The Bureaucratic Role Conflict Model 83 actors that are more generally interested in foreign policy (State), US balance of payments issues which tend to view soft loans and grants, and local-cost financing as outflows of US dollars (Treasury), trade (Commerce), and security (DOD, President and Congressional Armed Services Committees). So, State Department, for example, is a foreign aid “advocate” (Morss and Morss, 1982:75) but it also “guards” heavily its foreign policy definition of aid and ensures that AID projects are coinsistent with it. We therefore include State as guardian because, in one sense, all actors advocate foreign aid (usually for the wrong reasons!). Until FY 82, a clear problem of interagency coordination existed between actors that sought more funding, operating responsibility and control over specific projects. State Department, with the aid of its new “integrated” request process, is still responsible for coordination of scattered foreign aid efforts. The net effect of this mixed-motive guardianship is a superficial kind of conflict over resources to increase turf, build larger projects, and displace developmental goals with hard and fast rules. As if the domestic side of the foreign aid equation were not sufficiently damaging to rational policy-making, foreign aid also includes recipient actors who conduct their own annual turf and resource wars. The scope and intensity of guardian-spender relationships is often influenced by host-donor relations. In particular, the US administers aid to recipients in three ways: (1) Jointly, such as the Joint Commission on Rural Reconstruction in Taiwan (1948), (2) Counterpart, where local partners work with US technicians for purposes of intergovernmental coordination and exchange of techncial data (the most common), and (3) Liaison, where the host plans and executes the project and the US approves general design, such as in Burma, 1956 (Montgomery, 1962:161–165) or Israel, 1986. Over the years, recipient representatives have learned how to play the aid bureaucratic game, using US goals, rules, and lack of realistic communication among key actors to their own advantage. As will be argued below under “spenders”, the prestige, career advancement and foreign travel opportunities for recipient civil servants associated with foreign aid projects means that “any but the most xenophobic” (McNeil, 1981, 56) would favor acceptance of foreign aid. However, civil servants in recipient staff ministries such as Treasury, Planning or Finance tend to be less interested in maximizing amounts of aid to their agencies and more concerned with the “relative attractiveness of the terms of aid and priority of the sector concerned” (1981:56). They are spenders in the sense that they want more aid, but guardians (Figure 3. 3) in that they are concerned with its effects on the national economy. They may try to minimize the debt service burden to their countries and broker the best terms possible with donors such as USAID. Though guardians have a spending “streak” in them (mixed roles), spenders have few tendencies to hold off and “guard”. Their incentives are to maximize amounts of resources and influence for their respective institutions. As in the case of guardians, spenders in most cases represent interests several steps removed from tasks or activities likely to produce “development” or “stability”. They are concerned with short run self-interest maximization in a policy area that requires some attempt to also maximize the recipient’s long-term interests. Neither guardians nor spenders cover this “value”. Nor is it the logical product of conflict between self-interest maximizers, where those interests are structured by rules and repertoires to avoid the substance of foreign aid issues.

84

The Politics of United States Foreign Aid

In that 68.7% of total foreign aid funding requested for FY 87 will be security-related (up from 67% in FY 86) and 42% of all foreign aid funds were appropriated for the FMSC program in FY 86 (USAID, 1986:6) it can be said that DOD is perhaps the most successful advocate in the foreign aid process. Its interest in weapons and materiel transfers via MAP and FMSC is complemented by congressional interest in procurement patterns that benefit home district constituents. With inordinate budgetary influence, as noted, comes excessive policy-making influence. Problems of aid to the illegitimate regime, such as Vietnam (Montgomery, 1967:55), Somoza’s Nicaragua, Marcos’ Philippines, and Duvalier’s Haiti, were really problems of fending off what could have been total military solutions, advocated by DOD in most cases, and creating balance between host regime interests in obvious selfpreservation and US concerns for economic development. Whether or not a precise proportion of economic-military assistance can be programmed in advance, the two viewpoints on this (“hard” versus “soft” development schools as noted in Chapter 1) should be able to argue their cases before neutral tribunals. But, despite new programming and budgeting procedures in the executive, Congress still responds to security pressures more readily. AID must still clear its projects with State and Treasury and OMB for all inter-fund transfers and savings. Congress tends to be “hard” schooloriented unless a crisis occurs in which confused reversals of position follow. Guardians, in other words, still do not realistically compete with spenders on an equal plane despite apparent consensus indicated in budget requests and approvals (Table 3.1). Naturally the major US foreign aid advocate is USAID, for DA and its own definition of how ESF funds should be allocated. It seeks to maximize approved projects consistent with existing rules and goals set for it by significant others. AID is politically weak in the federal bureaucracy and among congressional officials. Headed by a semi-cabinet level person of usually nondescript status, AID is “second fiddle” to State, and sometimes Treasury and Commerce in foreign aid policy-making. Since FY 81, as noted, AID actually profits from the obscurity of its requests buried in and approved by continuing “appropriations” resolutions. Lacking a substantive mandate from the product of interest group competition or through thorough congressional review of each country’s claim to aid, AID relies on an internal group conflict process to produce its requests. But, as noted, this process has been merged with the claims of more powerful actors, like DOD. As the security proportion of foreign aid increases, this suggests that even as an advocate, AID merely sings what it is told by others even more distant from the everyday reality of the Third World. To the extent that the “hard” school is an inappropriate means of developing the Third World via aid, AID as an advocate is passive and foreign aid is on the wrong track. One group of spenders that cannot be characterized as “passive” are the contractors or intermediaries between donor and recipients that carry out the actual work of development in most cases. Many contractors perform excellent work and can be considered a vast improvement over work that would have been performed by AID itself. Often contractors recruit university personnel that work on AID projects for research and real world learning experiences, the results of which can later be integrated into coursework or publications manuscripts. Contributing to their widespread availability is the fact that academics on lean, nine-month salaries can always find time to earn extra money. So contractors can deliver usually excellent personnel and this is evident at the project bidding stage when competing

The Bureaucratic Role Conflict Model 85 proposals are examined by committees of AID personnel and recipient representatives. Occasionally AID will attempt to push in its retirees or near retirees to head up projects by defining the positions needed in the “request for proposal” to match their qualifications. But otherwise, the contractor can usually form their own team. On the other hand, contractors are also profit-seeking firms (RCA International, Booz-Allen, Arthur Young, etc.) and to the extent that “they have freedom of decision— for example with regard to the detailed design of a road, or the technology to be used—their self-interest may run counter to that of the prospective beneficiaries of the project” (McNeil, 1981:57). The degree to which they can distort a project depends largely on will and capacity of AID to evaluate their qualifications (during bidding) and recommendations (during execution). Frequently, contractors can determine the course of both economic and security assistance and this, of course, raises substantive questions on accountability in a democratic society. At the least, contractors can rip off taxpayers by doing “feasibility” work that AID professionals should be doing themselves, such as preparing PIDs and PPs, and evaluating projects, rather than spending the day in CDSS or ABS meetings. At the worst, contractors can change policies including technology for which they are the primary beneficiaries through increased sales, e. g. RCA electronic equipment. The tendency of contractors to direct foreign aid is unfortunate in that it adds another layer of bureaucratic chaos to an already fragmented arena. AID is already directed from above by several mutually conflicting advocates, such as DOD and State. Despite the case one could make for checks and balances from multiple interest groups, foreign aid policy-making tends to become less rational and coordinated with the inclusion of each new actor. Consultants are advocates but to the extent that they respond to the financing agency instead of recipient interests, they are part of the confused web of participants that contribute to incoherent foreign aid policy-making and results. Indeed, McNeil (1981:60) finds consultants to be an embodiment of the major contradictions in foreign aid. They are not accountable to the donor taxpayer or the recipient people. They are not even accountable to recipient political institutions. They are accountable to the donor which, in the case of AID, probably means little for policy determination, but more in contribution to waste and frivolity. Consultants to DOD or State, by contrast, might actually contribute to policy-making. As indicated repeatedly, domestic US interest groups are peripheral to the policy process in general. They lobby for changes in aid to benefit such causes as human rights (cutoffs to Guatemala) women in development, environmental protection, US farmer commodity export growth or protection, and so on. But they are single interest groups and where they have no stake, they do not participate in foreign aid directly or indirectly. Though this is entirely rational and consistent with the Downsian notion of a representative democracy (1957), it leaves foreign aid policy-making to either bureaucratic whim or to the most powerful groups. In this case, public policy is determined by DOD-State security advocates, powerful single-interest lobbies such as AIPAC (Israel), and a usually acquiescent and illinformed Congress. Again, the pattern of advocacy in foreign aid is fragmented enough. The picture is complicated further by recipient politicians and civil servants, both of whom stand to gain as much or more than US consultants. Third World politicians have a vested interest in delivering the goods to cheering constituents. Politicians and ministers often have constituency groups (voters, business groups, farmers, the military) to please as well as personal material interests.

86

The Politics of United States Foreign Aid

McNeil (1981:54) suggests that “self-interested concerns by those in power have led to some of the more stark failures in development aid. A steel works may be located far from both deep water and iron ore deposits but within the electorate of the minister concerned. Clearly aid donors are not unaware of such facts, but they often find it hard to resist the pressure. “Project location is then often distorted. But recipient politicians also seek visible projects with quick returns for their own benefit. These interests mesh nicely with current foreign aid rules that favor precisely such projects in the interest of security, stability and avoidance of giveaways by financing local recurrent costs. Recipient line ministries also advocate aid as a means of maximizing their agency budgets as well as for personal prestige. As noted, staff ministries can resist the more paltry reasons for hustling aid, such as an agro-industrial project in the home district. But they too gain in prestige, driving around in FAO, UNDP, or AID-ornamented wagons, or wearing an IADB lapel pin or suitcase tag, which mean lots in some places! Perhaps association with the well-meaning but lackluster AID-types brings less prestige and even terrorist threats of late. For most career civil servants, the ticket to career success is indeed a foreign, i. e. First World, experience such as study or employment. The next-best item is home country work with a US-sponsored aid project. Many Third World countries restrict line abilities to contract foreign loans since they recognize the open-ended texture of such incentives. Again, spenders maximize selective self-interests, so do guardians. Usually excluded are debates about the substantive foreign aid issues—institutional arrangements that would actually work instead of trying to please all parties, creative financing, why consultants should be doing training and technical assistance, sufficiency of institutional capacity, and so on. Guardian and spender roles are not sufficiently developed to deal with them at present. In this regard, the foreign aid policy process, like the processes of recipients, is underdeveloped.

Conclusion To conclude this chapter, the BRC model suggests that so long as role conflict is superficial, archaic rules and goals will contribute to foreign aid results that encourage dependency, complexity and unintended consequences. Many success stories are currently achieved in spite of the policy process that provided aid funding. Uninformed guardians in Congress tend to try and maximize US constituent interest with legislative conditions to which contractors must conform. These disincentives for development often lead to large capital projects with security emphases that prove damaging to the recipient’s view of other “soft” DA projects. While establishing the overall framework, Congress delegates responsibility for foreign aid policy-making to advocates such as DOD and State which reinforce the more irrational global aspects of the policy process. The BRC model predicts continued difficulty for foreign aid unless roles are clarified and institutional incentives are changed. The recommended changes will be presented in the final chapter following a “test” of the BRC framework in three US foreign aid settings: Latin America, Asia and the Middle East.

Chapter Four US AID TO LATIN AMERICA

Introduction In many ways, the present foreign aid program is an outgrowth of the model developed during World War II with Latin America. The US provided technical assistance loans and grants from the Reconstruction Finance Corporation “to speed up the production of raw materials” (Montgomery, 1967:38). The goal was some combination of Latin American socioeconomic development and support of US security interests, in particular successful Allied conduct of the war. Very early, foreign aid was caught in between two increasingly contradictory objectives: Latin American development which implies independence and sovereignty, and US regional security interests which implies dependence on US military technology to deal with internal (revolutionary) and external (border conflict) threats. Military aid through FMSC and MAP, for example, to Latin American military regimes have generally not led to socioeconomic development (however defined)—usually quite the reverse. Though Latin America was really the first regional foreign aid recipient, it has received the least funding of all regions except Africa. From 1946–1984, Latin America received only $14.4 billion in loans and grants of military and economic assistance, including loan repayments and interest (USAID, 1984:33). By comparison, in the same period, the Middle East received $71.4 billion, Asia $59.8 billion, Europe $37.1 billion and Africa $12.3 billion (USAID, 1984). The same pattern is repeated for aggregate economic (DA, PL 480 and ESF) and military assistance (FMSC, and MAP). For FY 86, it is estimated that Asia and the Near East will receive the lion’s share of both, $4.7 billion and $5.3 billion respectively. Africa will receive the least amount, $825.7 million and $96.6 million respectively. Latin America and the Caribbean will receive $1.3 billion and $234.7 million respectively (USAID, 1986:17). However, of the $833 million requested in ESF funds for FY 86 for Latin America. 72% or $600 million is proposed for Central America, meaning security-related projects. In fact, the recent increase in DA funding for the region, from 20% in FY 84 to 29% of all regional allocations in FY 87 (USAID, 1986 Congressional Presentation, p. 11), was based less on real socioeconomic need than on the political instability and potential security threat from the region. Latin American poverty is quite real. But it doesn’t compare with the abject squalor of Asia and Africa. Nevertheless, according to one AID official (Interview, March, 1986) only New Directions programs attempted to correct this imbalance. FY 1977 was the lowest DA funding for Latin America in decades as the legislation attempted to redirect funding toward real need areas. But coinciding with the withdrawal of the British from Central America in the late 1970s, and deterioration of the sociopolitical situation there, the US increased funding until, by the early 1980s, Central America became the largest recipient of funding in

DOI: 10.4324/9780203840184-4

88

The Politics of United States Foreign Aid

the Latin American regional bureau. For FY 87, Central America will receive 62.5% of all Latin American funds (Nowels, 1986:15). Recent funding efforts simply continue the historical pattern of supplying aid funds where security crises occur. As indicated in Figure 1.2, the Alliance for Progress period generated substantial aid funding to prevent more hemispheric Cubas, and aid to Central America has been growing astronomically since 1980 to prevent the region from igniting into total war. As aid to El Salvador increases, for example, the ratio of military to economic aid has increased from 1:1 in FY 81 to 4:1 for FY 84 (DSAA, 1986 Congressional Presentation, Volume 2, p. 47). El Salvador now ranks 5th in recipients of US aid and Costa Rica moved up the ladder to 13th place (ESF and DA-PL 480) (Nowels, 1986:16), to prevent the crises in Nicaragua and El Salvador from spilling over into the one consistently functioning democracy in the entire region. By contrast, with the exception of this year’s congressional attempt to earmark $15 million to Uruguay, South America has been receiving less aid funding. Brazil, Argentina and Uruguay are the regional “graduates” of the US foreign aid program which now rely almost exclusively on harder loans from private, mostly US, international banks. This is not supposed to be another book about foreign policy. Aid is a significant component of US foreign policy but it is usually an effect and not a cause. In most foreign policy matters, military issues tend to be quickly resolved. Security needs can be quickly met by rapid transfers of direct FMSC or MAP assistance, indirect ESF, or a quasi-secret transfer such as unappropriated Special Defense Acquisition Funds to “meet urgent foreign needs for military equipment” (DSAA, 1986, Volume 1, p. 108). By contrast, trade and economic issues such as lifting quotas or renegotiating tariffs must pass through endless Presidential and Congressional debates which take time. This leaves the bulk of immediate problems to be dealt with by extending or withdrawing DA. Hence, foreign policy consists largely of the rapid use of security aid and the slower use of economic aid coupled with trade summits. By itself then, economic aid achieves few successes so dramatic that they infect the rest of foreign policy, or few failures that jeopardize otherwise appropriate foreign policies. As will be evident from our survey of Latin America, this situation is unfortunate. Economic aid, with select military uses, should be the major tool for developing poorer countries. As proposed by the BRC model, economic aid has not been able to achieve this potential because of (1) the predominance of security objectives in foreign aid and policy, (2) defensive bureaucratic routines that stress project formulation but not execution, and (3) superficial conflict and distrust between major foreign aid policy-makers. To repeat an earlier statement, foreign aid has been used as a “political clay pigeon” and not as an end in itself.

Results Foreign aid programs and projects seem to succeed in Latin America as the larger thrust of foreign policy fails to stabilize or democratize the region. This paradox is due in part to the tendency of foreign aid projects to (1) generate excessive expectations, and (2) succeed according to some of the narrow project performance criteria noted in the last chapter (which are designed less to indicate performance than to fend off critics). Though gross

US Aid to Latin America

89

results claimed by AID in its “success stories” are largely accurate, these are individual project results. The net results of individual project or sectoral evaluations that tend to ignore the larger context of underdevelopment in this region are: dependency, complexity, and large doses of unintended consequences. As indicated, these results may not all be negative; some may even be required for development. But the major implication is that foreign aid is planned according to US bureaucratic and political needs, and executed without the local assistance of those who represent real clients and their needs. Like most societies, Latin American political cultures are riddled with contradictions. To the Western observer, they are more obvious in that they get in the way of doing business like “cultural red tape”. Many argue that foreign aid results are mostly a function of meeting and controlling Latin American political cultures. Harrison (1986), for instance, argues that Hispanic cultures, values, and attitudes “are the principal obstacle to progress in Latin America”. He suggests that underdevelopment is not a problem of dependency on US foreign aid but rather “Hispanic culture which nurtures authoritarianism, an excessive individualism, mistrust, corruption, and a fatalistic world view, all of which work against political pluralism and economic and social progress. “These are not original observations. Hirschman (1971:275) suggests that the pre-World War I “age of self-incrimination”, self-laceration and pessimism “can be traced back to Simon Bolivar and to his famous statement that in Latin America ‘treaties are pieces of paper, constitutions are books, elections are fights, liberty is anarchy and life a torment,’” The principal weakness of the “self-incrimination” thesis is that writers such as Harrison (1986) view these phenomena as permanent effects rather than correctible symptoms of poverty and lack of education. The characteristics are valid surface descriptions. Foreign aid has tried to eliminate them with varying degrees of success, partly because some policy-makers view them as causes, others as effects. Any foreign aid program that attempts to strengthen part of this contradictory sociopolitical context for its own purposes (and make a profit as well) is likely to end up riddled with contradictions itself. Dissimilar quality organizations that attempt to perform the same task, such as the private United Parcel Service and the public Postal Service in the US, often end up acquiring the same operational features and performance results. US foreign aid practices and policies in behalf of the Third World, tend to “go native” and acquire similarly underdeveloped features, such as dependency, complexity and unintended consequences. US military and economic assistance to Latin America tends to acquire some of these characteristics and merge them with the worst features of US foreign and domestic policy-making. Contradictions emanating from both the recipient culture and the context of American politics tend to produce programs with many dysfunctional results. First, let us examine how US foreign aid has intensified Latin American “dependency”. It can be argued that while economic assistance did not create dependency, the total security-economic package tends to do so by reinforcing structural contradictions within existing political systems. Economic aid generates dependency by such features as “tied aid” loan and grant provisions. But most First World bilateral programs tie aid. Since a given recipient is tied to multiple donors, formally or informally, it cannot be said that US economic aid generates much dependency in the sense of absorbing local capital and investment opportunities.

90

The Politics of United States Foreign Aid

The strengths and weaknesses of the “dependency” perspective were noted in Chapter 2. Nevertheless, the view that “international capitalism had through a long process exploited Latin America and institutionalized poverty by making the continent dependent on foreigncontrolled trade and investment” (LaFaber, 1984:220) was and is taken seriously by many Latin American observers and participants. LaFaber (1984:16) is correct in noting that “it stands as the most important and provocative method of interpreting US-Latin American relations. “Ironically, US foreign aid has been used to both increase and decrease dependency. Because the US is overly-dependent on Bolivian cocaine supplies, the US has tried to leverage a 10% reduction in the coca crop there by suspending all new military and economic aid (Atlanta Constitution, March 4, 1986). The substantial academic weaknesses of the perspective (see Kurth, 1974; Deutsch, 1974; and Odell, 1974) do not weaken the almost religious hold of this view on many groups in Latin America (“the outsiders are to blame for our wretched condition”). With the addition of military aid to the total aid package, dependency proponents have enough evidence in many cases, to validate their claims. A premise of the BRC model is that US foreign aid programs are largely determined by patterns of coalition formation, problem definition and search for solutions in the context of competing bureaucratic actors that conform (or should) to institutional roles. But such critics of this “bureaucratic politics” approach as O’Donnell (1974:169) argue that the conflict is superficial and really over subgoals such as health, housing, and agriculture, etc. More “dominant goals” exist such as “security” and “hemispheric hegemony” (1974:174) which are shared by the major bureaucratic actors despite their appearance of superficial conflict. US policy-making institutions act “incrementally” to achieve dominant goals through normal practices of “satisficing”. The result is a program that is more concerned with dominant US security objectives than Latin American development. The more harmful features of US economic aid derive not from AID plans but rather from the emphasis on their approval in a highly politicized process that ignores the need for implementation in a sensitive environment. AID in particular has difficulty establishing working relationships with local counterparts on a consistent basis. This relates to lack of expert continuity (rotation) and use of consultants with wide differences in capability to deal with local complexities. Given these problems and the availability of large sums of committed funds for recipient development, local institutions seek dependency because it meshes nicely with their own resource and status needs. In an Ecuadorian PID proposal, “tunnelling” under the Ministry of Agriculture was proposed to get around the problem of dependency as a false solution to political instability (the suggestion by some was that aid funding would stabilize the appropriate people in their jobs and permit the program to go forward). In short, the economic aid (DA) programs often “feather the wrong nests” and in so doing increase local institutional dependency on the foreign aid process. Since the policy process de-emphasizes implementation, funding is permitted to generate counterpart dependence which reinforces the existing power structure. For example, the Alliance for Progress tried to force Guatemala into land and tax reforms in the mid-1950s and early 1960s. In both cases, the results could be described as failures. 4000 families received land in “government-owned, undeveloped and inaccessible jungle land” Poverty, illiteracy, superstition and sickness continued to work on these subsistence farmers while 50% of

US Aid to Latin America

91

the country’s best land continued to be farmed by 1100 leading families. Alliance officials continued to believe that “counterrevolution could create a happier Guatemala” (LaFaber, 1984:166). Between 1954–1970, the US pumped more dollars into Guatemala than any other Central American nation. Nevertheless, high population growth kept the aid per capita ratio to the lowest in Central America. But even the elected President Julio Cesar Mendez, tried to meet Alliance demands by passing a “miniscule tax reform” program. The “oligarchy recoiled in such horror that his finance minister who urged the program went into exile overseas” (1984:166). The Alliance worsened the disparity between rich and poor and the percentage of the Guatemalan budget devoted to social services ranked among the lowest of the five Central American nations in 1966 (1984:168). Conversely, the percentage of the budget spent on the military was the highest (1984:168). But any dependency generated by the proposed tax and land reforms paled in developmental significance to the harm created by US security assistance in its various forms. In Guatemala, US security aid often worked at cross purposes to DA in practice by reinforcing repressive structures. In short, dependency on military not economic assistance was inhibiting socioeconomic development. US military assistance via training (IMET), and weapons supplies (FMSC, MAP) is based on the “hard school” (Montgomery, 1986:67) view that political stability is the precondition of successful government. Since real development cannot occur in an atmosphere of lawlessness and violence, obviously, the predominant US emphasis remains on security and stability even while espousing revolutionary aims through such programs as the Alliance. But the primary result is institutional dependency between US and Latin American militaries and little sustained improvement in socioeconomic development directly attributable to the US aid program as a whole. The AID “success stories” (USAID, Congressional Presentation, 1986:629–639) for this region are mainly infrastructure (El Salvador), training (Dominican Republic), education (Guatemala), agriculture (Belize) and nutrition (Haiti). While these are important accomplishments, the aid works to increase the strength of the host regime over the lives of its citizens. In many cases, sustained development depends on the continued existence of a regime that should not be sustained for many reasons. Nevertheless, US enrichment of Latin American military leaders contributes to creation of the “most coherent institution in an increasingly fragmented society” (1984:298). Coherence is consistent with AID routines that get projects approved and open the possibility of long-term change. But, in the short run, the aid reinforces powerfully secretive institutions that rationally view civil (or military) opposition and dissent as violations of the chain of command, and social problems as issues of stamina and discipline. With the power and resources of the national treasury at their disposal, the halls of government (Nicaragua, Guatemala, El Salvador, Uruguay, and Chile all at various times in their histories) become barracks and encampments. More politically experienced (and US-aided) militaries such as Uruguay’s now recognize the no-win rewards of political rule without resolution of law and civil rights issues (McDonald, 1985). But regardless of Congressional restrictions on security assistance through the Foreign Assistance Act (preventing DOD from telling recipients what to buy or what to do with purchases), fiscal support of Latin American military commands distorts local inter-group rivalries and, more importantly, gives the US military excessive influence in determining Latin American public policies.

92

The Politics of United States Foreign Aid

The unintended consequences of a program that needs experience with isolated field development problems but gains military experience instead, are many. In the economic sphere, DA programs do not create harmful dependency but can be associated with many unintended consequences. For example, Currie (1981:132) found that neither deliberate government policy nor foreign advice can claim much credit in the growth of Colombian agricultural productivity and in the massive transition from a predominantly rural and agricultural society to a predominantly urban one with an accompanying decline in the birth rate. For, he found that foreign advisors recommended retention of a high percentage of the total workforce as small farmers. Technicians jettisoned analysis of economic forces in favor of quick fix solutions to increase productivity. This suggests that foreign aid projects often rest on few tested hypotheses and are driven by impatience with both bureaucratic and economic forces. The rules and repertoires of foreign aid, together with goals that are skewed against field socioeconomic development work often produce projects that do not draw on theory, economic history or the “current experience of a large number of countries” (Currie, 1981:134). Development, by this view, often occurs in spite of US foreign aid expertise not because of it! But with sufficient aid resources, accumulated comparative experience, and absence of intervening factors other than poverty, disease, malnutrition, and illiteracy in Latin America (which intervene cumulatively to impede projects designed to eliminate them in piecemeal), one could argue that eventually “subsidies” will turn to “sales” and development will occur. Brazil has often been considered a “graduate” among countries that have received large amounts of US aid (CBO, 1980:4) in that increased foreign aid resources increased the Brazilian rate of investment which stimulated other investments in agriculture, forestry, and social infrastructure. Unfortunately, other things are rarely equal and economic assistance results are strongly influenced by military assistance as well as intervening social variables. One consequence is the rewarding of repressive governments that commit human rights violations with US aid. Schoultz (1981:157) noted that US aid in 1975–76 flowed “disproportionately to the hemisphere’s relatively egregious violators of fundamental human rights”, such as Haiti, Chile, Bolivia, and Guatemala. Since military aid dominates the foreign aid program in countries with military governments, this should be no surprise. But the effects are unintended; aid did not cause the violations but rather legitimized and reinforced them— perhaps increasing productivity! The El Salvadorean military for instance, was far along in the field of repression before US military aid was stepped up in the 1960s (LaFaber, 1984:243). Further, in 1972, Guatemala bought $6. 5 million in military equipment (mostly on credit) for participation in the Alliance’s “Civic Action Program”. This is a Pentagon term for involving the military deeply in economic and social projects. The FY 87 African Civic Action Program is designed, for instance, to “encourage military institutions to participate in economically productive activities”, and to “foster greater understanding between the military and civilian sectors in Africa” (DOD, Volume 2, 1986:28). According to LaFaber (1984:203) : “The Alliance had designed these programs to improve the military’s image by involving soldiers in building schools and roads. Civic action, however, did little for economic balance and democracy. The program turned civilian jobs over to government troops even

US Aid to Latin America

93

as unemployment was rampant. It promoted militarism and gave the generals an increased voice in economic and social policies. And it publicly allied the US with the Central American military at the grass roots level, even as revolutions picked up momentum in the towns and villages. Other side-effects appeared. The number-three man in the Salvadorean military was convicted in New York City during 1976 for attempting to sell 10, 000 machine guns—apparently Salvadoren surplus—to East Coast mobsters.” Ironically, the domination of the economic aid program by the security assistance program stimulated the very revolutionary activity it was designed to restrain. In Nicaragua, the US created Somoza’s National Guard and sent more officers to the US School of the Americas in the Canal Zone than any other nation in the hemisphere (2969 between 1949–1964). This helped establish the Nicaraguan military oligarchy that later received economic aid from the Alliance for Progress. Much of the Alliance’s “new cold war” money of the 1970s directly enriched the Somoza family and the Guard officers (LaFaber, 1984:226). The Alliance provided $50 million in loans between 1961–1967 primarily for economic “diversification” in the same way that the US is currently providing $10.5 million in DA and ESF to Belize for diversification from sugar (DOD, Volume 2, 1986:10). But the Nicaraguan oligarchy conveniently interpreted this to mean cotton instead of coffee (LaFaber, 1984:162). As cotton boomed, many small farmers were evicted from their grain-producing lands. Nicaragua turned to net import status in basic grains, the economy ground to a halt in 1967, and unemployment skyrocketed. That same year, the National Sandinista Liberation Front (FSLN) “began operations” (1984:163). Similarly, the 1960s Alliance for Progress used revolutionary rhetoric (“Let us again transform the hemisphere into a vast crucible of revolutionary ideas and efforts”—JFK cited by LaFaber, 1984:148) but spent money to train armies and police to prevent revolutions, which ultimately led to the revolutions of the late 1970s and early 1980s! For example, in 1968 the US-fed Guatemalan army “developed interests that diverged from the Alliance’s programs”, such as becoming major landholders. The army aligned with “more reactionary economic factions”, in an almost classic Latin American pattern, and vetoed the Alliance’s tax reform program by firing the finance minister. Just as conservatives could not decide whether the minister’s tax proposals “originated in Washington or Moscow”, the US ended up by being “hated with equal intensity by both sides” (1984:171). US military aid led directly to involvement with “unelected militaristic governments” (1984:153) and therefore the repression and killing of their opponents. In 1963, the military government of Guatemala “told the US to provide supplies and training but no advice, as its troops killed thousands of peasants and political opponents in the name of ‘counterinsurgency’” (1984:257). Domination of the foreign aid program by military assistance also led to the paradox of harmful dependence but lack of US leverage over recipient activities. For instance, following the army slaughter of more than 90 Salvadorean people in 1977, President Carter wanted to cut off military aid. But it had been cut off in 1976 when US authorities, as noted, “caught Salvadorean officers selling surplus weapons to North American gangsters.” General Carlos Humberto “then announced that he did not want Washington’s military goods anyway.. He turned to West Europeans and Israeli arms merchants” (1984:246).

94

The Politics of United States Foreign Aid

Similarly, military aid can unexpectedly benefit the opposition, as the US discovered in Vietnam (1975), and Nicaragua (1978). Despite massive amounts of US military aid to El Salvador ($1.96 billion in 1984 compared to only $22.6 million in 1981; USAID, 1984:47) guerrillas expanded their control over the country. According to LaFaber (1984:313), “about 40% of the arms Washington sent to the Salvadorean army ended up with the revolutionaries through sale or capture. The US taxpayer actually gave more aid to the rebels on a per capita basis than to the Salvadorean military. The folly of reliance on military aid to achieve social objectives should have been learned from the experience with Fidel Castro. From 1949–1961, the US provided $16 million in military aid to Cuba. This was 80% of all aid furnished to Cuba in that period (USAID, 1984:44)! By 1956, the US controlled 80% of Cuba’s utilities, 90% of its mines, and cattle ranches, nearly all of its oil and 40% of its sugar (Spanier, 1983:150). But the US-supported Batista dictatorship could neither control nor serve the majority of Cuban people. Transacted across large, formal organizations, the aid relationship intensifies counterpart dependency. The task environment of planning and budgeting aid amidst distrustful, powerseeking actors, tends to draw the US into an almost permanent relationship with its once-pliant wards. Under these conditions, leverage for any purpose is no longer possible. In El Salvador, US officials were “too deeply involved to demand very much” (LaFaber, 1984:255). By 1982, increased US investment in Guatemala led to the search for quick wealth and then to more slaughter. Functionally-oriented “mafias”, e. g. the military, bankers, importers, and industry, engaged in warfare for power and plunder. The US was faced with the dilemma of extending military and economic aid “so the army could fight the growing revolution, but threaten to cut off aid if the rival mafias did not stop murdering Indians, labor leaders, educators, lawyers, and each other” (1984:261). As LaFaber put it so well: “much as one hesitates to stop feeding a pet boa constrictor” the US was “reluctant to cut off aid and face the consequences” (1984:260). But the addition of US trade and investment in Latin America has historically and presently made it difficult to use aid as leverage. Contrary to the simplistic radical notion that aid, trade and investment are mutually consistent policies, they are in frequent conflict and not over just “subgoals” (O’Donnell in Cotler and Fagen, 1974:173). Feinberg (1983:39) notes that the efforts of the Carter Administration to improve human rights performance in Argentina, Chile and Uruguay “were undermined by private international capital markets.” He noted that “The small amounts of bilateral economic or military aid that were withheld paled in comparison to the hundreds of millions of dollars each of these governments was borrowing from private creditors.” As LDCs develop economically, if not democratically, their ability to gain access to “harder” loans puts them out of the “soft” loan and grant program of USAID. Conversely, private bankers and firms withdrew lines of credit and other liquid assets from El Salvador (1979–80) which harmed US efforts to “rebuild the country’s tottering political stability” (1983:39) via injections of aid. These were hardly intended consequences of aid planners. Reinforcing dependency and unintended consequences is the aidprocess itself which produces a state of the art but intensely complex product. Projects and programs are rife with excessive specialization and hierarchical relationships that encourage formalism, lack of communication, fragmentation and frustration on the part of they who want the projects to succeed. Rules and repertoires employed by actors in conflict over planning

US Aid to Latin America

95

and budgeting resources tend to encourage a standardized approach to project design and implementaion. Though this wastes administrative resources, it nevertheless enables USAID to communicate formally and consistently with recipient counterparts. But the price paid for this formality is inflexibility. Course corrections are much more difficult to make where, for example, evidence exists that aid is exceeding local absorptive capacity or that surplus is being siphoned off for the wrong purposes. Hence, in addition to dependency and unintended consequences, US aid programs tend to replicate the complexity of the process that created them in the field. This is due in part to the rules which favor foreign exchange cost financing only, a relatively large in-country AID presence, and the strong security emphasis to foreign aid. The first and third points have been discussed. Extensive statutory requirements necessitate staff to plan according to them. The financing limitation is based on the notion that capital development preceeds socioeconomic development. The AID decision structure also encourages project complexity. Tendler (1975:13) argues that the novelty of the task and decentralized mission structures contributes to a less hierarchical character than in other federal agencies. While this is true, the task novelty also encourages a quest for rule certainty, evident in project planning. The projects are traditional and display an almost machine-like conformity regardless of the rhetoric surrounding them. The result of a decentralized line seeking certainty from the rest of the aid policy bureaucracy is goal displacement, fragmented decision-making and field coordination problems that lower overall effectiveness of aid projects. Of AID’s 3748 people 59% or 2212 are in AID/W, and 41% or 1536 spread around 61 world-wide missions. With the exception of West Germany which has 45.8% or 1415 aid personnel attached to its embassies, the US has more in-country presence than any major donor (Canada, France, Sweden, UK). But the average cost of maintaining each direct-hire American employee overseas is about $100, 000 per year. Operating expenses for field missions is nearly double that in Washington ($231 million versus $124. 6 million FY 79–83; GA0, 1983:13). More importantly, administrative costs for US aid are clearly the highest of any donor: $320.5 million in 1980 compared with $38.9 million by Canada, $109 million by France, $39.2 million by Sweden, $69.9 million by the UK & $49.2 million by West Germany. Similarly, the ratio of aid delivered to costs incurred is the lowest of all, 14:1 as opposed to 47:1 by West Germany which also maintains a large field presence (GA0, 1983:13). The expenses vary, of course, by types of aid. Where DA and PL 480 is emphasized, costs are higher because of rules and repertoires which encourage slow disbursement with staff-intensive programming. Conversely, ESF programs are less staff intensive and amount to either direct payments (budgetary support) or capital project execution (GA0, 1983:14). Excessive reporting requirements lead to over-organization and control emphases in programming. These same staff levels must then monitor implementation. Given the volatility of the Third World environment in which projects are to be implemented, the result is either rule gridlock or complete staff indifference to results. AID’s congressionallyimposed rules tax its staff beyond their absorptive capacity and lead to projects that only superficially attain their objectives. The “pipeline” keeps the money flowing to questionable recipients for the wrong types of projects (large scale, capital intensive) and this produces, especially in the military

96

The Politics of United States Foreign Aid

sphere, harmful results as well as a tightening circle of dependency. The pull of military foreign assistance is so strong that President Reagan has proposed a 12% increase for FY 87 in the face of growing opposition to his non-defense cuts, ostensibly to satisfy the Gramm-Rudman deficit reduction targets.

Goals We have noted the overtly political character of US aid allocations. Any attempt to correlate economic aid receipts and recipient per capita income, for example, is inhibited by introduction of the security assistance component of aid (Asher, 1961:65). The history of US foreign aid is replete with examples of extra-project funds simply “because the US has decided to support a given country” (Stallings cited in Fagen, 1979:226). Assistance, by ESF in most cases, is for security-defense reasons including investment protection, ostensibly because the host government requested aid from the US. Nevertheless, in the 1960s, the US applied openly political criteria” (Stallings, 1979:226) to assist such countries as Chile and Brazil with (1) military aid to secure the cone region from internal subversion, (2) tied aid loans to provide business for US exporters, and (3) development assistance to improve the lot of the poor and reduce instability. The BRC model goals (security, profitability, and welfare) are not the only goals pursued by the US in Latin America. Schoultz (1981:150) suggests that such political purposes or “nondevelopmental” goals have included: “aid to help allies win elections (Chile, 1964), to consolidate military power after regime seizure (Brazil after 1964) and to survive crises which threaten continued exercise of military power (Dominican Republic after 1965; Bolivia after 1971).” He also notes that “aid has been used to purchase OAS and UN votes, to secure military base rights, and to obtain the help of foreign troops.” More recently, the US used the threat of withholding aid to demand an apology for an anti-US speech by Zimbabwe (Alexander, Atlanta Constitution, July 11, 1986). However, it would seem that such diverse goals can be telescoped into security, profitability and welfare without distorting or simplifying what US aid is actually used for. The goal of welfare or developmental need has been least emphasized in Latin America. This can be criticized as callous, and predictably capitalist or applauded as prudent savings that feathers the wrong oligarchic nests. It can also be criticized for turning the Latin poor into US welfare doles. Nevertheless, historically, the US provided rhetoric rather than aid resources to Latin America for social development. For their wartime cooperation in selling cheap to the US, Latin Americans hoped they would receive aid after World War II. But the US “assumed” political stability (dictatorially or democratically-enforced) and provided “little or no economic aid after the war” (LaFaber, 1984:89). Again, Latin America hoped that in return for military and economic cooperation in signing the Rio Pact of 1948 they would receive economic assistance. “No significant aid appeared” (1984:93). By 1952, State Department Latin American expert Thomas Mann admitted that “Our program of economic aid to Latin America is so small that it could almost be financed by ExportImport Bank profits from loans to Latin America alone” (1984:106). For FY 86, USAID proposed $1.2 billion in economic aid for the Latin American/ Caribbean Region. This aggregate figure was $190 million less than for FY 85. Of the $1.2 billion, $833 million or 64% was proposed for ESF projects (USAID, FY 86

US Aid to Latin America

97

Congressional Presentation, Volume 1, p. 23) and only $460 million or 35% for DA projects in the functional areas of: agriculture/rural development and nutrition; population planning; health; education and human resources; and selected development activities (USAID, FY 86 CP, p. 12). The “selected development activities” subprogram ($107 million or 8. 2% of the total request) consisted of a loan to the Caribbean Financial Services Corporation to finance new private sector enterprises and expand existing ones (USAID, FY 86 CP, p. 24). Put another way, International Security Assistance to Latin America accounts for 69% of the total request for foreign aid. Security assistance is dominant; private sector profitability a close second and welfare and development a distant third objective. As evident in Figure 4.1, most of US aid to Latin America flows now to El Salvador and Honduras. The rhetoric of developmental activities has changed over the years. Marshall Plan capital growth emphases were democratic and social development objectives in the Alliance years, which turned to security development by the 1970s. Failures along these fronts produced new rhetoric on basic needs and human rights for the mid-1970s. Despite remaining emphases on women in development, the current rhetorical emphasis is on development by private sector growth, roughly the Marshall Plan revisited but for the absence of “capital push”. But project activities to achieve development have remained largely constant. The guiding notion is that training, technical assistance or transfer of US know how (often hidden in “train the trainers” language) to develop agriculture, health, education and natural resources will move Latin American societies from traditional to modern western allied status. So the tools have remained the same, only the names have been changed to protect the guilty, e. g. farmers are now rural entrepreneurs (Newfarmer, 1983:118). Put simply, the foreign aid machinery projected the US view of how it developed (which was itself changed by the ideological perspectives of the administration in power) onto Latin America. For instance, we were dealing with “reasonable” governments like our own. Therefore any unreasonable governments that wanted leftist or totalitarian solutions rather than gradualist US democratic capitalist evolution, must somehow be influenced by “outside agitators” (Guess, 1984). The outside agitator premise allowed US security assistance to hold the line against communism (Containment and Frontiersmanship as noted in Chapter 1). Burns notes (1985:152) how Preident LBJ treated Third World leaders like US senators. “He presumed they were all reasonable men who could be persuaded to compromise on almost any issue if the right combination of threats and incentives were employed. “Not that LBJ was the typical American leader, but the decision premise is shared by many foreign aid policy-makers. Most aid programs depend on the recipient and US private sectors as well as government sectors for development. Their frequent mutuality of interest, as noted, often excluded the majority. Whether this “trickle-down” approach was correct depended on the differing contexts provided by the diversity of Latin countries. However, the instability and political chaos of most Latin societies rarely gave such long-term traditional approaches a chance to succeed. As early as the Bogota Economic Charter of 1948, the US was insisting that “private enterprise” be the “means for developing the southern hemisphere, while the Latins wanted to utilize more state power” (LaFaber, 1984:97). But the Latin private sectors were weak, oligarchic and tied inevitably to their armies for protection. The US private sector

Figure 4.1: Fiscal Year 1986 Request For Economic and Military Assistance Programs (Scaled To Program Size) Source: Defence Security Assistance Agency, US Department of Defense, “Congressional Presentation For Security Assistance Programs,” Volume 1, Fiscal Year 1987, p. 16

US Aid to Latin America

99

could invest, co-finance, and enter joint partnerships, while beneficiaries would be few. Conversely, the goal of development was blocked when, circumventing the private sector, reformers found Latin governments to be more “kleptocratic” than welfare-minded. As Kennedy’s “New Frontiersmen” learned too late about the Alliance program, “Under the governmental systems in such areas as Central America, widespread meaningful reform was impossible, no matter how much money was thrown at the program” (LaFaber, 1984:142). Unable to stimulate reform policies by recipients without charges of neo-imperialism; unable to link aid funding to democratic evolution in practice; unable usually to get warring factions to talk to each other but fearful of outside agitators gaining too much influence, the US usually responded in frustration by turning on all the aid valves which overwhelmed recipient absorptive capacity. The key to success (or major failure avoidance) has been the US ability to combine security-profitability for developmental objectives across the aid policy machinery. This requires a good crisis, clearly delineated problems, and aggressive actors to break down the weight of traditional rules, repertoires, and absence of realistic communication. US El Salvadorean aid policy has been more successful (“helping democratic Salvadorean government defend itself against antidemocratic insurgency;support consolidation of constitutional democratic process; and maintain production, employment and support adjustments to encourage growth”, DOD, 1986, CP, Volume 2, p. 46) in the last three years (1983–86) in creating a centrist government with increasing control over the countryside primarily because its enormous dollar impact has provided leverage for the US. In contrast with many recipient cases such as Vietnam, the US has exercised the leverage against the military and the government almost without regard to criticism, in part, from fear of the consequences of failure for the rest of the region. Throughout the history of US-Latin American relations, the only consistent aid objective has been to support the military. This is true despite the fact that military assistance to Latin America has been low relative to other regions. In FY 84, US military aid as a percentage of total aid to Latin America was only 22. 7%, while in the Near East it was 60. 6% and East Asia 58.8% (USAID, 1984). As indicated in Figure 4.1, most of the security assistance is concentrated in Central America. Though the bulk of foreign aid is ESF, DA and PL 480, more marginal injections of military support tend to have a catalytic effect on local politics. In 1969, the Rockefeller Report noted 17 military coups since the Alliance for Progress began eight years earlier. Based on this report, President Nixon concluded that the problem was law and order (“hard school” logic) and recommended more military aid. Like many others in the foreign aid policy process, the President believed the military to be “the essential force of constructive social change” (LaFaber, 1984:202). This fits nicely with the Nixon Doctrine of 1969 that delegated the US policeman role to local allied security forces, such as the Brazilian generals and Central American military regimes, and fed them with security assistance. Unfortunately, as in many foreign aid cases, the US policy-makers may have had little choice. According to LaFaber (1984:269) US officials were often trapped by history and circumstance. “They lacked both the imagination and perspective to understand how they less resembled the defenders of a free world than those who defended the late Roman Empire or the British Empire in Egypt. “He might have added that even with imagination and perspective, policy-makers would be constrained by ill-defined institutional roles and

100

The Politics of United States Foreign Aid

dysfunctional rules. While everyone agrees that the US pushes “hemispheric hegemony” (O’Donnell, cited in Cotler and Fagen, 1974:174) and security leverage as its dominant foreign aid goals, given the common propensity of US investment and banking to work at cross purposes (security versus profitability) and increasing domestic criticism of military solutions, a closer look at the origins and alternatives to the security goal in Latin America is warranted. As early as 1954, President Eisenhower admitted the weaknesses of a militarydominated foreign aid policy. In a 1954 report, the Administration noted the core dilemma: Latin America required stability to develop. But underdevelopment and poverty produced military coups and instability. On the other hand, revolutionary rhetoric accompanying the traditional foreign aid package, such as the Alliance, fueled expectations and added to the instability brought about by poverty—now “conscious” poverty, a force for active social change. Extremist factions on both left and right often played to these expectations while directly attempting to blackmail and weaken any positive effects US development aid might achieve. This presented US policy-makers with the paradox of requiring security aid to make economic aid work, knowing that integration of the two was extremely difficult. As could be predicted, the elusive quest for stability via security aid in societies noted for their particular volatility proved to be a long-term contradiction. Whether indigenous turmoil or generated by Soviet proxy aid (the Brzezinski and now Reagan East-West framework for judging Latin American problems to be caused by “outside agitation”), the aid response was quite consistent-security assistance via some combination of FMSC, MAP and ESF. Programs designed to produce development amidst terrorist attacks, such as the AID “Policy Improvement Program Assistance” project in Peru turned out to be more of the same: balance of payment and budgetary support to “cushion the disruptions caused by terrorism” (Sendero Luminoso). The premise of the ESF program in Peru ($45 million loan) was and is that “shortage of counterpart funds slowed the implementation of many critical investment projects” (USAID, FY 86 CP, Volume 1, p. 204). The premise may indeed be valid. If so, the question is why the last $1.75 billion in US aid (1962–84) to Peru did not address these problems? Though the US provided only $10 million in security assistance to Peru over the same period (USAID, 1984) it is clear that they receive other sources of security aid, such as French fighter aircraft, and that a substantial portion of the Peruvian budget is still devoted to military expenditures (15% in 1977 according to IMF, 1982:31). Though every society has law and order problems that need to be addressed by well-trained police and security forces, to bolster security as the major vehicle for national development is merely to attack the short-term effects of deeper underlying problems. Why the US continues to be led into this conceptual trap is not a problem of individual policy-maker imagination or perspective, as LaFaber suggested (1984:269), but one of institutional paralysis around often inappropriate program priorities and project means. But here it must be noted that the security premise runs deep among both civilian and military policy-makers. More recently, the US military has been more cautious about its “theaters” from fear of losing, or at least not winning, a war and jeopardizing their enormously successful defense “spend up” ($300 billion per year without a “real” buildup in readiness, according to Fallows, 1986:27). They have cautioned US involvement in potentially damaging conflicts, meaning greater reliance on locals via FMSC, and MAP,

US Aid to Latin America

101

or the Nixon Doctrine revisited. A good example of the running security premise is the Kissinger Report of 1984 which noted that while poverty invited revolution, “hostile outside forces—Cuba backed by the Soviet Union and now working through Nicaragua” (LaFaber, 1984:308) were destabilizing Central America. The Report therefore recommended increased arms aid to Honduras, Guatemala and especially El Salvador (see Figure 4.1). This represents the “outside agitator” premise noted previously which tends to create selffulfilling prophecies as outsiders provide aid to opponents of repressive “ward” regimes of the US. Rather than push for regime structural change via linkages of human rights to military aid (which President Reagan opposes as would the recipient regime), the US almost tragically provides small doses of traditional economic aid (DA, PL 480, and ESF) which stabilizes the operating environment for usually insensitive regimes propped up by US military aid (FMSC, IMET and MAP). The regime then conducts opponent “witch hunts” without US approval but with color of authority provided by past US aid. Among some policy-makers, the security premise is acute. For example, one cautious Pentagon opponent of US military involvement in Central America, Rear Admiral (ret) Eugene Carroll, Jr., noted that “Guerrilla forces in El Salvador, once their command posts and major concentrations are overwhelmed, will yield easily. Latin American troops don’t have their heart in any kind of war where there’s a lot of killing: that’s their history. They’re not like Asiatics” (Greve, Miami Herald, August 3, 1983). Hence underlying many US policy-maker decisions is the notion that Latin American development results from “winnable” guerrilla wars that leave the populace supportive of the US and its regime, in a society washed clean of outside agitators (meaning most regime opponents)! More often than not, the results of this kind of aid policy are pathetic and wasteful, but not without levity. Montgomery (1986:58) recalls how the late Wesley Fishel of the Michigan State advisory party to Vietnam “would sit slightly out of Diem’s view (Somoza?) mocking his gestures and making hilarious pantomine faces at his descriptions of the high purpose of his regime and his expression of concern for the welfare of his countrymen. “Most of the US aid program in Latin America, if properly edited, could read like a script for a Woody Allen or Monty Python film. In 1960, for example, the US-supported “clown” in Guatemala was Ydigoras Fuentes. The US sent Cubans there to train for the later Bay of Pigs fiasco. But “wealthy and reckless”, the early “freedom fighters” lived it up beforehand, resulting in many deaths by car accident on Guatemalan highways. This conscious flaunting of wealth upset the Guatemalan army which then revolted against Ydigoras. To fight his own army, Ydigoras filled up a transport plane with the Cubans (LaFaber, 1984:165)! Only intercession by the US ambassador this time prevented the use of US military aid to support both sides of a war. To round out the story, the air force revolted in 1962 anyway and was soundly thrashed by the Guatemalan army. Amidst this macho-flexing, the populace eagerly await their chance to develop democratically with US aid. But despite these abuses, suppose the US had not provided military assistance to Latin America? Would all of Latin America resemble the serenity of Costa Rica or perhaps Venezuela? Probably not. Five points need to be made which suggest that foreign aid is not an either military/or economic aid proposition but rather, for Latin America, a matter of stimulating the development of democratically authoritarian structures with wise injections of both. First, US military aid did not create the extensive armed forces in, for example, Chile or Uruguay. The US periodically contributed assistance, much like throwing hats

102

The Politics of United States Foreign Aid

into a flood to try and guide its direction. Second, historically unique military situations exist in selected Latin American countries that may not be transferable or interchangeable by aid. The modern anti-military Costa Rican political culture has a healthy suspicion of military “heroes” and views its developmental progress as independent of military fortune. In Venezuela, the democratic AD (Accion Democratica) Party gained office twice after the military overthrew dictatorships in 1945 and 1958. In Peru, the military overthrew President Belaunde Terry in 1968 and changed the political rules of the game to favor revolutionary or at least progressive programs. These unique and often favorable military relationships to civilian political institutions were not and probably cannot be created by aid programs. Third, at frequent intervals, Latin American polities tend to jettison the political rules of the game and take to the street, installing populist but autocratic replacements for the slow-moving machinery of democracy. Rarely do opponents stick to the electoral rules of the game where the easy option is to ally with the military for a coup. This applies whether the coup is from the right (Pinochet’s Chile), center (AD Venezuela) or left (Frente Amplio in Uruguay tried to make a deal with the military before the 1973 coup there). This makes the stabilizing effect of military aid difficult to predict. But fourth, without military aid (cutoffs), the Latin American military can easily obtain arms from rival suppliers, often from other Latin American countries with their own manufacturing capacities, such as Brazil. The Sandinistas, cutoff from all US aid by mid-1981, received assistance from Western Europe, Mexico, the Soviet bloc, and Qaddafi’s Libya (LaFaber, 1984:295). Hence, the ultimate leverage gained by marginal injections of US military aid in most cases is very little. Conversely, the greater the aid level, the higher the leverage (El Salvador) though the results may not be favorable (Guatemala). Nevertheless, the fifth point is that US military aid can, if properly directed and monitored, ensure serious negotiations among warring factions. For example, without military assistance to the factions favoring centrist President Duarte of El Salvador, either the “killer right” of Roberto D’Aubuisson (Diskin, Miami Herald, March 18, 1983) or the chaotic coalition of leftists (FMLN) would have gained control, with all such extremist programs have brought for Latin American development in the past (General Velasco Alvarado’s Peru, Allende’s Chile, Pinochet’s Chile, Castro’s Cuba, and Sandinista Nicaragua). That is, the easy solution of eliminating US military aid and replacing it with economic aid ignores the reality of sabotage by extremists who view aid programs as political competition, insoluble by reasonable and peaceful means. Here historic partisan conflict between Republican military aid advocates and Democratic economic aid advocates becomes moot. Both may be needed and the issue is how and in what proportions? The view that military aid may be needed to offer incentives to both left and right in El Salvador (the best current regional example of US aid in a political muddle) is held by many outside the Pentagon, including Representative Clarence Long (D-MD), former chair of the House Appropriations Subcommittee on Foreign Operations (Miami Herald, May 1, 1983). Some have held that the developmental results of US aid may be improved by requiring progress reports to Congress on Salvadorean control of rightist death squads (Congressional Quarterly, July 20, 1985, p. 1424). Control can permit targeting military aid and avoids the more glaring fungibility problems associated with blanket military grants in a country where the military is a miniature left-right political spectrum and aid merely

US Aid to Latin America

103

intensifies the stakes of the power struggle. Others suggest the “hard school” view that the “prerequisite for economic growth and greater democracy is security of life and property” or that Central America needs both long-term economic aid and short-term military aid to “provide a shield behind which economic growth and democratic change can begin to take root” (Kemp, Miami Herald, August 1, 1983). Control over military aid execution in other than a narrow accounting perspective is required. But controllers themselves are often accused of paternalism and interference by both Congress and the recipient military. In final analysis, military and economic programs are not contradictory concepts in theory. They are complementary but the overwhelming size of the military program in many countries (see Figure 4.1) dictates overall strategy and often makes them contradictory in practice. The military is viewed as the strongest institution in Latin America; building it up increases chances of a victory. Conversely, withholding aid to the military can affect political development. For example, the suspension of $26. 6 million in economic and military aid to Haiti for human rights violations is credited with persuading Baby Doc Duvalier to finally give up power (Atlanta Constitution, February 27, 1986). Note that it was the sum which provided the leverage, not any presumed effects on development that influenced behavior. To make matters worse, the fiction of profitable aid needs to be maintained for Congress and a generally uninformed US public (the issue remains unimportant because few powerful interests stand to lose income from the absence of foreign aid other than in the realms of military assistance and perhaps food aid). Foreign Assistance Act Sections 612 (b) and 638 (h) require projects to assure that “to the maximum extent possible, the country is contributing local currencies to meet the costs of contractual and other services, and foreign currencies owned by the US are utilized instead of dollars” (GAO, 1983:84). This criteria is applicable to all economic aid programs. But in practice the fiction is maintained by disbursing loans for soft matters like health and education, while providing grants for “real needs” like arms and military training. In 1984, 50% of all economic assistance to Latin America was in the form of soft loans; from 1962–84, 61% of all economic assistance to Latin America was by loan; from 1962–84, 52% of all aid to Latin America was via loan (USAID, 1984:33). “Cost-sharing” provisions in AID projects (25% by the recipient for program, project or activity) can be waived only if the country is “relatively least developed” according to Foreign Assistance Act Section 110 (a) (GAO, 1983:85). That loans are only fictionally profitable is revealed by the fact that of all economic aid loans, Latin American countries repaid only 22% of the principal and interest from 1946–1984 (USAID, 1984:33).

Constraints General problems with rules that affect capacity to program, budget, execute and evaluate foreign aid programs were noted in the last chapter. Here, let us focus on specific statutory requirements and practices that constrain effectiveness over foreign aid in Latin America. It can be argued that the entire programming-implementation cycle encourages either overattention to either bureaucratic detail or complete indifference. According to GAO (1983:58), “The AID Administrator, in a recent communication to the field offices, acknowledged the criticism that, in the past, the program design process had become

104

The Politics of United States Foreign Aid

an end in itself rather than a tool with which to achieve developmental objectives.” Put another way, “projects are judged on criteria unrealistic in terms of implementation and are approved as long as they are well-articulated and presented in the proper ‘form’” (1983:58). The programming phase of project development consists of a lengthy “staff intensive” exercise of addressing specific issues mandated by congressional statutes on foreign aid over the last 20 years. In most cases, the requirements are excessive and extensive staff time is devoted to rationalizing how conformity can be achieved. For example, most Latin American countries have taken steps to prevent locally-produced narcotics from making it to the US market (Foreign Assistance Act Section 481). But how adequate have these steps been if the street value of cocaine continues to drop (40% less today than in 1984) in the US (Lieber, 1986:40)? As noted, Section 620 (s) of the Foreign assistance Act requires the Administrator to “take into account” the amount of foreign exchange or other resources which the country has spent on military equipment. Does this mean less economic aid if more budgetary resources are devoted to military purchases by the recipient? Though we have discussed the State Department’s SAPRWG process of coordinating security aid into an “integrated request” which includes economic aid, no precise criteria or weights for fixing proportions have been provided other than the commonsense notion of “consensus”. The more important but unexamined question is how that consensus is attained. Further, for capital project assistance over $1 million (Foreign Assistance Act Section 611e), the Mission Director and Regional assistant Directors are supposed to “take into consideration the country’s capability to maintain and utilize the project. “This, of course, related to the recipient’s absorptive capacity. But it should be evident that if such criteria were actually taken into account, both economic and military aid would stop to many Latin countries, especially in Central America. How effectively can a country the size of El Salvador spend the requested $240 million in ESF and $134 million in MAP during FY 87 (USAID, 1986:26)? GAO concluded that “neither El Salvador nor the other small nations in the region have the capacity to absorb large amounts of foreign aid” and “..aid for several nations is already getting backed up in the pipeline” (Chardy, Miami Herald, February 25, 1984). Despite the immensity of bureaucratic constraint on viable AID programming, many ESF funds flow in as “cash transfers” and mysteriously disappear. Foreign Assistance Act 121 (d) requires a determination that Sahel projects have an “adequate system for accounting for and controlling receipts and expenditure of project funds” (GAO, 1983:84). But the capability of any system to function depends on the size of resources it must process. In Central America, the task is clearly excessive. Other items that must be programmed into the project include: environmental and tropical forestry issues (Foreign Assistance Act 121 (d), “New Directions” criteria such as appropriate/labor-using technology, democratic local government institutions, women, participation, cooperatives, loan repayment capacity, as well as proof that the US will not be involved in the purchase of agricultural lands. Though the Reagan Administration is clearly less interested in New Directions criteria, many of its considerations remain as subjects for issue papers and debate between Missions, AID/W and other foreign aid actors. For example, AID programmers must prepare mini-EIS’s for each project to assure every interest that the project could not possibly destroy the environment and tropical forests. But almost every agricultural development project will affect the environment by increasing

US Aid to Latin America

105

“point sources” to pollute water, or eliminate tropical forests for cropping lands, and so on. Most road projects will negatively affect both. Often AID prefers computer models, such as the Resources For Awareness of Population Impacts on Development (RAPID) developed predictably by a consultant, to tell them what most AID people already know: natural recovery of land, water, and forest resources is decreasing due to population pressures (Nesmith, 1985). Simply interviewing farmers and peasants near a proposed project would produce the same conclusion and probably some good suggestions on how to deal with them—at much less cost! But project planners must invest staff time on this because it is federal law. Similarly, the likelihood of local governmental interests being developed by aid to Latin America (which are strongly unitary societies with no local government taxation power), or women changing their roles through education and training is not very high in the short run. Requiring analysis of these and other issues like them to please congressional mandates wastes staff time in planning better projects. The problem with such statutory “checklists” is that once “checked off” they become written in granite by Congress for the managers in the AID Missions and AID/W, who have a difficult time obtaining clearance to deviate from them for misjudgments in project design. The design process is structured according to extraneous considerations which require posturing before Congress to demonstrate that all is foreseen (even though lead time between project identification and implementation can be up to 3 years) (GAO, 1983:55)! Much of the data required to satisfy these requirements is simply unknown and resources are used up to “cook” impressive information for the programming and congressional presentation process. This is an enormous constraint on better use of foreign aid resources—even moreso when it is recognized that of 3748 AID people in 1983 (GAO, 1983:10), 2212 (59%) were in Washington overseeing and ensuring compliance by the other 41% (1415) in the field! It should also be remembered that USAID is not the only actor programming foreign aid. State, DOD, Congress and the President also make decisions which affect aid programs and render coordination difficult. For example, in crisis situations such as El Salvador, different supplemental appropriation requests flow from the President ($53 million after 13 people including 4 US Marines were killed recently); DOD ($482 million for the same reason; overruled by the White House), and State ($1.2 billion in DA/ESF/PL 480 aid to cover the level recommended by the Kissinger Commission) (Congressional Quarterly, August 24, 1985, p. 1849). Similarly, a 1985 House Intelligence Committee conference report provided for administration of $27 million in Nicaraguan contra aid by State instead of CIA which limited to information only. But State could include trucks as part of humanitarian aid, which can only increase future fungibility problems (CQ, November 11, 1985, p. 2451). The CIA is supposed to seek reprogramming and transfer authority from the House Intelligence Committee whenever it wants to aid contras “directly”! Finally, it should be noted that AID Missions also have considerable discretion to “program” projects to achieve ends that may not be consistent with vigorous and innovative implementation of foreign aid projects. Recently, for example, the Belize AID Mission (with an $11.4 million portfolio for FY 86 of which $6.5 million or 57% is DA) required contractors bidding for the “Training For Employment and Productivity Project” (86–007) to eliminate their candidates for “chiefs of party” and other lead personnel and to find candidates with more US foreign aid project management experience. This strongly suggests that AID is seeking people who speak its

106

The Politics of United States Foreign Aid

language, such as AID retirees or near-retirees. Such attempts to program the appearance of aid projects accounts for the lackluster uniformity of the product and often the less than spectacular results both during execution and after the funds have been spent. As indicated in the last chapter, the AID budget process tends to drive the rest of the programming and execution cycle. Deadlines and rules ensure this result while delays and deviations threaten the apparent efficiency of resource flows. It is often said that budgeting reflects public policy. But policy should not be made by the budget process alone because it is normally a narrow and blunt instrument that makes tough though not necessarily wise choices. The ABS is geared toward the congressional appropriations process, stressing financial information with little or no narrative or explanation. As noted in the last chapter, the ABS was originally intended to have a programmatic as well as budgetary role (GA0, 1985:17). Though programming issues are raised, the reviews emphasize issues of budgetary control, rather than managerial, planning or implementation questions. In general, projects are not deleted or modified from the aid portfolio as the result of ABS analyses. With emphasis on straight financial issues, such as budget authority-obligation balances, it is not hard to understand how this process recycles mistakes based on faulty policy premises. For example, how locked in should aid priorities be in El Salvador where US aid represents 25%–30% of the Salvadorean government budget and budget politics notoriously takes on a life of its own? Additionally, the US aid effort in Haiti since 1973 has stressed PL 480 and agricultural development projects (DA). While road maintenance and construction projects assisted by AID have been “most successful,” the overall AID program has had little impact on Haiti’s “dire poverty” because of implementation problems. For example, an “Integrated Agricultural Development Project” for $12 million was initiated in 1976 to develop the institutional capacity of the Ministry of Agriculture and community organizations to deliver productive resources and services to small farmers. By 1981, only about $2 million had been spent (GA0, 1982:5–7). Because of AID emphasis on increasing local institutional capacity to spend pipeline funds, “absorptive capacity” problems impeded effective results in Haiti. In 1975–77, money obligated for AID projects but not spent increased to $38 million (GA0, 1982:16). The foreign aid program must still go through an annual reauthorization by Congress during which AID develops its internal budget with State Department, taking into account security and development issues. Since the executive side budget process tends to squeeze policy questions out, the appropriate time for raising larger questions of program effectiveness should be during reauthorization. But, as noted, since Congress has not been able to complete reauthorizations in recent fiscal years, emphasis has shifted to the appropriations process. Though not technically legal (authorizations must be a separate prior action), the result has been integration of financial and larger policy issues in one abbreviated process. After reauthorization, Congress then appropriates funds (provides budget authority) by sector or functional account. The 1973 New Directions legislation established categories for activities such as: agriculture and nutrition, and population planning. With the exception of “energy and PVOs”, the functional categories are the same for FY 86. While the purpose of these accounts is to enhance congressional control over the direction of DA funds (tracking DA obligations for consistency with appropriations), it hamstrings AID ability to respond to changing development needs (GA0, 1983:56). The broad functional

US Aid to Latin America

107

approach to appropriations for AID is often inconsistent with projects developed at the Mission level. Missions often rewrite projects to fit existing accounts from fear of losing funds (unliquidated budget authority returns to the Treasury), even though the sector may not need that activity according to the Mission. Hence, the AID programming-budgeting and congressional reauthorization process tends to be an automatic affair that allocates resources according to internal bureaucratic needs and larger security interests rather than developmental need. A serious limitation on project efficacy in Latin America is that the allocation systems presume developmental results in one fiscal year. If obligated funds are not spent for a variety of legitimate reasons (and cannot be deobligated and reobligated to other projects), the funds revert to the Treasury or general fund. In FY 85, AID deobligated $29.5 million from DA and reobligated S59.3 million to ESF. It also returned to the Treasury $1.7 million (USAID, 1986:12). The rule on annual budget authority operates as a disincentive to terminate unsatisfactory projects and works against real foreign aid efficacy. The basis for this practice is distrust among foreign aid actors (AID and OMB) as noted in the last chapter. AID authority to reobligate funds that have been deobligated could reduce incentives for careful programming and diminish both OMB and congressional control over foreign aid (GAO, 1983:57). Further, AID budgeting and programming are not the only determinants of foreign aid results. Latin American countries and AID consultants prepare budgets for implementation. According to Wynia (1972:175) consultants in a Central American farm to market road program deliberately underestimated costs to increase the benefit-cost ratio, increasing project profitability and likelihood of funding. Foreign aid results, of course, are also determined by country fiscal conditions which are not directly controllable by the US. A downturn in commodity, oil and mineral prices can wipe out local counterpart funds needed for project completion. Hence, foreign aid efficacy depends on local political economy issues as well as domestic US foreign aid politics. Nobody has yet accused foreign aid of being a science! The final potential constraint to foreign aid efficacy is the evaluation phase. The efficacy of future aid programs depends in large part on critical comparative information derived from past failures and successes. AID monitors project activities by periodic site visits to “ensure regulations are observed” by contractors and PVOs (GA0, 1983:57). GAO suggests that “the need for program oversight is further supported by the fact that AID works in a high-risk environment with a high vulnerability to waste, fraud, and inefficiency” (1983:57). Much of the monitoring effort is then designed to control use of obligations and disbursements in a narrow accounting sense. On the one hand, this kind of approach can lead to ineffective results (penny wise, pound foolish) where narrow professionals assess innovative approaches to the administration of difficult problems in, for instance, agricultural development. Montgomery (1967:76) noted how US aid was withdrawn from a successful Latin American program of “servicios” (special organizations) because they were suspected of “loose accounting practices. “On the other hand, where large amounts of US funding flow into a crisis area such as El Salvador, the possibility of real accounting control diminishes while the need increases. Erik-Nelson (Miami Herald, November 8, 1983) argued that the US should send aid administrators (accountants) out into the Salvadorean countryside to supervise US-sponsored

108

The Politics of United States Foreign Aid

projects because the more than $1 billion in aid, entering on the pretext of land reform and economic development, was not reaching the intended beneficiaries. Similarly, AID evaluations tend to be financial rather than program oriented. Aid conducts “ongoing” project evaluations at the mid-point of project life and “impact” evaluations at project completion, focusing on issues of interest to higher management. The AID Office of Inspector General also evaluates by audits and investigations of aid programs (GA0, 1983:57–58). But most evaluations are either too narrow or general for managerial use. Evaluations are particularly important since if done properly, and actually used, they could remove many of the constraints caused by problems in programming, budgeting and execution. Regardless of the institutional dysfunctions that generate hastily-prepared often ill-directed projects, if evaluations revealed this to both AID and to Congress, the foreign aid policy process could benefit from making course corrections. But little evidence of systemic learning exists. One AID official, responsible over a lengthy period for many project preparations and approvals for El Salvador, expressed disbelief to me in 1981 that the country could be going under after all the “successful” projects that had been completed there! The possibility that success was measured in too easy terms, or that success of one sectoral project could be unrelated to the larger programmatic picture of poverty and underdevelopment in El Salvador, was an alien and quite repugnant thought to this official. He had done his job well; if El Salvador folded, something was obviously wrong with El Salvador that neither he (his job description or role) nor the US government could be responsible for! Under these conditions, emphasis on the short-term and tangible genie of military assistance becomes more understandable. In short, AID monitors and evaluates its programs/projects on criteria that are required by its programming process. Projects must appear to accomplish stated objectives in quantifiable form. For example, AID Administrator McPherson recently argued that El Salvdorean land reform was a “success” because, among other things, 50, 000 applications for land title “had been filed” by campesinos since 1982 and 3700 families had been “reinstated” to lands from which they had been evicted illegally by former owners (McPherson, Washington Post, March 20, 1983). Such indicators, however, do not measure titles “issued” or indicate how long “reinstatements” last, though this data could easily be obtained. But they sound forceful and positive at ABS meetings and this provides the rationale for increased funding, i. e. recycled premises into new programs and budgets. The evaluations follow rules that tend to squeeze out more of reality as information rises in the system. Missions, allegedly, have the broadest view of the locale. However, Missions can fall prey to the perspectives of US contractors operating locally. Further, rules require classification of reality into projects that fit appropriation categories. The central office (AID/W) is top heavy with rule-oriented staff (despite the fact that many occupied Mission roles before as part of normal AID Foreign Service rotations) and this encourages machine-like production of traditional projects for the pipeline. Congress has paradoxically the broadest and shallowest view and could require substantive evaluations. Instead, Congress tends to view itself in the narrow but contradictory role of aid financial guardian for profitable development projects and military aid giveaways. Its presumed expertise in foreign affairs (which varies mostly by committee leadership), often boils down to statutory checklists that intrude on AID

US Aid to Latin America

109

management. This constraint renders aid more financially-oriented (in a federal budgetary sense) and more technocratic.

Role Conflict The BRC model presumes that foreign aid policy is the result of ways in which bureaucratic actors define their missions and develop their structures to carry them out (May, cited in Cotler and Fagen, 1974:141). A particular mixture of economic and military aid for a country is then a product of conflict over both institutional missions (turf) and rule structures. Much of this reflects relative power differences in foreign aid policy-making. For example, how much influence does AID have in the determination of level and kinds of military assistance to El Salvador? How much influence do AID Missions have in developing innovative projects inconsistent with current programming and budgeting practices at AID/W? If State and Defense determine the overall scope and purpose of aid on a country by country basis, why have an “independent” foreign aid agency (AID) at all? The BRC model suggests that the quest for secure and profitable development of agency resources by the present structure of “guardian” and “spender” roles in Congress and the executive branch produces aid results that are formalistic, unintended, and susceptible to recipient ward status. These results are generally damaging to both donor and recipient since the donor policy actors do not gain substantially at the expense of other donors, trade competitors or military actors, and the recipient societies do not maximize use of aid resources. While foreign aid policy-making is apparently fragmented into competing fiefdoms, real control over the level and scope of the program is maintained by only a few actors, notably DOD and the Presidency. That conflict is between actors with ill-defined roles affects the substance of policy debate by excluding key values, such as developmental need and how best to attain it (including non-capitalist means). According to “incremental” budgetary theory, the clash of clear guardian and spender roles eventually (perhaps several annual budget cycles) will take most values into account. It should also be noted that some actors have fixed and clear roles but remain “officially” outside the policy process. For example, nongovernmental actors such as banks, unions, multinational corporations and universities may have an influence on foreign aid programming and results that the foreign aid policy machinery cannot directly control (Feinberg, 1983:39). On some issues, such as access to capital markets and technology, the US government may have less influence than these non-governmental actors (Lowenthal, 1983:322). Let us examine how pursuit of goals and subgoals by actors with “mixed” guardianspender roles affects foreign aid to Latin America. USAID, for example, has been both guardian and spender for its programs in this region. For instance, AID advocates more projects each FY as proposals are generated and debated within its Missions, Regional Bureaus and Washington office (centrally funded projects). Over the last 6 years, with “continuing appropriations resolutions” as the rule, total foreign aid request increases have remained relatively constant at about 8%-10% annually. But total foreign aid funds have also been allocated to “hot spots” such as El Salvador and the Middle East not because of the AID project cycle but rather from US strategic interests (DOD, President and Congressional) in those areas. Much of the friction between the actors preferences is reduced by the vehicle of “supplemental appropriations”, where Congress provides additional budget authority

110

The Politics of United States Foreign Aid

after the date of original appropriations acts, which may or may not be carried forward from one FY to the next. For instance, in FY 85, $1.1 in supplemental appropriation funds (ESF) and $68.7 million (DA), mostly for El Salvador and the Middle East, were carried “out” of the fiscal year into the next (USAID, 1986:12). Thus, AID “guards” its missions and programs in conjunction with other actors such as Congress, but also “advocates” greater funding during the request phase. Department of State is an important “guardian” of foreign aid since AID’s resource levels are set “in consultation” with it as overseer of the process (GAO, 1985:3). Though State Department support of aid packages tends to vary by issue and country, it often lines up behind DOD security solutions. This is due in part to a split between “old Latin American hands” (LaFaber, 1984:246) and “soft school” advocates within State. However, in a 1977 debate over use of aid for the San Lorenzo dam project as leverage against recent victors in a fraudulent El Salvadorean election, State human rights advocates who wanted a total cutoff finally won. State was then overrruled by President Carter and aid was resumed (LaFaber, 1984:247). Here guardians and spenders were not debating the substance of the aid (as noted before) but rather its leverage value to General Romero and against potential loss of influence in the region if European aid agencies built the dam instead. Where potentially revolutionary or anti-current military regime threats exist, State tends to support the regime, while AID often fragments in different directions: top appointees back DOD and CIA; Missions and Regional Bureaus often back regime opponents to the extent possible within this system. In some instances State is actually more security-oriented than DOD. In the 1940s, State opposed Pentagon plans for militarization of Latin America on the grounds that the Latin American nations could not afford the weapons. “They would go bankrupt” (LaFaber, 1984:92). But beyond the narrow scope of “profitability”, where funds existed (FMSC “soft sales” or MAP grants), State supports security solutions. For instance, despite the fact that 2% of the Guatemalan population owned 72% of the land in 1952 (1984:115), State believed that popularly elected President Jose Arevalo’s “restrictions on oligarchic property rights smacked of communism” (1984:119). By the 1980s, the Pentagon had learned the budgetary value of not losing wars and therefore opposed Secretary of State Haig’s military (East-West) approach to Central America. One of the ironies of foreign aid is that civilian military theorists, i. e. academic Pentagon consultants, often demonstrate this quality of reckless bravado in their policy recommendations to a far greater extent than seasoned military professionals (or “lifers”). Still, in 1979, State opposed US support for remnants of Somoza’s National Guard who were simply looting, murdering and dive-bombing slums by this time in history. DOD, Carter-Brzezinski and CIA support for “preserving order” via support to the National Guard prevailed this time over State’s more reasoned position. Since USAID is technically in State, the shifting politics of State coalitions tends to increase difficulties of programming aid over the long term. On the one hand, since State must react to daily foreign policy crises and facts vary enormously, any rational attempt to program aid in long-term aid to State inputs must end up in confusion. This is the problem of mixed roles for both agencies: AID and State are sometimes advocates and guardians of aid for differing reasons. The eventual consensus achieved has more to do with power politics than technical judgments, meaning that weaker actors such as AID tend to have less control over the program.

US Aid to Latin America

111

As noted before, the aid program is designed by many hands from different agencies but AID is responsible largely for execution. Admittedly, responsibility is often hard to pin down, as in ESF which is “administered by AID and the Department of State” (Development Coordination Committee, 1985:110). Regardless of which State faction prevails, it still tends toward militaristic solutions in most cases. For instance, Costa Rica has always presented problems for more simplistic State Department policies of support for military regimes and dispensing aid to keep things under control. In his pre-Presidential role as heroic social reformer and romantic revolutionary, Jose Figueres had signed the 1947 Pact of the Caribbean in part to rid Central America of dictators like Somoza. Though the State Department did not openly support Figueres against Somoza, it refused to help anti-Somoza (early contra contras!) forces because he was merely authoritarian and not totalitarian (LaFaber, 1984:104). In the 1950s, State could not classify Figueres as a “fish such as Arevalo nor a fowl such as Somoza” (1984:105). State was suspicious of the region’s major non-communist, non-dictatorial leader. This policy continues down to the present day attitude of taking them for granted as a pro-US Central American democracy. For FY 85, Costa Rica nevertheless received the third highest level of US aid in Latin America, $216 million, of which $160 million (74%) was ESF and only $20 million (9. 2%) was DA (USAID, 1986:15). The relatively high aid level is explicable by the regional turmoil in Nicaragua and El Salvador, and by the quick-thinking AID staffers who, after learning in 1978 that Costa Rica’s per capita income was above the US aid cut-off point, redefined Costa Rica as a “middle income developing nation” to keep the funds flowing to probably the most civilized AID location anywhere. But most Presidents have taken the Costa Ricans for granted. Nixon preferred Somoza and Carter focused on Nicaragua and El Salvador while turning the Honduran military into his closest ally in the region. “Not even Costa Rica’s unblemished human rights record, a record light-years ahead of any other in Latin America, won any special favors. Central America’s lone democracy was too anti-military, too anti-dictatorial to fit US policy” (LaFaber, 1984:268). The other major “mixed-role” agency in foreign aid policy-making is DOD. Defense plays the role of advocate for increased military aid (MAP, FMSC, IMET), to receptive regimes. These turn out to be, in most cases, repressive authoritarian regimes (Guatemala) or military institutions within democratic authoritarian governments (Mexico, Uruguay). DOD is also guardian against provision of DA to revolutionary or deviant governments whose policies could generate political instability. DOD equips, trains, and arms Latin American governments in their quest against subversives, dissidents and other regime opponents. For FY 87, requested budget authority for security assistance (including ESF) to Latin America was $1.3 billion, second only to the Near East and South Asia ($6.1 billion) (DOD, 1986, Volume 2, p. 185). The repressive practices of Latin American authoritarian governments have been relatively clear in the dramatic cases of: Uruguay against the Tupamaros, Argentina against the Monteneros, and Chile, and Brazil against their respective leftists. In many cases, these governments had legitimate claims to preventing violence and terror by what turned out to be overzealous bands of revolutionaries. But in all cases the governments began, with US assistance, to practice a form of counter-terrorism just as indiscriminate and damaging to the fabric of law. As stated by Klare and Aronson (cited in Fagen, 1979:139) “The US has supplied military juntas that have taken over country after country in Latin America with a steady stream of arms and military aid. “A similar pattern

112

The Politics of United States Foreign Aid

has been described for the US in Central America. In both cases, though the US prefers middle of the road, liberal kinds of regimes, through military support of existing armies which eventually take civilian command, the US finds itself repeatedly locked in with the most reactionary elements (of both the military and civilian groups) in each country. Where this has not happened, as in Costa Rica, it is because the country has no army to begin with and consistently refuses support to build one up as a matter of democratic principle. It cannot be said that US military aid causes golpes or repression. Excluding ESF (as does the AID “Green Book”), US military aid to Latin America has been quite small. Of $70.1 billion in US military aid to all regions (1962–84), Latin America received only $2.4 billion (3.4%), while the Near East and South Asia received $36.8 billion (52.4%) (USAID, 1984). But it can be argued that US support for existing military practices tends to legitimate them and in that real sense, makes pluralist solutions harder to attain. Put another way, the high propensity to stage golpes in Latin America is partly the product of an extremely individualistic, macho-flexing political culture. But US support for factions of questionable democratic intention, e. g. present day Nicaraguan “contras”, adds to the instability except where the regime has been unreceptive to US military aid: Costa Rica, Venezuela, and Mexico. It could be argued that defense advocacy by DOD overwhelms economic aid (DA, PL 480) advocated by AID and this distorts results toward the security end of the spectrum. But this argument would be too simplistic. A more comprehensive explanation would note persistent goal (or role) displacement by foreign aid actors. The roles are ill-defined and mixed, in part, because actors tend to displace them with more feasible ones. Tendler (1975:49) notes how a constantly-criticized USAID began to institutionalize criticism by “coming to identify with the very interests of the bureaucrtic entities which it was trying to fend off. “For example, AID must confer with Treasury before making loans, and State in proportions of security-development aid packages each fiscal year (the noted SAPRWG process leading to the “integrated” aid request). Over time, AID begins to sound and act like its former policy adversaries. AID becomes an advocate for interests not directly related to DA, e. g. ESF for semi-defense purposes, and this weakens its capacity to serve as legitimate spokesman for Third World development. USAID complicity in military assistance (willing or unwilling, as in the Mitrione affair in Uruguay), reduces its legitimacy, and to the dismay of AID line practitioners who seek positive developmental results for their work, AID becomes another “tool of imperialism”. Similarly, DOD and the President, in most cases, guard against economic aid to revolutionary governments. The most significant short-term vehicle for US influence of a regime’s direction is military aid. Economic aid is mostly long-term and results, as noted, are a matter of intense debate. Arms sales, credits and training provide the easiest path to influence what is usually the future governing elites in the country’s strongest institution. Training at American-run military schools generates respect for US military prowess and, it is hoped, respect for the US concept of democracy which they defend. Where the political situation is not too unstable, military aid can provide quick returns in the short term for US influence, particularly trade and investment. Hence, USAID becomes an advocate against DOD guardians in cases where potentially unfriendly groups are likely to gain power. But the seemingly natural affinity for socially-oriented revolutionary opposition groups and the developmental goals of USAID, is tempered by the existence of rigid counterpart roles

US Aid to Latin America

113

on the one hand, and imbalanced conflict between mixed or ill-defined roles on the other. DOD has most of the resources, political support and persuasive arguments. USAID tires of taking the “soft school” approach at some point, and displaces its goals to become more consistent with the majority opinions on what aid should look like. Whatever problems of information distortion exist in executive branch foreign aid policy-making, they should be clarified during annual congressional reauthorization and appropriations processes. But congressional advocacy and guardianship roles are often based on very little of the prior executive policy process. Generally, Congress is less interested in the substantive connection between aid and development than with the political leverage that the dollar sum of aid funds can give to the US. It views aid as a means to this end. By contrast, executive actors tend to view aid as the end itself: security, food, health, education, housing, etc. For decades, US military and economic aid to Haiti was perceived as a means for leverage. The bulk of assistance (98.87 of which has been economic in the 1962–84 period, USAID, 1984:51) supported the heavy-handed dictatorship of Baby Doc Duvalier. Despite multilateral aid to the same regime, Duvalier was viewed by the international community as a US tool and puppet. And here it is that foreign aid goes down ignominiously with foreign policy failure. But the severity and desperation of the local situation, in the words of Montgomery (1986:90), may have actually “favored a bold effort. “Throughout the last six years, the performance of several USAID projects there has been spectacular though largely unnoticed. The question is how such successes can occur in a context where aid was often diverted by the former Duvalier regime. Several factors have been important. First, guardian and spender roles have been relatively clear. Congress has acted consistently as guardian on issues of human rights and accounting abuses in Haiti. The AID Mission in Haiti, imbued with strong leadership from 1981 to 1985 implemented a clear developmental plan for “bottom up” development often through PVOs and around established Haitian ministries. In Haiti, the congressional human rights “constraint” on AID programming actually worked as an “opportunity” to build political consensus for “successful intervention” to remove Baby Doc. Advocates in AID/Haiti, Ambassador Clayton McManaway, and congressional staffers toughened the language of the aid certification process (FY 82 Appropriation, Section 511), making free labor unions and democratic reforms a new requirement for US aid. Congress also earmarked $1 million for a literacy program at a time when the Duvalier regime was cracking down on the Catholic Church’s literacy program there (Richburg, Washington Post, February 23, 1986). In the end, even President Reagan was persuaded by the flow of exiles, the aid corruption in Haiti and the “unusual coalition” of human rights advocates to champion human rights himself! Despite the repressive context, development projects have succeeded in bringing income and employment benefits to specific target populations. For example, the AID Agro-Forestry project has been a showcase, relying upon small producer interests to expand fuelwood marketing opportunities. In 1981, USAID gave $8 million to Haiti to halt deforestation and increase small farmer income and employment opportunities. By 1984, 20, 000 farmers had planted 13 million seedlings—5 million more than the original goal (Leigh and Alexander, Atlanta Constitution, December 10, 1985). Though Haiti is still rapidly losing topsoil, population is growing rapidly and less food is being produced (Simons, New York Times, June 15, 1986), the AID project is a micro example of balanced conflict. Spenders (small

114

The Politics of United States Foreign Aid

fuelwood producers) clashed realistically with guardians (buyers, mills, processors) over prices and quality, leading to incremental mutual gains by all parties. Second, the goals have been clear (economic development) and rules have not impeded action consistent with that objective. US military assistance was extremely small. In contrast with the El Salvador model, Haiti did not experience a left, right, center clash over allegiance from the poor and dispossessed. The poor are relatively homogeneous (most of the country) and this greatly simplifies planning. Finally, strong AID Mission leadership guided local projects through and around the AID/W mazeway to increase chances of success. The relentless quest for certainty by AID/W programmers did not impede field flexibility because AID/W “old hands” were in Haiti steering projects between the waves. The ingredients of successful aid policy then: (1) clear guardian and spender roles at all levels of policy-making, (2) strong AID Mission advocacy of interests consistent with recipient needs (meaning realistic communication between AID and local counterparts), and (3) absence of a serious security threat that repeatedly blackmails aid success. Despite the clarity of congressional action in Haiti, however, generally Congress views aid as a means of leverage rather than a substantive end in itself (or means to development). There are several reasons for this behavior. First, Congress is disinterested and/or hostile to foreign aid because: (1) Congress lacks consensus on the rationale for foreign aid, (2) foreign aid supporters cannot really demonstrate successful impacts, and (3) most domestic constituencies are either opposed or too weak to matter. Second, given the surreal quality of this issue before Congress, i. e. no tangible gains but potential losses connected to votes, foreign aid is susceptible to ideological posturing. It is not a substantive policy issue to Congress on which much political capital is going to be risked one way or the other. Third, the House Foreign Affairs Committee tries to annually reauthorize the foreign aid program. But according to Feuerwerger (1979:64), this is not a highly respected committee, meaning that its measures are tainted with unpopularity because of the program and the committee assignment itself. Finally, the Executive dominates Congress with respect to diplomacy and arms sales because it controls the basic tools of foreign policy initiation: (1) information, (2) a staff with the duty of formulating policy (programming, budgeting foreign aid), and (3) constitutional prerogative (1979:178). In short, Congress can add to executive requests or thwart them. But it usually cannot play a positive role because it lacks capacity under current practices to set the foreign aid agenda. Nor, in many cases, does Congress know where the money goes after appropriations. Specifically, Congress has such difficulty controlling reprogramming and fund transfers by DOD and the President that it, like OMB, tends to cast a jaundiced eye on all of them even where necessary for managerial flexibility. According to Nathan and Oliver (1983:128) Congress lost control of US foreign aid in the 1960s and 1970s as the military aid program was used to “establish surrogates for US presence in the Third World.” Congress treats foreign aid superficially unless a crisis occurs, at which point it often overreacts with ideological purity and political posturing. It normally is both guardian and spender on foreign aid authorizations. As noted, the major congressional actors are: the authorizing committees (Foreign Relations and Foreign Affairs), the appropriations committees together with their more influential subcommittees (Foreign Operations, and Inter-American Affairs), and of course, the Armed Service Committees. Given the use of foreign aid as a means to narrow agency objectives that are controlled by dominant actors in the policy process, it is no surprise that the same interests influence Congress. Annual

US Aid to Latin America

115

reauthorization hearings tend to focus on crisis issues such as El Salvador, and “aid” to the Nicaraguan contras. As noted, these committees have neither the time nor interest in conducting in-depth country analyses of foreign aid. Over the last several fiscal years, as noted, they have not even finished reauthorization work and the Reagan Administration has largely ignored them despite the legal basis of their actions as prerequisiste for budget authority. In the long run, other than for crisis areas and special interest “earmarks” such as Israel and Egypt, individual country analysis and lobbying is discouraged by the entire foreign aid policy process. This lowers the level of public information and interest in the subject and works to the advantage of dominant interests (DOD) in the executive branch. In contrast with the low level of debate and analysis of economic aid, greater interest is generated by security assistance appropriations. This is for two reasons. First, the linkage between aid and results is palpably clearer in the short run. The Reagan strategy of support for “freedom fighters” such as the contras is viewed (even by House Democrats which have approved FY 87 aid to them) as a “rapid return” use of aid. However, Congressional weakness at guarding against “covert” aid appropriations may be due in part to lobbying by Sandinistas in Congress (Cannon and Hoffman, Washington Post, February 9, 1986). Further, support for the more moderate elements of the El Salvadorean military may have finally paid dividends in permitting the elected President Duarte to carry out democratic policy reforms without left-right reprisals that blocked change in the past. Finally, lobbying by Angolan guerrilla leader Jonas Savimbi (the “Che Guevara of the right”) in Congress may have produced $13 million in a 2-stage “covert” aid program for his freedom fighters (Tyler and Ottaway, Washington Post, February 9, 1986). In short, lobbying by narrow defense aid interests produces greater dividends than that mustered by economic aid splinter groups which lack power and demonstrable results unless a crisis exists. Second, political power and information are intertwined in the DOD-Armed Service Committee relationship. Military aid programs emphasize procurement and sale of weapons to Latin American and other Third World military institutions. Both Armed Services and Appropriations Committees tend to approve DOD (impliedly armament industry) requests on this (Woll, 1985:449). Constituent support for wider defense expenditures stems from clear in-district benefits. The committees advocate military aid intensely and they can usually substantiate their requests with hard data (though always with an unexamined premise of soft results like “political moderation and regional stability” (DOD, 1986, Volume 1, p. 32).

Conclusion In sum, aid to Latin America is the product of a process that rigidly examines the scope of the program according to criteria that favor security or “hard” solutions in crisis contexts. The bulk of countries are not examined regularly by the process as a whole in other than superficial terms. Reacting to individual country crises or events, Congress proceeds to expand the list of statutory requirements to which AID must conform. This satisfies constituents on high-visibility issues such as El Salvador and Haiti. But it tends to inhibit communication between executive policy-makers and Congress. Congress thus weakens its own role and diminishes the value of foreign aid by refusing to treat economic aid as the primary means to development, rather than as diplomatic leverage. In Latin America, the

116

The Politics of United States Foreign Aid

lack of policy continuity, indicated by repeated and sudden injections of non-developmental aid via supplemental appropriations, is largely the result of superficial conflict between a rule-bound executive and a Congress which retaliates with more rules to keep ostensible control of foreign aid. It cannot be said that more thorough debate of foreign aid funding to each Latin American country by both the executive and Congress, beginning in 1944, would have eliminated all developmental obstacles and produced functioning political democracies in each case. But it would have helped!

Chapter Five US AID TO ASIA

Introduction Wolf (1960:249) suggests that “The objectives of foreign aid are part of, but not identical with the objectives of foreign policy. ‘He noted over 25 years ago what is still valid today, that some foreign policy objectives, such as reunification of Germany or Korea, have little or nothing to do with foreign aid. But the implication is that short of such larger goals, foreign aid and foreign policy objectives are or should be consistent. Throughout the long history of US foreign aid to Asia, the presumption has been that foreign aid was foreign policy. The mostly military objectives in this region since World War II have been linked to the foreign aid program. In practice, failures of the former have been blamed on the latter; the spectacular successes of the latter have been attributed to the wisdom and foresight of the former. Such fallacious thinking has been detrimental to planning and execution of the US aid program in this region. With the exception of the Middle East (Chapter Six), in no other region have foreign aid objectives been twisted so far out of shape to meet foreign policy goals. Because the machinery of foreign aid (institution-building projects, PL 480, FMSC, MAP, and ESF) is perceived as the executor of foreign policy, it has been said that “Nowhere in the Third World has there been a greater gap between American ambitions and accomplishments, or between the price America has paid (in blood, treasure and domestic division) and the return on its ‘investment’ than in Asia” (Gurtov, 1974:126). At the same time, nowhere in the Third World have foreign aid programs achieved such spectacular successes. Using the BRC model, this chapter will attempt to explain some of these successes and failures. As indicated in Figure 1.2, the postwar expansion of US aid to Asia coincided with expansion of aid to the Middle East and reduction of aid to Europe. From 1946–84, US aid to Asia reached approximately 23.6% ($66.4 billion) of the worldwide total ($281.1 billion) or second in regional priority behind the Middle East (about 32.7%; $91.8 billion) and far ahead of Latin America discussed in Chapter Four (about 6.8%; $19.1 billion) (USAID, 1984). Of the $66.4 billion to Asia, 57.4% ($38.1 billion) was military aid, not including ESF, and 42.6% ($28.3 billion) was economic aid. Only four countries received 66% of all regional aid. South Korea, Taiwan, the Philippines, and Vietnam received 80% of regional military aid and 62.4% of the economic aid in the 1946–84 period (USAID, 1984). By 1984, US aid to Asia dropped in significance to roughly 5.0% ($751.3 million obligated) of the total ($15.3 billion). This figure was far behind the Middle East (51.6%; $7.8 billion), Latin America (10.0% ;$1.5 billion) and even Africa (10.0%; $1.5 billion) (USAID, 1984). Much of this was explicable by the drop in perceived regional security threats after the Vietnam War. Regional aid will likely increase from FY 86, largely to stabilize the Philippines, again for security reasons. Consistent with this assertion, for FY DOI: 10.4324/9780203840184-5

118

The Politics of United States Foreign Aid

87, planned US aid to Asia increased to 14% (see Figure 5.1) of the total which moves Asia aid back into third place in LDC bilateral aid behind the Middle East and Latin (mainly Central) America. Foreign aid to the Far East and East Asia was employed initially as a means or strategy of “containing” Soviet expansion in this region. The Mutual Security Act period from 1952–1957 consisted mainly of military grants to the Middle and Far East in support of US military alliances. Institutionally, US aid to Asia is now merged with aid to the “Near East” in the new AID Bureau for Asia and the Near East. But AID still distinguishes programs to South and Southeast Asia from the Near East-North African area (AID Congressional Presentation, FY 87:367–76). It will be argued that aid to Asia exemplifies the BRC model proposition that where institutional roles are not defined and extreme power balances persist, informational distortions contribute to results that encourage dependency, excessive technological complexity and unintended consequences. Conflict is unrealistic in the sense that information presented by relevant actors is not received on its merits, but tends to be accorded the same status as the agency or committee which presents it.

Figure 5.1: Allocation of Bilateral Foreign Aid By Major Region, FY 1987 Source: Larry Q.Nowels, “US Foreign Assistance In An Era of Declining Resources: Issues For Congress In 1986”, Washington, D.C.: Congressional Research Service of the Library of Congress, p. 15.

US Aid to Asia

119

Foreign aid policy thus becomes rigid and often inconsistent with recipient needs. Where this has not occurred, conditions predicted by the BRC model are exceptional—clear roles, evolutionary conflict, institutional trust, and so on. Foreign aid failures in Asia can largely be explained by the singular US concern for security (meaning a foreign aid policy process that provides greatest access to this objective) and the discovery of “ward” regimes devoted to this cause. As one would expect, successes tend to occur when larger security problems have been eliminated and program interferences minimized.

Goals Since foreign aid served as the major vehicle through which larger US foreign policy operating concepts affected Asia, it may be useful to begin with “goals” and to explain how “results” were a product of visionary aims translated across rules, repertoires and institutional actors. In Asia, nearly all of the major US “liberal” foreign policy notions became central foreign aid programming (but not implementation) premises, such as: coalition governments-elections, profitable aid, humanitarian welfare, and political leverage via stability and security. Let us examine how the design and execution of economic and military aid to the four regional recipients: Korea, Taiwan, Philippines, and Vietnam, were affected by these premises. During and after World War II, the Soviets shifted pressure on the US from Europe to Asia. With an emerging “bi-polar” view (Sino-Soviet bloc) of the world as decision premise, the US stepped up military and economic aid to receptive leaders and regimes in this region in order to “leverage” its world view. Since most countries in Asia had only recently emerged from “Western colonialism” (Spanier, 1983:51), their nationalistic and anti-American feelings were very strong. Additionally, in contrast with Latin America where democratic and progressive political traditions often pre-dated US influence, political systems in Asia tended to be authoritarian, often with collective patterns of property ownership. The task of “conversion” to the American “liberal” way of life via aid would be more difficult and costly under these conditions. Despite more obvious contradictions between US “developmental” and strategic premises, US foreign policy forged ahead, applying “liberal” doctrines since 1947. As applied to Taiwan, South Korea, Vietnam, and the Philippines, these premises translated into the use of aid for security, profitability and developmental purposes. Packenham suggests that American political development doctrines reflect four key assumptions: (1) change and development abroad are easy goals to accomplish, (2) good things like economic and political development go together, (3) radicalism and revolutions are bad, and (4) power should not necessarily be centralized (1973: Chapter 3). For example, easy change and development is a premiere US premise because, with the exception of the Civil War, they have been comparatively easy to attain in the US (1973:112). This translated into such optimistic foreign aid assumptions as South Korean “economic self-reliance within a relatively few years” (1948–52), and the use of technical assistance to preserve moderate governments in Burma, Indochina, Thailand and the Philippines (1973:114). We have already noted this optimism toward the evolution of constitutional democracy in Latin America in the Alliance for Progress program. The optimistic assumption presumed that recipients wanted to change institutions and cultures to emulate the US middle -class capitalist democracy; those opposed had to be

120

The Politics of United States Foreign Aid

“outside agitators” or misguided deviant thinkers that could be educated back to sanity. So, for example, the US tried to leverage a unified China against the Soviets even before World War II ended. The US encouraged a “coalition government” to unite Nationalists (KMT) and Communists (CCP), which were actually two irreconcilable enemies committed to fight to the finish. Nevertheless, the US provided $2 billion in economic-military aid to a regime functioning in a country that was mostly anti-Nationalist. Lack of popular rural support for the Nationalists negated the significant numerical military advantages Chaing held in personnel and materiel (Spanier, 1983:59). Optimists remained convinced that Chaing’s corrupt, inefficient and reactionary government would provide a politically effective instrument to both carry—out socioeconomic reforms in China and support US “containment” policies against the Soviets. Unfortunately, permanent optimism (the sense of “divine mission” or “Manifest Destiny” to share the “white man’s burden”) as a premise opens one to easy manipulation by con -men who talk reform and anticommunism but employ aid to carry out their personal agendas or “crony capitalism”. Since the support of Chiang, the US has been plagued with the problem of finding effective conduits for its aid in Asia. President Truman released a White Paper arguing for US dissociation from Chiang and formal recognition of the Communist Government as the legitimate-official government of China. Had this been done, many of the dilemmas of the current two-China policy might have been avoided. But in the late 1940s, as in 1986, mixed-role conflict assured distrust among foreign aid policy actors and this permitted the most powerful and simplistic viewpoints to prevail in most cases. The advent of the Korean War in 1950 reinforced the liberal notion that China could still be reunited through Chiang, who was on permanent leave in Taiwan. Having learned very little from experience that translated into more flexible or informed foreign aid institutions, the US employed the same easy change assumptions in Korea, Vietnam and the Philippines without ever fully recognizing that constitutional democracy was perhaps the last goal of any of the contending groups in each country. The political conflicts amounted to the functional equivalent of tribal battles between feudal authoritarian groups (similar to faculty battles in many US universities) with the exception that in each case one group was able to manipulate aid from the US by saying the right things. In Korea, for example, the-US watched the growth of Soviet influence in North Korea with both trepidation and optimism. The US proposed to the UN General Assembly a free election for all Korea in 1947. After the Soviets refused to grant the UN commission access to North Korea, the US recognized South Korea as the offical republic and Syngman Rhee as its “legitimate representative” (Spanier, 1983:63). Even before the North Korean attack in 1950, the US had provided significant economic and military aid to Rhee to “bolster his non-communist government and help Korea establish a democratic society” (1983:63). Whether or not US aid would win the war for South Korea, and turn it into a wasteland or an industrial giant, genuine belief existed that aid could change an authoritarian society into a democratic one, largely propelled by the formative event of elections and coalition governments. More cynical observers, of course, viewed the election as a vehicle for legitimation of the US puppet regime or a “demonstration” election (Herman and Brodhead, 1984). The naive liberal view prevailed. Similarly, policy-makers believed that the US could turn Vietnam into a constitutional democracy with functioning parties and elections. The “loss” of China, attributed simplistically

US Aid to Asia

121

to the failure to pour more money down Chiang’s drain, together with the Korean stalemate at the 38th parallel, pushed US public opinion behind French efforts to regain colonial control over Indochina. The US provided economic-military aid to the French in 1954, to cover 75% of their war effort against the Vietminh. After the French wipeout at Dienbienphu, the optimistic US “easy change” assumption resurfaced in two places. First, the US would create another NATO in Southeast Asia called SEATO (despite the absence of a regional nationalistic consciousness as existed in Europe). Second, the US would support its imported leader Ngo Dinh Diem ($2 billion between 1955–60) in an effort to hold the “frontier” at the 17th parallel in Vietnam against communism. The legitimizing vehicle would be a 1956 election consistent with the 1954 Geneva Accords. Diem “abstained” from a national election and settled for continued US support as Premier. Whether he abstained against US wishes (Montgomery, 1986:54) or was restrained by the US from fear of jeopardizing its “frontiersmanship” strategy (Hernam and Brodhead, 1984:57), everyone (including President Eisenhower) knew he would have lost soundly to Ho Chi Minh. In short, the US imagined it could liberalize Vietnam though the vehicle of aid for an election to legitimize a politically bankrupt, isolated, and repressive regime. According to Montgomery (1986:54), Diem and his “American supporters viewed the massive southward migration of refugees from communism as a better indicator of popular preference than a staged election would be -a view shared by the French. “Unfortunately, refugee flows cannot substitute for a relationship of accountability and trust between societal groups and the regime. An election can be a formative event that can marginally improve chances for democratic development. But the US has almost consistently opposed “real” general elections (where the communist opposition actually participates) as “radical”, such as in support of Magsaysay in the 1950s and later Marcos in the Philippines against more popular candidates. The second premise of foreign aid policy in Asia has been that “good things” such as economic development, political stability and aid profitability, go together. We have discussed the stability-democracy presumption in Latin America and its more recent version as the “privatization” strategy of President Reagan: private sector growth leads to political democracy. Similar perspectives argue that loans (not grants) encourage thrift and responsible choices leading to self-help and democratic development. The mutual growth of nice things tends inevitably to be the result of technology transfer of products and techniques via US aid. Whatever natural order may exist between the concepts of growth, technology, development, and communist containment, US foreign aid policies have worked decidedly against “letting the chips fall where they may.” In the short run, aid seems to encourage growth and stability. Montgomery even suggests that early US successes in building the economies of Taiwan, Korea, the Philippines, Iran, and several Latin American countries may have contributed to overconfidence and excessive optimism (1986:79–80). The Truman Point 4 Plan for foreign aid rested squarely on the belief that “there was no technical problem in the newly emerging nations that American science could not address” (1986:80). But if aid technology transfer was successful in generating growth in Asia, it did not stimulate constitutional democratic development. Nor has it turned out, did aid “profit” the US in the narrow financial sense it was intended. One would expect, if good things go together, that aid would at least provide leverage to install democratic institutions.

122

The Politics of United States Foreign Aid

Since 1946, the US provided $6.5 billion in economic and military aid to Taiwan across Chiang and subsequent KMT (Kuomintang) “Chinese” party leaders (USAID, 1984:85). But Rumpf (Washington Post, 1984) argues that the Chinese have ruled Taiwan by martial law since 1949. Chiang’s forces killed an estimated 10,000–30,000 Taiwanese during anti-KMT riots in 1947. Chinese occupation of Taiwan after World War II was an Allied proposition. Thus, the KMT claim for continued US support there is extremely thin. Rumpf (1984) suggests that neither President Reagan nor Peking want a national referendum in which a majority of the Taiwanese people could choose between independence or “some kind of affiliation with the mainland. “Despite its liberal-premised aid, the US continues to support KMT “resistance to self-determination and nationalism.” Nor did aid to Vietnam ($23.3 billion and 35% of the regional total, 1946–84) or to Korea ($9.4 billion in the same period), provide any democratic reform leverage. By the 1950s, Saigon became the largest US aid mission and the largest military assistance advisory group in the world. In an atmosphere of “exhilaration” among US advisors, which Montgomery suggests may be a “partial explanation for the inability of Americans to perceive many advance warnings of failure” (1986:75), the “hard” (body count) and “soft” (education, health, literacy) schools of development clashed without having much effect on Vietnamese behavior. Just as each “school” had a separate appropriations account (DA and FMSC), in a figurative sense the recipient “accounts” were not integrated with these forms of aid by the US policy process. Montgomery suggests that in their delicate attempts not to interfere in Vietnamese affairs, US advisors often ended up advising each other instead (1986:70). Driven by optimistic and self-righteous support of Diem and later Thieu-Ky, the provision of massive amounts of aid could not prevent the combined political-military defeat of 1975. In Korea, the US also supported a string of corrupt dictators which, despite impressive economic gains not directly tied to US aid, simply maintained a repressive, anti-democratic regime. The US quest for stabilty to assure growth and development may be misplaced in South Korea. According to White (New York Times, March 5, 1985), it is the repressive Chun Doo Hwan regime which is unstable and the South Korean people that are industrious, educated and stable. He suggests that President Reagan is supporting a regime that openly maintains itself in power by “controlling the press, suppressing free trade unions, and depriving democratic leaders of their civil rights. Thus the regime actually threatens the stability of the nation..” Finally, he suggests (1985) that “History demonstrates that American assistance and advice rarely causes dictators to moderate their conduct when power is perceived to be at stake.” Clearly, US assistance and advice to the ex-regime of Ferdinand Marcos in the Philippines had little moderating effect. It can be argued more persuasively that the advice may have had more moderating effect on the Reagan Administration itself! Since 1946, the US provided $3.7 billion in aid (USAID, 1984) and additional revenues from maintenance of US bases (Clark and Subic Naval bases) in the Philippines. The US has applied its “containment” doctrine to the Pacific region for decades. Aid flowed to the Marcos regime since 1972 largely as “leverage” for maintenance of the bases. Though Marcos imposed martial law from 1972–77, USAID loans and grants “lept fivefold” (Lappe, Collins and Kinley, 1980:21). At the same time, real income and nutrition levels of workers and peasants steadily declined. By 1978, the Philippine per capita calorie consumption was second

US Aid to Asia

123

lowest only to war—torn Cambodia (1980:21). In 1979, President Carter agreed to provide $300 million in military aid and credit sales and over $l billion in economic aid for the next five years. Lappe et. al (1980:22) suggest that part of the military aid was used to defend against Filipino resistance to the Marcos regime. However, the “widespread fraud” in the February, 1986 (finally annulled) reelection of Marcos over Corazon Aquino (widow of assassinated opposition leader BenignoAquino), together with the recent and clear lessons of supporting unpopular dictators in Nicaragua, Iran, and Vietnam, may have had the dual effect of shifting US policies and sending a message to nearby regional autocrats (South Korea) that future US aid may dry up unless real reforms are made. Ballot box snatching by Marcos’ army and other procedural irregularities were widely reported (Branigin, Washington Post, February 9, 1986) in the US media. Despite all past US aid, Marcos’ “constitutional authoritarian” regime (his term) was plagued by capital flight, inflation, a declining economy and the status of being the only ASEAN nation with a growing communist insurgency (Branigin, Washington Post, August 12, 1984). High US policy-makers such as Secretary of State Schultz had suggested moving the military bases, while others such as Chairman of the House Foreign Affairs Committee Dante Fascell (D-Fla) recommended a total aid cutoff (Gwertzman, New York Times, February 20, 1986; Roberts, New York Times, February 20, 1986). Predictably, Marcos vowed to seek aid from the Soviets if such insults would occur. Convinced finally of his vulnerability, however, Marcos departed for Hawaii on a US Air Force jet on February 27, less than a month after Baby Doc Duvalier’s departure for France aboard a similar aircraft. The sudden shift in US foreign policy, brought about in part from recognition of the absence of leverage attained by billions of dollars in US aid, sent signals as far as the Chun regime in South Korea. Chun suddenly released 300 political dissidents from jail and now claims even to be in favor of constitutional reform, with the knowledge that US aid could be cut off ($231 million in military aid in 1984—none in economic aid) (Atlanta Constitution, February 26, 1986). Finally, recipient growth does not always lead to aid profitability. As noted, aid is often disbursed (by Congress, the President or State) because it is supposed to be financially and politically profitable to US interests. Though AID plans its projects on a “LOGFRAME” basis of rough cost effectiveness to recipient farmers, students, patients, workers, etc., Congress and the public (many of the same groups seeking Nirvana through a balanced federal budget) press for returns in the form of loan repayments and purchases of US goods and services (trade) to benefit US balance of payments. Financial profitability is even more important at present since the US is now a debtor nation. Historically, “tied aid” programs, PL 480 and soft loans have provided such returns to the US, often at the expense of indigenous development. Unfortunately, US aid to Asia has produced unintended effects for the US beyond the usual marginal welfare benefits and maintenance of the local regime. Ho (1978:117) suggests for Taiwan during the 1950s and 1960s “Without AID’s influence, and active intervention, the private sector would not have become Taiwan’s foremost source of economic growth. AID helped to create a more conducive atmosphere for economic growth, particularly for the expansion of private industries by: (1) financing government projects with strong external economies, (2) inducing the government to liberalize its economic policies, and (3) laying a constraining hand on military expenditures. “Judged according to growth and developmental criteria, US aid to Taiwan has been a spectacular success. But

124

The Politics of United States Foreign Aid

judged according to the liberal “all good things together” criteria, it is a failure because the regime remains authoritarian. First, Taiwan is a stable authoritarian regime which is unlikely to evolve into a constitutional democracy for some time. Use of aid for this larger kind of political leverage is resented and could be counter-productive to US-ROC relations. Second, the ROC like other countries in the area have sizeable positive balance of trade surpluses with the US. In effect, US aid has financed losses in US GNP by generating more efficient competition. Aid has been blamed for the Japanese success story, and now the emergence of a “second Japan”, South Korea! In 1985, South Korea exported $5 billion more to the US than it imported in steel, cars, electronic goods, and even photo-album covers (Auerbach, Washington Post, February 9, 1986). Part of the “profitability” problem with Taiwan may be that, in its quest for profit and success stories, the US terminated aid to ROC in 1963 (except for military grants). Kaplan (1967:279) noted the US “obsession” with success stories and recommended resumption of aid because it is a “major tool for protecting and advancing US interests. “Aid leverage might have prevented Korea from following Japanese practices of excluding US goods and targeting industries with licensing measures to keep out specific country products. The third liberal premise of US aid has been that revolution and radicalism are bad (Packenham, 1973:129). Though the first premise suggested that change was easy, the third defines acceptable change as coups, insurrections, and revolts but not revolutionary changes in social structure and values (1973:129). The distinction is similar to that of a marginal organizational change in accounting or budgetary practices as opposed to a major reorganization of functions. Where one concept begins and the other ends is often unclear in practice. Under the Alliance for Progress, as noted in the last chapter, the US pinned its hopes on the “democratic left” in Latin America. These were “reformers” not radicals— Haya de la Torre of Peru and Jose Figueres of Costa Rica (1973:145). But the record shows that the US has even remained suspicious of reformers such as Figueres because they could not be classified as “fish or fowl” (LaFaber, 1984:105). We have noted how “legitimate” aid recipients often liquidate “communist” opponents in Latin America with little outcry other than from US human rights advocates. The military government which overthrew Sukarno in Indonesia (1965) liquidated between 300, 000– 500, 000 “communists”. But the event went largely unnoticed in the US compared to leftwing terrorism by such groups as Fidel Castro and the Vietcong (Packenham, 1973:139). Sukarno’s “Guided Democracy” program was a broad coalition of political and social forces including the Communist Party (PKI) that could be termed “radical nationalist”. In 1957, President Eisenhower and his chief advisors planned a staged rebellion as an “ideological struggle” against a perceived threat to his “frontiersmanship” strategy. According to Gurtov (1974:189), they believed that “support of ‘anti-communist’ rebels might topple Sukarno, just as it had enabled Castillo Armas to overthrow Arbenz in Guatemala.” Sukarno, of course, was replaced by the Suharto dictatorship in 1965 and US military and economic aid was provided to such an extent that by 1980, Indonesia was in fourth place ($205 million) behind India as top recipient of US aid (Lappe et. al., 1980:18). (In FY 87, no Asian nation is among the top 13 recipients of US aid with the exception of Korea which is in 10th place;Nowels, 1986:16). US aid to Suharto was used largely to shore up political prisoners (100, 000 in 1977 is still the record for the greatest number of long-term

US Aid to Asia

125

prisoners). But Suharto blocked land reform proposals and squandered the nation’s then spectacular oil and gas export revenues ($10 billion in 1979) on luxury imports, military excesses and showy capital-intensive industrial projects. Roughly half the labor force was either unemployed or underemployed (Lappe, et. al., 1980:19). The US aid program there could hardly be considered successful in narrow LOGFRAME terms or in wider leverage considerations. Security has been the guiding premise of US aid to the Far East. The fear of revolution and radicalism tends to encourage aid relations that are based on false premises, misunderstandings, distrust and lack of mutual agreement on core aims. Such “mutuality of interests” must exist beforehand (Jacoby, 196&132; Montgomery, 1962: Chapter 2) or they must be reasonably attainable by appropriate role conflict among donor and recipient actors as argued here. In the Far East as elsewhere, the US tends to project its own image of political propriety (political democracy) and impropriety (radical outside agitators), while at the same time recognizing that it does not understand the local political culture. An AID offical once told me that at least in the Far East “they look different than we do so we can expect trouble. In Latin America we are constantly lulled into complacency because they look and even sound like us..” (!) In such ambiguous contexts, the easy solution is to build up the authoritarian center to hold society together in the hope that interest group pluralism can evolve and the center can relax its grip. This, of course, ignores the powerful effect of institutional roles on individual behavior—role redefinition becomes a deadly game of power between groups with substantially different resources, the military and civilians! So in “Eastasia” this simplistic objective has become remarkably consistent with the local regional tradition of one-party rule. But the authoritarian stability may also partly explain the economic competence that now surpasses the West (Hofheinz and Calder, 1982:71). It is unlikely that any amount of security assistance can buy the leverage that could be attained by first establishing mutuality of interests with recipients. For example, despite massive aid to the Marcos regime ($2.3 billion since 1962 of which 36% or $828 million was military aid), his continued political repression and related deterioration of the economy undermined the legitimacy of his own regime. Solarz (New York Times, August 8, 1984) noted correctly that “the Philippines bore a certain resemblance to South Vietnam in the late 1950s. Once again, the US appears tied to a corrupt and inefficient government that has lost the confidence of its people.” However, his recommendations illustrate the frequent contradictions of liberal US premises-condition future economic aid to Marcos on return to democracy and elimination of widespread corruption! Such naivete ignores both the near-insoluble recipient “fungibility” problem, and the existing lengthy Congressional statutory checklist prohibiting aid, for example, to monopolies, and countries that do not encourage “democratic private and local government institutions (Foreign Assistance Act, Section 102b). On paper, such reforms already exist. Instead, US aid has subsidized a well-documented decline in Philippine institutions. According to Branigin (Washington Post, March 25, 1984), the scale of military penetration into every aspect of society: stevedoring, postal, transportation, janitorial, has been unprecedented. From 1971–80, the armed forces expanded 279%, making it “the fastest growing military in Asia. “During the same period, military expenditures also increased the fastest of ASEAN nations while its economy grew the slowest. Yet somehow, the usual litany of administrative factors, such as: “poor maintenance and logistics”, “limited

126

The Politics of United States Foreign Aid

control and support”, “widespread corruption and lack of motivation” in the military, have weakened its ability to fight “a spreading communist insurgency” (Branigin, Washington Post, March 25, 1984). The suggestion that more aid (matriel and training) could remedy what has always been the contradictory provision of economic aid to stabilize the military regime of Marcos to serve US security ends was ludicrous. Despite US aid and advice (“training”, “technology transfer”, “institution-building”), the Philippines was and is “Asia’s economic deadbeat” (Branigin, Washington Post, August 12, 1984). The final liberal premise is that power should be distributed rather than accumulated in the hands of centralized political regimes. The problem of “government-building” for the US is not ‘creation of authority and accumulation of power (though absence of legitimacy seems to point in this direction), but rather abiding by the Lockean notions of limiting authority and dividing power. Institution-building strategies have always presumed they could increase service delivery capacities within a framework of checks and balances, separation of power, federalism, regular elections and competitive parties (Packenham, 1973:155). As noted in Chapter 4, the Alliance for Progress was the first real effort to discourage military coups and encourage elections and constitutionalism through appropriate disbursement of aid (1973:156). The other major policy expression was the Title IX amendment to the Foreign Assistance Act of 1961 (Section 601a) which sought to encourage the formation of groups, such as: unions, voluntary organizations, trade associations, savings and loan associations and credit unions. Here, the presumption is that centralizing power discourages popular participation and decentralization of power encourages it. This is an easy distinction, something like the familiar notions that: (1) all economic aid promotes good things like development, while military aid promotes bad things like dictatorships and oppression, (2) area specialists are more likely to be empathetic and knowledgeable about local political cultures than “crisis managers” who are chiefly into cost-benefit calculations, gamesmanship and power rivalries (Gurtov, 1974:206n), and (3) AID Missions are locally committed while AID/W people are committed mainly to careers and paperpushing. In fact, such distinctions are simplistic and represent more isolated (though memorable!) examples rather than systemic influences. Decentralized economic power arising from successful economic aid projects may lead to substantial growth and expansion of the middle class. However, this long-term result, achieved in the Philippines and Taiwan, may not lead to constitutional benefits as one would expect. Such decentralization of production may deadlock and diffuse power, which actually reinforces elite control. Paradoxically, short-term decentralization can expand access by privileged groups, whereas military aid may leverage out hard-line leaders such as Marcos or Duvalier, and their allies within the military. This, of course, requires extensive support from diverse US institutions including the President, all of which takes time to develop and is highly situational. Further, it suggests that the US should target “hard school” aid for “soft school” purposes where political power is highly stratified, i. e. most recipients. This has been the premise of the Civic Action Program: to enroll the military in civic functions and acculturate them to needs of the populace. The “hard” approach also challenges the widespread notion that military aid produces the usual reign of terror and repression while military aid builds up countervailing power that restrains the military by offering them a share of the gains.

US Aid to Asia

127

US aid (military or economic) has produced few federal systems where groups other than the ruling clique and national capital city retains wealth and power. This statement also presumes knowledge of how to measure progress toward such broader objectives from aid expenditures. But we don’t. According to Wolf (1960:25) “The higher we ascend in the hierarchy of objectives, the less useful the objectives are for determining preferred solutions to specific decision-making problems. “So we are back to the problem of how many landed small farmers can assure political democracy. Nevertheless, over time the effects of both types of aid can encourage mutual interaction. Business opposition to Marcos (much of which has been generated by US economic aid, especially PL 480), and military opposition to the use of force against Marcos’ opponents (sustained by US military aid) may combine to produce a broad consensus for constitutional democracy now that he has gone. For example, on February 23, 1986, the Minister of Defense and Deputy Armed Forces Chief of the Philippine Army, along with several thousand soldiers demanded the resignation of President Marcos and recognized opposition leader Corazon Aquino as “legitimately elected President” (Jones, Atlanta Constitution, February 23, 1986). The US responded with a clear signal of nonsupport for Marcos, giving the green light to the opposition. Though development is clearly last place among US aid goals as noted, in the complex and uncertain world of development assistance, it may be that the direct route (intended consequences) leads to growth but not development. Paradoxically, under certain conditions, investment in the right “centrist” military faction can lead to both growth and development (unintended consequences)! Brazilian and Venezuelan experiences with authoritarian rule shifting to permanent civilian-military coalitions or civilian institutions tends to support this proposition. Additionally, the US aversion (for DA, PL 480 and other forms of economic aid) to centralized power is inconsistent with regional Eastasian traditions of “dominant central administration” (Hofheinz and Calder, 1982:71). Power in Japan, South Korea, Taiwan, and elsewhere is concentrated in “metropolitan” hands which makes for policy “homogeneity” analogous to France (1982:71). This works against the US liberal premise, but is consistent with the top-down, bureaucratic workings of foreign aid policy-making as well as the intended effects of US military aid. Under the best of circumstances, foreign aid thus reinforces local structures in the hopes of changing them!

Results The results presumed by the BRC model (complexity, dependency and unintended consequences) have been shown to be a product of ill-defined budgetary guardian and spender roles translated across foreign aid institutions with extreme power imbalances. This volatile political context encourages AID to defensively emphasize planning routines, with the result that plans and implementation needs often contradict each other. Consequently, as noted in Chapter 4, economic aid programs often produce “successful” if unspectacular results in a narrow sense, and few positive effects on political development variables, such as constitutional government. Where security problems are absent, as in Thailand in the 1950s, this provided an “occasion for deliberate experimentation in the development and administration of technical and economic aid” (Montgomery, 1962:29). However, even under such favorable circumstances, successes on the US side must be matched by will and

128

The Politics of United States Foreign Aid

capacity of the recipient. For example, 500 fire hydrants were installed in solid concrete in Bangkok with no connections to the city water supply (1962:110). Success is often defined narrowly to mean “planning”, when the above example suggests implementation is just as important. Further, the relationship to many such project successes and evolution of the right to exercise political freedoms needs to be documented. Similarly, US military aid (to Latin America) has not produced regime stability; in most cases it has contributed to The cultural tendency to have golpes instead of elections. But in Asia there have been some dramatic economic aid successes; the Philippine example may be one of the few cases where military aid has provided leverage that moved the system toward democracy. The central question then is what distinguishes success from failure? It is argued that the most important factor is the existence of complementary USrecipient interests. Disagreements over project details can occur without threatening the agreed -upon boundaries of the aid project or program (Jacoby, 1966:132). For example, aid to Taiwan was administered through the extraordinary device of an autonomous institution rather than the national budget (1966:222). This served to build consensus around aid goals and reduced temptations to spend money for political purposes, i. e. military expansion. The BRC model seeks to answer the prior question of where such agreement comes from and whether it can be created? As suggested by Figure 3.1, a substantial part of donorrecipient agreement derives from realistic communication and trust that is less a product of accident than constructive conflict over time. Where roles remain “mixed”, rules and power imbalances inhibit effective planning and execution of development projects. This jeopardizes attainment of security, profitability and developmental objectives. As in Latin America, US aid has contributed to three related effects. Program complexity and frequent irrelevance to need, e. g. trucks with automatic transmissions for the Ecuadorian sierra region, are often a product of rigid rule interpretation by AID in the volatile context of Congressional and State Department oversight. Dependency also results between the US and its recipients, which tends to produce mutual losses and unintended consequences. For example, many aid projects suffer from over-complexity and bureaucratic topheaviness. High staff to project ratios are due in part to the foreign aid bureaucracy’s emphasis on capital over recurrent projects and foreign exchange-only financing (McNeil, 1981:79). The institutional incentive is to build-in administrative overkill in the form of short-long term consultants and Mission specialists to score high on preset “productivity” measures, such as visits, contact hours, etc. Even such “bottom-up” sounding projects as “agroforestry” end up utilizing the “Beltway Bandit” network for planning and implementation. Wolf (cited in Lyden and Miller, 1982:106) described the “internalities”, or incentives for social decision-makers to include private or organizational costs and benefits in their decision calculi, which stimulate “hi-tech”, “sophisticated”, and “modern” solutions. He notes that in the Vietnam War, excessive zeal for the complex and novel, together with mindless opposition to what was simple and familiar, encouraged use of jet fighters instead of more efficacious modified propeller-driven cargo aircraft (1982:106). While he suggests that governments tend to “operate through large organizations using blunt instruments”, such as FAO using 80% of its budget for staff costs in Rome, it is also true that US aid is implemented by private contractors. They also try to maximize sophistication and staff at the expense of simplicity and recipient needs. This means that every incentive is to interpret the rules to maximize profits/look good for Congress next

US Aid to Asia

129

fiscal year, rather than take risks for long-term development. For instance, placing a few anthropologists at several possible project sites for several years before project design (preprogramming) would not fit existing categories and would be viewed by both AID and Congress as either pure research or a lark. Hence, the tendency is to transfer US urban bureaucratics to Third World rural development contexts in the form of top-heavy projects. Projects to increase farmer productivity, for example, produce trainees but also cold storage units, trucks and other capital equipment which maximize performance on measures of: sites visited, farmers trained, seeds planted, and perhaps even crops harvested. But such short-term measures tend to ignore the frequent spread of landlessness, owing frequently to the fact that small farmers do not own enough land for economic results. But, the US has also planned and executed rather simple and highly effective economic aid projects. In Vietnam, a $20, 000 Tilapia fish project (“miracle fish” from Africa transplanted to Asia) helped many Indochinese peasants by 1952. But North Vietnamese propaganda that eating the fish produced leprosy scared many away and cut deeply into program effectiveness (Montgomery, 1962:39). As suggested, dependency is an almost predictable result of aid across donor-recipient bureaucracies. It was noted in the last chapter that the continuing dependency of many Central American nations on US aid serves as a powerful argument for use of the “dependency” perspective in that region. However, aid results have been different enough in Asia to raise the question whether dependency is not a transitional phase, maintained more by planning and execution failures, than any deterministic relationship with capitalist aid programs? The familiar litany of criticism can be cited for historic “aid” results in Vietnam and the Philippines. But having catalogued them, it becomes apparent that something more is operative here—positive results based on mutual understanding between donor and recipient. Decades of military and economic aid to South Vietnam did little to leverage US objectives and produced dependency in a variety of forms. Le Hoang Trong (1975:281), for example, argued that dependency was created by administration of aid, and that its termination under these circumstances would collapse both “military and economic fronts. “He noted that the US poured billions of dollars into South Vietnam in non-military aid (1964–74) without any Vietnamese plan for its use. Most of the aid financed consumer goods and ended up in the “pockets of the powerful in Saigon and abroad.” (That little has been learned about funding questionable causes with loose controls is indicated by the current case of contra fund abuse. According to McCarthy (Washington Post, July 6, 1986), “Less than half of the $27 million in aid meant for food, clothing, and medicines has gone into the contra jungle camps. Instead, it has been tracked to secret bank accounts in the Cayman Islands, to the Honduran army or individuals or firms that the GAO, in generous restraint, said ‘do not appear to be suppliers in the region.’”) In South Vietnam, only a small percentage of US economic aid was invested in production and development. With US aid, South Vietnam “almost exclusively consumed without producing much of anything—except war” (1975:294). South Vietnam became dependent on aid which diminished local incentive to produce. The Commodity Import Program (CIP) and Food For Peace (PL 480) programs provided cheaper and better imports at lower prices than local commodities such as rice. So, a major rice exporter after World

130

The Politics of United States Foreign Aid

War II, South Vietnam imported rice from the US at the rate of 300, 500 metric tons in 1974 (Le Hoang Trong, 1975:295). Rice production was also affected by the “Green Revolution” which introduced “miracle rice” (R8) to South Vietnam. Though it tripled yield per hectare, it required large quantities of imported fertilizers and insecticides. This increased dependency on US fertilizer imports which raised food staple prices and made it economically unavailable to non-farmers. Similarly, South Vietnam raised imported US chickens and had to import costly chicken feed as well. US aid-financed imports dominated all fields of Vietnamese industrial production, such as sugar, condensed milk, paper, textiles, clocks and bicycles. This widened the gap between rich and poor “upon which Communism thrives” (1975:297). More importantly, in this context, the SVN army thrived on hardware but was dependent on US aid for operations and maintenance, spare parts, ammo, and training (1975:298). Unfortunately, such “dependency” analyses often contradict themselves in the solutions section, where either the break with world capitalism is advocated (leading to dependency on communist institutions), or almost incredibly, more aid is recommendeded to solve problems caused by prior aid. The Hoang Trong analysis does the latter. But the problem is that dependency tends to be transitional, not an intended end state. It should be correctible in mid-course like any other program failure, without jettisoning the entire system. Montgomery (1962:88), for example, noted that the Commercial Import Program was designed to encourage large-scale private expansion in Vietnam. But capital goods over $500, 000 in value could not be imported without special permission from Washington. He called this an “administrative rigidity” that kept South Vietnam dependent on imports because they lacked the aid-financed capital to substitute them. Put another way, the aid that could have reduced dependency significantly was being wasted on import financing. South Vietnamese dependency was unfortunate from a developmental perspective. But the problems created could not have been resolved simply by (1) cutting off aid, (2) never giving aid in the first place, or (3) increasing aid prior to the North Vietnamese victory. Reduction of US aid brought the SVN economy to its knees by decreasing purchasing power for US imports. But the dependency was created by aid for “consumption and defense” rather than for “development and self-reliance” (1975:281). Nor could the US overcome the ideological-security draw of “communist containment” and “domino theories” to avoid following the French in after 1954. US foreign aid institutions, then and now, favored defense advocacy regardless of recipient needs or interests, and this insured a large “aid” program controlled and defined by defense (DOD) purposes. Serious opposition to SVN defense support did not permeate American public opinion and shift the balance of power from defense advocacy to consensus for total withdrawal for 10 more years. Le Hoang Trong argued reasonably for more aid (soft loans instead of grants) to build up import substitution capacity and end dependency on US imports. But this meant that imported fertilizer dependency would be ended by US aid loans for construction of fertilizer factories in SVN. This would simply shift dependency to a different account that would still require loan repayment. It also ignores the fact that the government of SVN in 1973– 74 lacked absorptive capacity to spend even the existing pipeline funds. Hoang Trong (1975:297) argued for more aid but also noted that SVN couldn’t use the funds it had! Like many before him writing in the midst of a political tragedy-a war-torn nation lacking a

US Aid to Asia

131

viable center, fending off left and right extremists, overwhelmed with US dollars, but with little control over their use—he argued for a “more creative approach to aid” (1975:297). But this was not to be. The US poured in aid money and support for a non-existent political center, occupied by the authoritarian regimes of Diem and later Thieu-Ky. This costly blunder (a US fear of using leverage from even greater unintended consequences) affected intangibles such as military morale and civilian support, rendering useless any attempt to carry on even the most primitive policy analysis. For example, assessing the “relative costs and gains to the US from using aid to affect (military) capabilities.. to respond to local aggression” (Wolf, 1960:286), at this point, was an exercise in futility. The problems of using aid properly stemmed not from absence of the right budgetary techniques such as “marginal utility analysis” (Lewis, cited in Lyden and Miller, 1982:264) or cost-benefit analysis. Rather, aid was going awry from failure to recognize initially the institutional constraints which stemmed from prior lack of agreement on objectives by the US and SVN governments. This stemmed from lack of clear guardian and spender roles and extreme power imbalances which blocked effective communication and the development of trust. US aid provided little leverage for the US in SVN. Advice from Washington was either too late or doctrainaire to be acceptable to Diem; advice from local US advisors was often directed at other advisors to avoid the stigma of intervention in local affairs. The result was that Diem was shielded from US advice and control but was maintained by force (symbolic and real) of US support. In economic aid, for example, Diem often made his own choices, which however costly or absurd, the US followed. In one instance, Diem selected refugee resettlement villages without fertile soil or water. Despite US technical efforts to suspend support or withhold procurement funds for this purpose, Diem prevailed each time. According to Montgomery (1986:71), since aid technicians couldn’t (and would not) justify the resettlement communities economically, the funds were charged to military “supporting assistance” funds. (DOD appropriations have three-year obligation authority which provides greater “flexibility” for their use than annual DA obligation authority). “But no such accounting device protected the settlers themselves” (1986:71). Hence, US aid to Asia produced both economic and military dependency in the short and medium term. Regimes in South Korea, China (ROC), Vietnam (SVN) and the Philippines received large amounts of US aid after World War II (see Figure 1.2). Sylvan (1976) even correleates increases in US military aid (1946–70) to 15 Asian countries with increased propensity to engage in military conflicts. He concluded that among other things, military aid decreases risk-aversion to military conflict as a solution to conflict resolution (1976). But, where aid is terminated, either from agreed upon successes (ROC), or after military defeat (SVN), dependency on US aid is diminished substantially because role clarity is increased. For example, after termination of Burmese aid in 1953 because of growing hostility and suspicion of US efforts which were perceived to compromise their neutrality in Southeast Asia, Burma asked for reinstatement in 1957. Understanding and aid effectiveness increased, although Burmese dependence on US for capital grants reached 82% by 1960 (Montgomery, 1962:31). Conversely, continuation of aid without agreed-upon political objectives perpetuates harmful dependency and can reduce both growth and development. AID influence was “generally beneficient” upon the “formation of Chinese economic policies” (Jacoby,

132

The Politics of United States Foreign Aid

1966:132). But prior Japanese concessional development financing in the 1890s and early 1900s stimulated high rates of capital formation and large export surpluses simultaneously (1966:76). This early form of foreign aid lasted only for about 10 years. US foreign aid after World War II built upon this foundation and the existing powerful Chinese interest in economic growth (but not democratic political development). Dependency was minimal and growth spectacular. Similar kinds of results were achieved in South Korea after World War II. With massive US military and economic aid, South Korea attained an average of 6.9% per capita real GNP growth rate between 1960–78 (USAID, FY 82 Congressional Presentation, Asia: 162). On the other hand, by cutting off support for Vietnamese and Philippine regimes, whose leaders used US aid primarily to solidify their own power at the expense of their people’s welfare, the US may have strengthened its influence in Asia. The end of the war cost North Vietnam $l billion in aid from China (PRC) and SVN about the same amount from the US (Becker, Washington Post, February 27, 1983). However, the two “donors” have been replaced functionally by annual Soviet aid for the same amount ($2 billion) (Gelb, New York Times, April 18, 1985). Dependency has ended in both the Philippines and Vietnam but, so far, growth has not resumed. The final effect of US aid in Asia has been a host of unintended consequences. Massive aid has been associated with military commitments in all four Asian countries, as well as Thailand, where military aid in FY 84 amounted to 70.6% of all US aid obligations (USAID, 1984:82). Both “security first” and “legitimacy first” schools combined to generate increased amounts of both kinds of aid. Where the military aid has been targeted and controlled, as in China and Korea, the effects of economic aid have been impressive. In both countries, trade and investment have largely replaced US economic aid—though not military grants. Conversely, where the military aid was provided as leverage payment to secure bases or regimes, as in Vietnam and the Philippines, the economic aid has not been effective in overcoming constraints to growth. Rather, the military assistance simply encouraged more conflict (first against domestic opponents then external threats) (Sylvan, 1976) instead of compromise. In the economic sphere, successes often paradoxically led to failure. As in El Salvador (discussed in the last chapter) the extreme right and left are both suspicious of economic benefits bestowed by external aid to a centrist regime. In the competition for allegiance which characterizes guerrilla war, successful DA or PL 480 projects (buttressed by ESF aid to keep the regime afloat) can swing beneficiaries in favor of government programs. As noted, the line between military and economic aid is often arbitrary since both programs add to the recipient’s resources, thereby freeing revenue and foreign exchange for other uses (Kaplan, 1967:283). But the “uses” often affect development in unintended ways. For example, National Institute of Administration trainees in Vietnam were successful in administering rural development projects, for which they were often eliminated as “elites” by the Vietcong who were fearful that their work would generate support for Diem. Similarly, according to Montgomery (1986:59) “The more we were able to improve conditions in the South of that divided country, the more the Vietcong would exploit the discontents of development by recruiting those whose expectations had risen too fast. “Success in rural development led to failure through the generation of extremist political blackmail.

US Aid to Asia

133

Finally, aid success also produced “failure” in generating competition with US trade and industry. Eventually, after passing through a “dependency stage”, aid can build commodity, textile, leather and other industries, whose products will predictably compete with US producers. In this fashion, aid generates its own political opposition in the US while its beneficiaries have little influence on US policy-making. Following the example of Japan, Korea and Taiwan are competing to become the “2nd Japan”, flooding US markets with exports in high-tech, autos, steel and textiles while jealously protecting their own domestic markets (Auerbach, Washington Post, February 9, 1986). These countries are “graduates” of the US aid program like Brazil, whose many “profitable” individual loans-grants that are argued before Congress each fiscal year, turn out to be mostly profitable for the recipients. Though US farmers, textile workers, steel technicians, and auto manufacturers would disagree, the aid has been profitable in a more abstract realm of “high politics”: these are thriving capitalist economies now, when in the immediate post World War II era, their chances of being communist satellites, economic deadbeats, or both, were much higher than today because of US aid. As USAID tries to expand its political base of support, such as by channelling foreign students through US universities, it ironically loses support where its successes are greatest—in spreading US “knowhow” to Asia!

Constraints Why the goals of foreign aid translate into successes and failures relates to systemic features of the US policy process. As indicated, the process is fragmented across a variety of agencies, interest groups and Congress. But in practice, absence of total program group support as well as consistent influence by Congress or USAID, shifts power rather permanently to the President, DOD and State Department. To the extent that these actors have a consistent decision premise, they have inordinate influence on policy-making if not results. Put another way, the foreign aid policy arena is a shifting coalition of executive branch actors that try to bargain and maximize their relative positions of power through the annual budget process. Without a coordinating actor or forum such as Congress or State, the process devolves authority to the most powerful actors in the coalition, e. g. DOD and the Presidency. Coordinating mechanisms in State (SAPRWG and Integrated Foreign Aid Requests) and in Congress (the appropriations process) tend to exclude values because of the significant power imbalances among bureaucratic actors and seem to change foreign aid allocations between economic and military programs very little. The procedures which control programming, execution and feedback from foreign aid outlays are affected by the clarity of agreements worked out between US bureaucratic actors and counterparts in recipient countries. This means that the clarity of institutional guardian and spender roles and the substance of conflict between them (opposing perspectives may be irrelevant for policy-making if they are not brought together in direct conflict) determines the scope and purpose of the foreign aid program. Here, we examine the rule constraints as they affect the program in Asia. Much has been written on the way foreign policy institutions acted to structure US choices in the Vietnam War. Much of this is applicable to US foreign aid policy-making in Asia as well. The thrust of much of this literature is that, acting to maximize institutional, personal and national power, the bureaucracy selectively distorted information on the

134

The Politics of United States Foreign Aid

benefits of disengagement or de-escalation and de-emphasized the costs of escalation (Gurtov, 1974:145). Dominated by Pentagon hard-liners at the time, bureaucratic pressures stressed action and the “exploitation of US power and military potential” (1974:155). Gurtov suggests that problems of duplicated effort and co-optation of policy dissidents were but “symptoms of a ‘disease’ indigenous to the decision-making process” (1974:155). It is argued here that the ‘disease’consists of imbalanced and ill-defined institutional roles which distort information and policy choices. In Vietnam, the foreign policy bureaucracy constantly limited the President’s choices to different kinds of escalation. (1974:156). Role definition and realistic conflict seem to cure the disease, leading to the strong likelihood of both economic and political development via aid in El Salvador, Haiti and the Philippines. However, in the 1970s the basis for aid leverage was almost purely military (communist containment) and the level of realistic communication between the US and recipients in Asia was low in most cases. Bureaucracies produce “decisions” and the more quantitatively certain they can be made, the greater the possible consensus since all can rationalize the action firmly. Hence “body-counts” prevailed over negotiations; status quo over cuttingoff SVN autocrats. By contrast, where communication was clear and realistic as in Taiwan, growth and some development took place. For example, Pentagon hubris throughout the Vietnam War produced military grants (MAP) almost without control. The result was lost leverage over both the course of the war and political development of Vietnam. Certainly a war is not a good time for thoughtful long-range aid programming. But anticipation of the severity of mistakes by experienced aid technicians would seem to dictate caution and conservatism, i. e. control of the funds consistent with the “profitability” goal noted above. That this did not happen suggests deeper problems than surface “coordination”. These deeper problems relate to role definition and realistic conflict by actors with vast differences in political and budgetary support. Two other related programming practices that constrain foreign aid effectiveness are: (1) congressional restrictions on aid use, and (2) USAID rules which seek certainty at the price of developmental results. “Red tape” is endemic to all intergovernmental loangrant programs; the necessity for control of expenditures likely to be inconsistent with legislative intent and the need to stimulate appropriate activities require proper grantor incentive structures. Foreign aid is not qualitatively different from domestic programs in this regard. But quantitative differences arise because of the immense complexity of the relatively small foreign aid program—many program authorities in addition to USAID; many types of assistance (FMSC, ESF, DA, etc.) and many different types of recipient political cultures. The chance of failure is quite high under such conditions and it is here that rules and repertoires can make a difference. But Congress tends to leave the decision structure intact after each experience. For instance, the New Directions legislation of 1973 was designed to redirect AID efforts toward the problems of chronic poverty and powerlessness. The “natural tendencies in the executive branch” (Montgomery, 1986:95) that directed funding through the largest recipient ministries and often toward narrow middle class beneficiaries were to be restrained by adding on new requirements, such as Foreign Assistance Act Sections 102b, 111, 113, and 281a to direct DA toward New Directions objectives. Similarly, military failures in Vietnam brought the War Powers Act of 1973, “a Quixotic attempt to restrain future presidents from using military power to bolster their image at home” (1986:95). But,

US Aid to Asia

135

though neither law would “affect the tendency to use power, whether military or economic, to assist the already powerful” (1986:95), the institutional effect was to increase AID staff time in trying to comply with the letter if not the spirit of the law. The congressional tendency to leave the decision-making structure intact and addon marginal requirements can be noted each fiscal year. Congress finds foreign aid a contentious, politically unpopular issue, despite its relatively small size (1.5% of the FY 86 US budget or $16.0 billion out of $1.05 trillion requested). It therefore falls easily into partisan squabbles which reduce chances of passing either foreign aid authorizations (1986 was the first one since 1981) or appropriations bills (none since 1981 which means “continuing apporpriations resolutions”). Part of this contentiousness, and the perceived powerlessness of Congress to control foreign affairs, is expressed in the many amendments and conditions added to the bills. The FY 86 authorization bill for $12.6 billion which finally cleared both chambers contained more than 60 amendments, including anti-family planning provisions for conservatives, more DA for liberals, and “freedom fighter” aid for Nicaraguan, Afghan and Cambodian “contras” (Congressional Quarterly, July 13, 1985:1359). The subsequent appropriations bill markup session in the House Appropriations Subcommittee on Foreign Operations added 30 more amendments, including aid prohibitions to countries where airport security against terrorists is lax, and the requirement that military aid to El Salvador be controlled in semiannual reports to Congress on success in curbing rightist death squads (CQ, July 20, 1985:1424). Individually the conditions are valuable contributions to civilized politics; in aggregate they serve as a blueprint for failure by constraining a small agency (USAID) charged with implementing a key portion of foreign policy. In Vietnam, AID was charged with administering many “politically inspired conditions on the use of American aid” such as one-half of commodity imports in US “bottoms” and facilitation of bids on commodity imports by US small businesses (Montgomery, 1986:74). USAID, like its Third World counterparts, lacks absorptive capacity to interpret these rules consistently for effective program implementation. Since Congress provides the money and USAID lacks a consistent base of client political support, AID attempts to comply with these myriad rules in ways that, as noted, displace larger goals. Development of rules to satisfy Congress (New Directions, environmental, accounting concerns) served also as a defense mechanism to parry outside criticism. Compliance became easier than “developing” recipient societies, and provided tangible benefits in support by State and Congress for increased AID appropriations. Congressional rules then contributed to “goal displacement” by an agency that is more committed to bureaucratic efficiency and management than the flexibility, and innovation required for the elusive task of development. Some mistake cause and effect in blaming types of AID people or roles for the result. Gurtov (1974:206n), for example, distinguished area specialists from crisis managers that are more interested in gamesmanship. But AID rotates its foreign service personnel from Washington to Missions, meaning that roles and perspectives become mixed over time. AID programming, as noted, is rigid and based on approval of detailed project blueprints in Washington. The approval process for CDSSs becomes more important in this “relentless search for certainty” (Montgomery, 1986:92) than the results of AID operations. Even mid-course corrections in field projects, necessary for managerial flexibility, are often

136

The Politics of United States Foreign Aid

viewed by AID as threats after staff time and political capital are expended to get the project approved. Thus, Montgomery (1986:91) notes that AID did not want a procedure to incorporate information about “second-generation” effects of the Mali Livestock II project in Africa. Because of multiple pressures to program precisely in a highly uncertain recipient context, USAID staff time is “displaced” from concern with project implementation to project design and approval. An additional consequence of this kind of programming is that AID has every incentive to rely on “large, safe, generally accepted project designs” (Montgomery, 1986:93). We have noted that AID budgets often reflect these narrow and often contradictory programming pressures. Part of the problem is that AID budgeting is tied to congressional disincentives for effective budget policy-making. That is, Congress budgets foreign aid from a variety of political pressures, the most prevalent of which are DOD and the President with State Department advice. AID requests for DA are detailed, specific and well-substantiated in its annual “Congressional Presentations”. But, as noted, annual or biennial reauthorization time is spent not on analysis of prior aid results or on learning of specific country course corrections needed to improve aid results. Rather, authorization hearings are mostly an opportunity to engage in ideological abstractions linked usually to regional “hot spots” such as the Philippines. The other programs (the established “base”) are not examined in traditional incremental budgetary fashion because there is no time or resources for the immense task—largely because they are wasted on the wrong issues. AID budget formats moved from the object of expenditure to PPBS in the 1970s. But given the resistance to both accounting for results within the agency and “crosswalking” such information to Congress, AID now practices a mixture of line-item and “performance” type budgeting. This permits Congress and “significant others” to use either type of information. AID budgets use “programs” and “projects” as the functional equivalent of clusters of line-items. Thus, AID recognizes that Congress will award budget authority not for the complexities of implementation, but rather for good plans in line-items on which trades can be made, conditions attached, etc. Plans are politically safer than results. It is only logical then that AID programming would not provide the “time or climate conducive to an objective appraisal of actual experience with a view toward identifying and transferring lessons learned” (Packenham, 1973:119). Budget approval time is not the season to display one’s dirty laundry—consensus and impressive plans bring the fiscal results, not arcane questions about development! Mistakes are often recycled into the budget and submerged under broad existing program labels. Under standard US policy-making incentives, AID managers and program people are encouraged to sound authoritative, present slick charts and visual documentation, and obligate the money by the end of the fiscal year. Again, this is no different from any other US program area. On the other hand, US foreign aid programming and budgeting is affected to some extent by recipient behavior. Caiden (1985:36) notes that budgetary systems of ASEAN countries are “not easily described by existing comparative models” since they “do not conform to the model of poor countries nor do they exemplify normative models of the literature. “Here again, the US experience of donor control and bias of recipient policy choices by intergovernmental grant programs may be a better descriptor of what foreign aid does to local budgets and conversely how recipient “receptivity” to donor projects encourages more

US Aid to Asia

137

US programming for them. McNeil (1981:20) notes how foreign-financed donor projects have a better chance of being included in the local budgets “against possibly better projects which are locally-funded. “Problems of local coordination between the operating budget and the “development” or capital budget in Malaysia, for example, (Caiden, 1985:34) may be affected by the programming and fungibility problems related to the $11 million US aid program there (USAID, 1984:78). Since most recipient budgets are executive branch-dominated processes (unsullied by legislative interference), administrative advocacy for “status” foreign aid projects is also operative and this distorts the use of aid for national needs. This kind of “political capital trap” also works against programmatic and budgetary flexibility by AID. For instance, disagreements between US and SVN officials over the scope and purpose of the Vietnamese Land Development Program resulted in US withdrawal of funds for the project. The US sought agricultural development and settlement of farmers on 3–5 hectare plots; Diem wanted security objectives to control and no more than 1 hectare per farmer. In 1957, the US withdrew funds. But since the congressional appropriations process had already fixed the dollar amount for the Vietnam “program”, the Diem government simply transferred the funds to other aid operations. The ability of recipients to practice such countercontrol techniques continues today and distorts a foreign aid program already politically disfigured before it leaves Washington for the field. During the execution phase or “fiscal year” in most cases for DA (many DOD programs have 3 year obligational authority), existing congressional rules tend to work against developmental results. In the Vietnam War, for example, absorptive capacity of both donor and recipients was a serious constraint to effectiveness. Several billion dollars flowed to Vietnam with either no strings or too many attached. Despite vigorous efforts to control aid use by attaching more statutory requirements, Congress provides funds to the US President on a discretionary basis. Under the Foreign Assistance Act and Defense Appropriations Acts, Congress established “contingency and emergency funds” simply because “neither Congress nor the President can foresee all conceivable circumstances that might arise in a given fiscal year” (Nathan and Oliver, 1983:125). Among the discretionary budget techniques available to the President (and foreign aid officals) are: transfer authority, reprogramming, control of excess stocks, carryover and pipeline funds. The exercise of these powers suggests that a “fungibility” control problem exists at both ends of the foreign aid transaction (donor and recipient) and that statutory remedies often multiply loopholes through which funds can be re-allocated after congressional appropriations had apparently settled the matter. While most of the egregious examples of what were varying degrees of “impoundment” occurred during the Vietnam War, many contemporary examples can be found in country programs, particularly El Salvador, and the Philippines. So “transfer of excess stocks” (within the security assistance program) has been used commonly by CIA to conduct covert operations or continue supplies of “aid” to governments where Congress has reduced military assistance. In FY 74, for instance, $450 million in local currencies generated by the Food for Peace program were “transferred” to military accounts in Vietnam and Cambodia. In FY 73, 60% of the DOD budget ($43 billion then) remained unobligated; much of this was carried over to the next fiscal year and “transferred” between appropriations accounts to bomb Cambodia. It should also be noted that carryover funds do not have to be voted

138

The Politics of United States Foreign Aid

on the floor of Congress. This is one more way in which Congress acquiesces in executive foreign policy-making. Further, overestimating budget authority needs during the fiscal year is a ploy that can add to individual agency contingency fund surpluses. When Congress refused to appropriate any more aid funds for the falling Cambodian government of Lol Nol in 1975, DOD miraculously saved the day by finding extra funds from “overanticipated” inflation effects on the ammunition line-item (Nathan and Oliver, 1983:127). As the Saigon regime was collapsing in March, 1975, the Ford Administration sought $300 million more in “supplemental appropriations” for Vietnamese foreign aid. But it was found that only $178 million of the FY 75 appropriation of $750 million had even been obligated! The Administration was effectively seeking to increase its “pipeline” funds for “later emergencies” by about $500 million (1983:128). In 1985, the House Intelligence Committee held that the CIA must have reprogramming and transfer authority requests approved by it for use of any part of the $27 million in approved aid to the Nicaraguan “contras” (CQ, November 23, 1985:2451). Nevertheless, as noted, less than 50% of this “aid” has served non-lethal purposes. Instead, most of it ended up in secret bank accounts including the Honduran army (McCarthy, Washington Post, July 6, 1986). The appropriations process is still easily circumvented and this fragments both programming and budgeting of future aid projects even further.

Role Conflict We have discussed the effects of goals and rule constraints on foreign aid results. Let us now examine how the entire process, from program to results, is a product of the incentives and disincentives provided by the role behavior of bureaucratic actors in conflict over issues of foreign aid. It has been recognized that foreign aid is unique to American politics as an unpopular program with lack of domestic support or national consensus on its political, economic or military objectives. A host of US institutions are responsible for foreign aid under State Department coordination but intense turf battles still persist over specific policy responsibilities and funding. Ostensible narrowing of agency differences indicated by request and appropriations patterns in Table 3.1 may have moved differences to lower organizational levels; they may be merged and screened out by the level of aggregation achieved in the “Integrated Request”. Clearly, further research on mid-level foreign aid decision premises and budgetary priorities is necessary. But presently, neither advocates for more aid (USAID, DOD) nor guardians (Congress and State) have a clear picture of the interrelationships between the multiple foreign aid subprograms. Rather, the important political issues for getting the program reauthorized and funded are narrow statutory requirements, e. g. US small business, environmental protection, and AID routines (CDSSs, ABSs) that tend to displace systemic concerns for recipient development. For FY 86, more than 200 items remained in disagreement among House and Senate conferees alone in the foreign aid authorizations bill. But most related to more contentious questions such as family planning (contraception and abortion) and aid to Nicaraguan contras, instead of the substantive details of programming and effectiveness of the individual projects as they applied to specific countries. Like the budget process itself, the authorization and

US Aid to Asia

139

appropriations processes are iterative and proceed on the basis of approving or modifying recommendations from prior stages, most of which amount to political compromises. Small mistakes are incrementally carried over year by year until a crisis (Marcos’ election fraud) either causes reexamination of some part of the program (support for Aquino) or it doesn’t (support for Somoza’s National Guard and the Sandinistas). Incremental policy-making is the basic descriptor of American politics. The model works well where all sides of a controversy are represented by aggressive advocates, information is available, and power is not permanently imbalanced. Incremental policy effectiveness thus depends on the structure of the forum. This means not simply the size or number of actors but the clarity of the roles and the likelihood that they can deal with the substantive issues before policy is recycled for another fiscal year. Montgomery (1962:104) in a classic study of the determinants of US foreign aid policies in Southeast Asia in the 1950s notes that the “Major achievements of US foreign aid have emerged out of genuine cooperation between aid missions and the governments receiving aid. “He suggests that the elusive quality of “mutuality” is critical and that “diplomatic enthusiasm” often exaggerates the “common purposes” of “cooperating countries” (1962:61). But tensions and misunderstandings between AID Missions and host counterparts are not qualitatively different than tensions and conflict between US policymaking actors over foreign aid authorizations and appropriations. A core assumption of this book is that compromises and consensus cannot simply be mandated but must arise out of a period of possibly wasteful and harmful conflict. To the extent that clear positions meet directly with clear responses by relevant institutions, the process should lead (see Figure 3.3) toward reduction of distrust and enhanced communication, which are the fundamental bases for exchange of knowledge and cooperation. Hence, recipient and donor conflicts must be viewed as part of the same problem of designing institutional incentive structures to squarely face and resolve conflicts. In this framework, the confusion over foreign aid goals, as well as rules and repertoires which encourage such confusion, are less important for improved policy-making than the “ecology” of role conflict in which they function. Changing the structure of conflict, for example by providing USAID additional authority to set its own budgetary ceilings (now set by State) and country priorities (now established by State with major voice exercised by DOD), could change incentives to develop more innovative approaches to aid. In this section, we will focus first on the types of conflicts between US and recipient that has impeded results. We then turn to the problem of how destructive and superficial conflicts have impeded aid results in Asia, like Latin America. US aid projects are often affected by the application of US rules arising out of domestic political concerns. These are sources of distrust and superficial conflict. Not uncommonly, US aid technicians apply American standards to recipient aid projects. For example, aid is often conditional on reform of recipient institutions, such as tax and fiscal reforms, and personnel administration practices. But such countries as Thailand, Taiwan, and Vietnam showed no interest in personnel reforms because political considerations prevailed (Montgomery, 1962:119) and still do. Nor do recipients become ecstatic at the thought of achieving administrative efficacy via bringing “off-budget” agencies and government enterprises on budget, or at least under direct Treasury department control. Caiden (1985:33) noted this counter-control tendency in Malaysia. For example, one wonders how eager

140

The Politics of United States Foreign Aid

Uruguay would be to receive aid for more centralized control of its public finances, where the “public sector” is already subclassified into categories like “central administration” (also called “central government” which includes autonomous agencies and the City of Montevideo), and “general government” (which includes the “nonfinancial public sector” which also includes state enterprises) (IMF, 1985:83). Such minute distinctions are not accidental and reflect deeper power divisions within their government (if “government” means all of the public sector!). Part of the problem of reducing conflict and increasing understanding is the professional gulf between State diplomacy and high politics and AID technical assistance or lower order activities. Diplomatic pressure to conform to an aid condition is often hard to muster because “Few ambassadors are prepared to take a positive stand on a technical issue” (Montgomery, 1962:118). Those that do such as Robert White’s stand on land reform in El Salvador find themselves on various “hit lists”, including those of US political opponents. Similarly, lack of political consensus in the US on the scope, purpose, or results of foreign aid, permits attachment of many conditions to aid motivated largely by domestic politics. US concern for financial profitability, women’s rights, environmental protection and human rights often translate abroad into resentment and misunderstanding when forced on aid recipients. As indicated, such statutory requirements are explicable in part by congressional frustration at inability to do anything really substantive under present policy-making structures. More importantly, recipients and AID simply disagree and no forum exists for conflict resolution (they exist ostensibly during project design but not execution) to the ultimate detriment of both US objectives and recipient needs in the programmed area. For example, in a telling early conflict over the meaning of the “internal security” that was supposed to be provided by military aid to Vietnam, President Diem defined it as aid to his Civil Guard to become a loyal “paramilitary” component of his army. The US position was that aid should be used to develop an autonomous police force (Montgomery, 1962:68) that would decentralize its operations into “village security” (1962:69). US aid technicians had an early glimpse into how the US security objective would be compromised by mis-directed military aid. Political security for the puppet regime against legitimate domestic opponents became more important than national security against subversion and terrorism. The US was largely dragged into this “reverse ward” relationship by providing no enforceable mechanism by which aid technicians could muster enough institutional power to overrule the rest of the system. That US aid inevitably was tied to an authoritarian regime with little interest except its own survival generated high opportunity costs that were simply not payable later. For instance, the Vietnamese Civic Action Program originated in 1954–55 out of a military effort to rebuild war-damaged public facilities (1962:70). The successful local program was later expanded to distribution of seeds and medical supplies to rural areas (an early “pacification” program). But Vietnamese public ministries protested alleged infringements of their responsibilities by the program. The US nevertheless continued to provide aid in 1956 that was channelled not through the CAP program but across the ministries. The program was thus effectively destroyed by “bureaucratic rivalries” (1962:70) and an excellent opportunity to gain early support from the countryside was lost.

US Aid to Asia

141

But the biggest disappointment caused by ill-defined roles and resultant distrust was the Vietnamese Land Development Program. In 1954–55, roughly 300, 000 refugees from North Vietnam were settled on abandoned land tracts. The US wanted to use the remaining millions of hectares of rich, idle farmland for distribution to peasants for agricultural development on plots of 3–5 hectares. But Diem wanted the Land Development Program to be for security purposes, conceived as a “wall of humanity’ along the borders (1962:73) with plots of no more than 1 hectare each. Later Diem used the Land Development facilities as “re-education” centers and “barracks” for political prisoners. The US opposed these actions and withdrew support. But, as noted several times, the foreign aid budget process had already fixed the amount available for obligation, and “withdrawal” amounted to no more than a Vietnamese fund transfer to other projects (1962:72) with presumably similar kinds of objectives. The ultimate result, of course, was that some agricultural production occurred with little military security since the whole idea was “uneconomic” and merely intensified rural dissatisfaction with the Diem regime (also a large opportunity cost, payable in full, Spring, 1975). Such examples suggest the futility and frustration of aid technicians providing local advice to deaf ears, who are also unable to exact real compromises from either US foreign aid policy-makers (because Congress and DOD are miles above them in the “staff” world) or local policy-makers. Individuals and institutions often behave unpredictably in such situations. Institutions exhibit donor fatigue and either “withdraw” fecklessly or pour money in indifferently. Individuals follow suit. One AID colleague, as noted, went AWOL from the Philippine AID Mission and conducted his own agroforestry program! It follows that such conflicts or standoffs are intimately related to the degree of trust and understanding of foreign aid by the larger actors which make and approve aid policies. Under current practices, staff time is devoted largely to the superficial tasks of debating “riders” to foreign aid bills, programming details into revised CDSSs, and budgeting longterm projects over the next 12 months in a tightly constrained bureaucratic environment where little can easily be changed once approval is gained. These types of control-oriented practices are effects of prior mistrust between institutional actors that ultimately derives from the ambiguity of total foreign aid objectives and results. Let us review some of the larger institutional problems, referring where possible to the more recent and dramatic case of US-Philippine aid. As in other country programs, the US attempted to achieve stability, profitability and development through manipulation of the terms and proportions of economic and military aid in the Philippines. Over the Foreign Assistance Act period (1962–84), military aid amounted to 35.9% ($828.8 million) and economic aid 63.9% ($ L47 billion) of the total ($2.3 billion). These proportions have remained about constant from 1946 to 1984. (USAID, 1984:79). However, military aid increased to $54.7 million in FY 86 from $51.5 million in FY 84. President Reagan had requested double that amount ($102.8 million) for FY 87 even before Ferdinand Marcos fell from power (Fuerbringer, New York Times, February 26, 1986). The Reagan military request was based on a 1983 agreement with ex-President Marcos that committed the US to provide $900 million in aid from 1983–88 in return for the use of the 2 military bases. Though the lion’s share of the aid had been economic, the degree of trust or mutuality between the US and Marcos was never strong. Much of this aid had been “transferred” to military uses by Marcos’ banking institutions (the fungibility

142

The Politics of United States Foreign Aid

problem again) and some portion had contributed to real estate shopping sprees of his immediate family. Even in more precisely controlled areas governed by the IMF, such as the amount of Philippine foreign debt, it was discovered that prior to the assassination of Benigno Aquino in 1983 “the Filipino economy was going downhill” (Rowen, Washington Post, January 29, 1984). In 1984, IMF found that the Philippine Central Bank had overstated its reserves by $600 million! This revelation added to capital flight and to larger questions about the use of US foreign aid here. But throughout the Marcos years (President from 1965; Martial Law from 1972 to 1977; “constitutional authoritarian” President until 1986), despite technical dissents by USAID on rural development project design and execution, the bulk of US foreign aid actors played their budgetary policy roles, the effect of which was to continue US support for his regime. The story of US reversal of support tends to demonstrate our thesis that over time role clarity and more realistic conflict (better information) lead to greater institutional trust and better policy decisions. Most of the actors were institutions in both the US and Philippines. However, individual leaders such as Marcos critic, Representative Stephen Solarz (D-NY) (whose subcommittee voted 9–0 to suspend all military aid and channel all non-military aid through PVOs, Fuerbringer, New York Times, February 26, 1986), Marcos’ friend Daniel Inouye (D-Hawaii) who continued to insist Marcos was our friend and to “have patience” right until the end (CQ, November 16, 1985:2390), White House “troubleshooter” Philip Habib and Senator Paul Laxalt (R-Nevada), as well as the dramatic non-violent resistance of Corazon Aquino, contributed to the formation of US policy consensus for a shift of loyalties. The crisis precipitated locally by the Philippine rejection of Marcos served as notice to US institutions that their contrary impulses for agency reasons would produce collective disaster as a continued policy. Hence, Filipino pressures enabled “mutual disengagement” from Marcos to serve as a common bond. Given this “core understanding”, roles and agency pressures became more consistent and rational. Similar kinds of understandings were reached by different means in South Korea, Taiwan, Brazil, Venezuela, Mexico, and Costa Rica. The core understanding in each case was that the recipient would not compromise certain political values for the sake of economic gain and political dependency. In all cases, systemic learning by the US was precipitated by external crisis rather than shifts from processing its own information. But the ultimate learning occurred from the conflict between actors debating the substance of new information. The 1983 assassination of Aquino set off protests around the world that helped convince President Reagan to “drop the policy of uncritical support for Marcos” (CQ, November 9, 1985:2287). The policy conflict took place between continued advocates of providing more economic aid around Marcos’ institutions (USAID, State, Congress), and advocates of more military aid to appropriate factions within Marcos’ army to defend against the growing insurgency, and “guardians” of US long-term regional security and developmental interests (Congress, DOD). The growing economic and political crisis, dating from 1983, helped clarify positions and aided the policy process in part by providing clearer information to partisans of each position. International media attention and the congressional forum assured maintenance of the issue of aid to a legitimate Philippine government on the US policy agenda.

US Aid to Asia

143

While AID-State interests sought more economic aid, increasing distrust of Philippine institutions generated support for disbursement through PVOs and the private sector. This implied lack of support for Marcos. Even DOD, perennial advocate for more military aid, tailored its request to moderate elements in the military. That the “hard” school can produce “soft” results was again demonstrated by the critical “mutinies” of Defense Minister Juan Ponce Enrile and Lieutenant General Fidel Ramos who threw their support to Aquino and the the cause of “peaceful civil disobedience” which brought down Marcos. But the outcome might have been different (the US might have stayed with Marcos to the end) had not Congress served a recent but constant “guardian” role on the Philippine aid issue. Congress, as noted, does not examine the substance of aid by country unless a crisis or controversy exists (otherwise it is a weak guardian or mixed role) which usually means ex post policy guidance. Here, the crisis produced an agenda for 3 years during which public opinion was galvanized behind the right US decision for long-term security and developmental interests. The lengthy crisis clarified the congressional role as well as that of State, AID, and DOD. Beginning in FY 84, Congress mandated aid cuts despite Reagan Administration requests for more Philippine aid. So, in the FY 85 continuing appropriations resolution (CAR), Congress cut $45 million from the Reagan request and shifted it to economic aid (DA). In FY 86, Congress cut $30 million from the Reagan request of $100 million, leaving $70 million which Congress again shifted to DA (CQ, November 9, 1985:2288). Late; Congress informed Marcos via letter that aid would be ended if the 1986 elections were fraudulent (CQ, November 16, 1985:2390). At this stage, congressional Marcos ally Daniel Inouye dissented and asked for “patience” from his colleagues. However, Marcos followed by reinstating General Fabian Ver as armed forces chief, an incredible move to even US supporters given his questionable acquittal on the charge of contibuting to the assassination of Aquino in 1983. Congress then tied more retrictions to a reduced amount of military aid (CQ, December 21, 1985:2689). The paradox, as noted, is such a ploy actually gave Marcos more freedom to spend economic aid money. But AID responded by intensifying efforts to run economic aid around his ministries and through PVOs.

Conclusion Realistic conflict in the Philippine aid case increased mutual consensus by US actors, eventually including President Reagan himself, that Marcos should fall. According to Apple (New York Times, February 26, 1986) “Never before had the US pushed so publicly for the removal of a leader of a major ally. “Though he notes that Washington eventually conspired to secretly end the regimes of Rhee in South Korea, the Shah of Iran, Duvalier in Haiti, Somoza in Nicaragua, and Diem of Vietnam, the Philippine case was different because a middle class alternative existed (and still does if Aquino falls) to Marcos. The 1986 Philippine aid case suggests that where roles are clarified by lengthy institutional conflict, communication and trust improve which increases consensus for foreign aid policy-making. Future US provision of aid to the Aquino and subsequent Filipino governments will occur in an improved “ecology” of policy-making. US aid to Asia should gain in improved chances of producing more actual security and development. Foreign aid economic “success stories” where political structures have not conformed to US “liberal premises” are now on notice that the US may well support reformist but not

144

The Politics of United States Foreign Aid

revolutionary alternatives. South Korean President Chun had the Kim opponents jailed for attempting to petition South Korean voters for a presidential election. Recently, after the demise of Marcos, he had the Kims released. The “second Japan” may find that the road to Tokyo does not pass through Manila (New York Times, February 26, 1986). In the Philippines, to contrast with earlier US aid to Vietnam, “genuine cooperation” and “continuity of purpose” (Montgomery, 1962:104) were attained by bureaucratic role conflict. The lesson, of course, is that US institutions should be receptive to the same realistic debate on all recipients before crises occur. As indicated by institutional behavior in Iran and Nicaragua, the US ignores most crises until they are irreversible. Where no political necessity exists for a hard look before a hard choice, foreign aid institutions maximize their respective turfs, Congress votes authorizations and appropriations on a superficial uninformed basis, and foreign aid continues its course despite evidence of major shortcomings.

Chapter Six US AID TO THE MIDDLE EAST

Introduction We have seen that foreign aid efficacy depends largely on mutuality of aims among donor actors and between donor and recipient. In Asia and Latin America, the effects of US aid have been similar in most cases to throwing hats in raging floods. In some areas, the flood was going the US course for a time—Taiwan, South Korea, Costa Rica, Brazil, and Chile. In other areas, the current was not so strong that it could not be redirected by mutual agreement to alter the course of the flood—El Salvador, Philippines, and Haiti. Elsewhere, US aid simply reinforced the destructive effects of flooding by throwing in the hat, when most observers had agreed that it was going in the wrong direction—Somoza’s Nicaragua, Vietnam and Batista’s Cuba. Other viable alternatives existed in each case but were excluded by the US aid policy process. Success and failure turned on the decision-making capacity of US actors to correctly program and execute both economic and military aid. In most cases, the local political currents were rather clearly directed and the only question was whether US policy could adapt to them. Successes or failures to do so were a product of US institutional factors and to a lesser extent problems of understanding local political cultures and practices. In the Middle East, the problem of US aid efficacy is both more simple and more complex. It is more basic than aid elsewhere in the world because the overriding objective is security for NATO, European and US oil supplies and against Soviet interference with that continued supply (see Figure 4.1). Accordingly, 5 out of the top 6 US aid recipients for FY 87 were Middle Eastern countries (Israel, Egypt, Turkey, Pakistan, and Greece) and the bulk of their aid was security assistance (FMS/MAP financing) (Nowels, 1986:16). Development (DA) of already oil-rich countries, that even receive aid from OPEC itself, is clearly less important here to the US than in Latin America or Asia. But the problem for US aid, boiled down to its most tangible form of “political capital” for leverage purposes, is that legitimate authority structures are mostly absent in this region. Consensus, broad bases of legitimate political authority, even limited government and freedom of the press can be found in varying degrees amidst the turmoil and factional infighting of Asia and Latin America. In the Middle East, to continue the flood metaphor, aid is similar to throwing a hat into the bottom of a waterfall, with contradictory and powerful waves smashing it against the rocks. Bilateral aid is supposed to “buy” security leverage in a region characterized by centuries of distrust often between feudal warlords fired up by religious zeal. The Middle East is like Hardin’s “Tragedy of the Commond’ (1968) where many actors maximize their power via terror and coercion to the ultimate ruin of all. Absent usually is a sense of mutual restraint that can act for the benefit of the region or individual nation. Regional political, ethnic, and religious conflicts tend to DOI: 10.4324/9780203840184-6

146

The Politics of United States Foreign Aid

override consensus on governing institutions. Hence, the “prime beneficiary of foreign aid”, the recipient “government” (Asher, 1961:25) tends to be missing! Though it may be a difference of degree and not kind from Asia and Latin America, in some cases no central government exists at all (a “libertarian’s” dream!) with all this implies for the breakdown of civilized values and rule of law. Almost total absence of mutuality between donor-recipient suggests that aid will contribute to many unintended consequences (surprise successes and failures). The history of US foreign aid in this region is almost precisely that: the futile quest for security and stability via projects planned and implemented amidst volatile shifting factions on which little national consensus exists. US aid to the Middle East may be no more than a proverbial hat in a political waterfall. But, unfortunately, it is the largest of US foreign aid hats. As noted in Figure 1.2, the Middle East replaced Asia as the leading recipient of US aid around 1975, and by 1982 received around 50% of the total. In FY 84, the US spent or “obligated” 73.5% of all world-wide military aid in the Middle East ($4.76 billion) and 34.9% of its economic aid ($3.09 billion including ESF). This was absolutely the highest amount in both categories. Since much of Middle Eastern “economic” aid is ESF, most amounts to either cash transfers or capital projects, such as electricity and port development in Egypt. For example, US foreign aid now amounts to 19% of the Israeli budget (Meyer, Washington Post, December 15, 1985). Of the total USAID Middle East program for FY 82, 98% was ESF and only 2% for functional DA projects (USAID, FY 82 CP: 3). The more astonishing fact is that few Middle East countries “need” US foreign aid— most are “upper middle income” (Iran, Iraq, Jordan, Lebanon, and Syria) or “high income” countries (Israel, Saudi Arabia, United Arab Emirates) (Lewis and Kallab, 1983). In FY 85, Israel alone received $2.6 billion ($1.2 billion economic and $1.4 billion military) or 33% of all US Middle Eastern aid. Aid to Israel is so popular in Congress that even in the new era of fiscal austerity characterized by Gramm-Rudman-Hollings legislation, Congress still provides more money than the President requests (CQ, February 2, 1985)! The Israeli aid estimate for FY 86 increased to $3.6 billion, in only one year (USAID, 1986:20). Curiously, as Congress berates Latin debtors to pay up, Israel has the highest per capita foreign debt in the world. Israel is also the largest world-wide recipient of US aid. For many in Congress, votes for or against military aid is equivalent to a vote on military aid to Israel since this is about 50% of the total (Fuerwerger, 1979:69) Looked at more crudely, the aid amounts to payoffs to negotiate with antagonisitc Arab neighbors, particularly Egypt, or to withhold the use of force to keep conflict from spreading. Egypt receives $2.4 billion each year (USAID, 1986:20) for the same reasons. By FY 85, Israel and Egypt received $4.6 billion or 47% of the entire security assistance program (Shaw, Wall Street Journal, May 29, 1985). Israel and Egypt, the top 2 recipients, account for 41% of FY 87 bilateral allocations (see Figure 5.1) and 32% of the entire foreign aid budget (Nowels, 1986:15). US aid is devoted to these two countries largely for political leverage and security reasons. Israel actually returned $51.6 million in aid money to the US in FY 86 to help its congressional supporters make Gramm-Rudman cuts more equitably (CQ, January 25, 1986:170)! In this final chapter on US regional aid, we again examine the determinants and results relationship through the BRC model. In particular, we will examine the problems and prospects for US bilateral aid to: Israel, Egypt, Lebanon, Turkey, India and Iran. For US aid

US Aid to the Middle East 147 planners, this region is characterized by more political, religious and ethnic conflicts per square mile than both other regions combined. Aid to the Middle East is less a country than a regional and local problem, where each actor tends to take its cues from what it perceives is happening next door (like a political stock market). Of the three regions examined, in the Middle East local factors may be as important to US aid results as the organization of US actors, roles and goals. As suggested in Chapter 1, “Success” here really depends on which level one is talking about: country, regional, project, or technician (Montgomery, 1967:73). Practically any kind of US aid is viewed by some faction as intervention and neoimperialism. In contrast with aid to Latin America and Asia where the US policy process largely makes the rules of foreign aid and therefore determines results, in the Middle East, the rules are less amenable to change since they are less institutionalized and tend to reflect deeper feudal and anarchistic rivalries.

Results It was noted earlier that US aid can be charged with few unequivocal successes or failures. Even if the measure is defined broadly as: aid which directly or indirectly induces behavior desired by the US government through its foreign policy, it is still difficult to attribute results to aid efforts. In Latin America, much less has been spent (until recently in Central America) to achieve its few success stories. Such countries as Brazil and Costa Rica were already predisposed to US policy objectives and aid may have simply reinforced existing behavior. Though El Salvador may ultimately be a success case of a liberal, stable political center constructed by US aid, the results may not be transferable to other countries. Similarly, in Asia, successful aid seems to have merely reinforced entrepreneurial behavior in Taiwan and South Korea. Aid did not provide the needed political leverage in Vietnam (or it did and the US failed to exercise it), and almost failed to do so in the Philippines (a last-second aid victory). “Failure” has been measured by: excessive dependency, overly-complex and bureaucratic projects, and many harmful unintended consequences. Conversely, “success” has been measured by: realistic disengagement, mutually autonomous decision-making, projects at appropriate scales to meet real needs, and achievement of intended results. The existence of realistic conflict and mutuality of interests largely determined the success or failure of US objectives in both Latin America and Asia. In the Middle East, aid successes have been harder to come by, largely because of the immensity of ethnic-politico-religious forces which intervene between the US and its intended recipients (in addition to the usual institutional problems). Nevertheless, where conflict has been lengthy and realistic, as in Egypt, US aid has been expensive but largely successful in attaining such objectives as: assisting the poor, managing conflicts, deterring aggression and maintaining access for US forces (DOD, Volume l, 1986:23). Aid, of course, has been purely political to achieve leverage for US-European oil supplies and to prevent Soviet regional expansion. Such a “thin” rationale opens the door to the easy charge of US neo-imperialistic expansion, a rallying cry useful in solidifying otherwise opposing “sectarian”, “subnational” or even “nationalist” groups. The rather consistent post-World War II objective has been to use aid to stabilize the region around governments friendly first to Israel (the principal ally), opposed to the USSR, and somehow civil to each other. Food aid, ESF, DA, and Security Assistance have all been used for these purposes.

148

The Politics of United States Foreign Aid

The lengthy and complex political history of the Middle East is beyond the scope of this work. However, the effects of US aid on that history have been substantial in some countries, notably Israel, Iran, Turkey, and Egypt. Let us therefore attempt to describe the aid successes and failures in these and other selected countries, and later explain them by reference to bureaucratic actor role conflict. We have noted how dependency seemed to be a characteristic of failed US aid programs in parts of Asia and Latin America. The US cultivated Latin dependency during World War II by encouraging production for US markets to finance imports (in part from the US) always on the implied premise that aid would encourage “disengagement” and a more autonomous relationship. In Asia, the US reinforced a situation with its global East-West conflict premises, whereby the Vietnamese economy imported US consumer items and produced only war. Hence, dependency, or the exchange of decision-making autonomy in the political and economic spheres for reliance on external direction, was an unintended result of the US economy to which foreign aid is inextricably linked. Curiously, as noted, dependency is both a symptom of failure and success in some instances. Latin America depended on US imports and capital for decades after World War II. Now productive Latin American economies, built partly by US aid, depend on US export markets which threaten US domestic market structures and encourage US protectionism. In a final irony, the US now subsidizes its commodity exports to compete with economies it once subsidized, such as beef to Brazil at lower prices than Uruguay (Auerbach, Washington Post, July 13, 1986)! Once the dependence relationship begins, it tends to be mutually problematic, unless replaced by transition to “interdependence”. In the Middle East, it is largely the US which relies upon the wishes of actors who threaten to continue to destroy the “political commons” if more aid is not forthcoming. Shaw (Wall Street Journal, May 29, 1985) labels Israel and Egypt as “permanently dependent client states” who use US aid to “put off facing up to the consequences of their economic folly. “Whenever Israeli security and economic aid goes up, i. e. every year, Egyptian aid also goes up (and Saudi Arabia and Jordan clamor for increase shares of the aid pie to “balance” power and “stabilize” the Middle East arms race). With triple digit inflation and the highest per capita foreign debt in the world, it is easy to suggest that Israeli “dependence” on US aid makes matters worse. The Israeli military is so well equipped that aid actually subsidizes the growth of its “expensive and inefficient defense industry” (Shaw, 1985). On the other hand, the US is committed to the strength of Israel against antagonists, perceived and real. The Middle East strategy funnels aid to Israel as the central fulcrum of a wider plan to contain the Soviets and to keep the oil flowing. So, the US ups the annual aid ante to Israel and Egypt (mostly ESF, FMSC and MAP), then Saudi Arabia (which will now receive more advanced US missiles) and Jordan (which continues to demand more arms to defend against the others!) (see Figure 4.1). These substantial aid levels become new budgetary “bases”, below which it is politically difficult to go in future years. As noted, Congress questions aid requests only at the margin. In times of fiscal austerity decisions increasingly are made at the top, meaning by Congress with heavy pressure from Middle East lobbies. Moreover, after the President (Reagan to a greater extent than Nixon, Ford or Carter) increases multi-year base rights commitments to countries such as: Spain, Portugal, Turkey, Greece, and the Philippines, Congress tends to rubber stamp them, preserving

US Aid to the Middle East 149 the “indefinite dependencies of the security assistance program” (Shaw, l985). The net result of the congressional failure to play a consistent guardian role here is that US policy dependency on the Middle East is increased. This is costly and dangerous since the US ends up paying everyone off in what becomes the functional equivalent of extortion money to prevent oil cutoffs and Soviet influence. Each party can threaten to violate one or more US objectives and continually receive more US aid. Failure of foreign aid actors to realistically clash over guardian and spender issues permits this volatile situation to continue. This, of course, jeopardizes the entire US aid program and could be dealt with by creation of an improved forum that would gradually clarify positions and promote substantive decisions. “Success” under these conditions is hard to define. But if it means “conformity with behavior desired by the US government” (Burns, 1985:5), aid seems to be working against that objective by ironically increasing US dependency on its provision of aid! If dependency from either direction (donor or recipient) has not been conducive to successful foreign aid results, neither has been the tendency to unload projects of immense scale and technical complexity on governments with marginal absorptive capacity. We have discussed the mechanisms which encourage capital assistance and the tendency to finance infrastructure projects over recurrent or operating needs. These devices are reinforced by the perceived need to make a quick, impressive political gain by both the US and recipient counterparts. For example, the many adverse though unintended consequences of the US refusal (John Foster Dulles) to finance the Egyptian Aswan High Dam project in 1960 have been reviewed extensively by many writers. The decision is generally tied in with an unfortunate chain of events, leading finally to what has been consistently termed a US “fiasco” (Burns, 1985: xv). Others, though generally agreeing that project complexity (inappropriate technology transfer) is a defect of US foreign aid around the world, suggest that occasionally a “showcase project” is needed and that 1960 was really a “lost opportunity” to “catch the US taxpayer’s imagination” (Goldman, 1967:84). The Aswan Dam, by this view, could have been that tangible foreign aid blockbuster—a real winner in contrast with the majority of projects that are neither real successes or failures in the eyes of the world (Like some people, to paraphrase LaRochefoucauld, foreign aid projects are so superficial that they have neither real defects nor virtues!). While the construction of the dam, like the pyramids before, might have had political value, here for US policy objectives, it was wellrecognized even then that the larger the project the greater the unintended consequences. For instance, early Egyptian conversion of part of the Sudan into a lake for the project, and displacement of 50, 000 inhabitants (Burns, 1985:41) would have likely produced charges of neo-imperialism and turned into another disaster for the US foreign aid program. As it stood, by refusing to finance the dam, the US only lost influence temporarily to the Soviets which it regained later with smaller aid projects. Because of the mixed role status of foreign aid institutions and the volatility of the Middle East itself, many if not most aid consequences have been of the unintended variety. We need not limit ourselves to Israel and Egypt to see this. For example, India (a low income country where per capita GNP is below $400) has also been a relatively large recipient of US economic aid ($8.6 billion between 1962–84), 57% of which ($4. 8 billion) has been PL 480 food aid (USAID, 1984:15). Though India received only $176 million in FY 85 ($85 million in DA and $90 million in PL 480 aid), it ranks in the top five total recipients of US aid since 1946. For all of this effort, the results have not been entirely favorable or

150

The Politics of United States Foreign Aid

expected. In India, the US has attempted to maintain a contradictory mix of Middle Eastern security (with no security assistance), Indian democracy, and aid profitability. Events refuse to allow achievement of these goals! The first two objectives are related in that aid is used to maintain regimes that sustain democracy. However, overt US support for regimes often “de-legitimates” them and strengthens the local nationalism they are designed to overcome (Rudolph and Rudolph, 1983:88). US aid to “client” regimes, as noted, often produces dependency which usually does not lead to sustained “growth and governability.” Though the Indian Administrative Service is highly qualified and professional (unlike the state in most aid recipients), like the rest of this region, the Indian state has been faced with the additional challenge of religious “revivalism” by Sikhs, Hindus, and Muslims (Rudolph and Rudolph, 1983:102). US aid has been a “subsidy for the practice of democracy under conditions of scarcity” (1983:103). But any aid-financed efforts that produce a strong Indian response to nationalism is likely to backfire, weakening both the US position and the Indian state. On the other hand, withdrawal of US aid to try and avoid such dilemmas (such as exchanging “unprofitable” concessional aid for hard commercial loans as proposed by President Reagan; 1983:45) could weaken the Indian economy (servicing the hard loans from limited export earnings). This could easily destroy the broad but tenuous political consensus that has enabled India to pursue a relatively stable democratic path to development. Under conditions of economic adversity and religious revivalism, the relationship between US aid and local “governability” becomes complex and full of surprise results. The effects of US aid in Lebanon have been even more unintended than in India. Lebanon has not been a “top 10” recipient of US largesse. Since 1946, the US has provided $543.3 million in military-economic loans and grants. Of this, $263.8 million or 49% has been military aid. But, in a twist of script, its mounting political problems may be due in part to US aid to its aggressive neighbors, such as Israel which often makes “security” claims to Southern Lebanon. Syria makes similar claims to the Bekaa region. But since US aid to Syria up to 1980 was 99% economic aid ($582.3 million; USAID, 1984:27), it cannot be listed as a causal factor in Syrian behavior toward Lebanon unless an enormous amount of fund transfers occurred. As far back as 1952, the US viewed Lebanon as a “neutral bridge” between the Arabs and Europe—it seemed to coexist nicely with Arab culture and markets and European-style democratic institutions. But the political existence of Lebanon has always been in question and this centrifugal feature has contributed to regional instability. According to Ignatius (1983:1140), the 1982 War turned Lebanon into a “battleground for Arab-Israeli and interArab conflict. “The “sovereignty” of the Lebanese government was held together mostly by 5000 US and European troops. The rest of Lebanon was parcelled out to: the Israeli Army (Southern Lebanon and 50% of Mt. Lebanon), the Christian “Lebanese Forces” (other half of Mt. Lebanon, home of Maronite Christian and Druze Muslims), pro-Syrian “warlords” and Palestinian guerrillas (Northern Lebanon), and the Syrian army (Bekaa, home of the Shi’ite Muslims and Greek Catholics). At least part of the polarization of Lebanese politics was due to the Arab-nationalist crusade of Gamal Nasser in the late 1950s (Ignatius, 1983:1145). Lebanese Christians saw this movement as an alternative to Islam for uniting Arabs. But by the 1960s, as Nasser’s movement “threatened Lebanon’s East-West identity”, some Maronites looked to Israel as

US Aid to the Middle East 151 a model of militant non-Muslim statehood. This tendency frightened the Lebanese Muslims who deepened their attachment to Egypt and Syria. This “explosive mix” was detonated by Palestinians, whose armed presence further radicalized the Lebanese Christian militia. The stage was then set for what became the 1975–76 and late-1982 civil wars. For US aid planners (the US replaced Israel as chief ally of Lebanese Christians), what is left is a “nation of villages without a strong sense of statehood or citizenship”, a “pattern of strong local loyalties and an irrelevant central government”, and a “quasi-feudal political structure” (1983:1148–49) complete with warlords and fiefdoms. Clearly, the US did not cause this complex state of affairs through its disbursement of aid. But the question is what, if anything, can aid do to remedy Middle East political deterioration? Past disbursements of aid for “institution-building” did not tie the country together in the face of powerful centrifugal forces. In fact, this may be an extreme case of an indistinguishable foreign aidforeign policy. Where does one begin and the other end? Foreign aid efforts in Lebanon have been limited to ESF grants ($18.7 million for FY 85) which will be used for political stabilization, low income housing, rural-urban health care, education and agriculture. The FY 82 AID Congressional Presentation recognized that “as the government does not have authority throughout the country and its public service capacity remains limited, AID will also continue to rely on the capability and organization of Private Voluntary Organizations (PVOs)” (AID, Annex IV, 1982:88). What kind of institution-building strategy would work here? Training? What effects would US institution-building aid to the Government of Lebanon have on sectarian factions that now claim control over most of Lebanon? The probable results would be countercontrol efforts against both the Lebanese government and its US benefactors. Hence, the problem of Middle East peace and continuity of foreign policy become directly related to the efficacy of foreign aid. The isolated AID country program cannot succeed without wider regional foreign policy efforts—also executed largely through disbursements of economic and military aid. Success here depends as much on control of feudalism and nationalism which is destroying the “commons”, as on the efficacy of US foreign aid. Though foreign aid should be supported as a separate “responsibility center”, it is clear that success depends on coordination of both foreign aid and foreign policy. Under current institutional practices, decision-making is not integrated. The challenge of achieving successful aid results is increased by the fact that under conditions of civil strife and urban guerrilla warfare, the relationship between aid and development may be inverse. As war ravaged the country’s physical infrastructure from 1976–82, for example, capital increased in the Lebanese banking sector by 600%. According to Ignatius (1983:1152) “The banks were flush with cash because of remittances from Lebanese who had fled the country to work abroad. By 1981, remittances accounted for about 45% of Lebanon’s national income… This inflow of cash allowed Lebanon to transform a regular balance of trade deficit into a substantial balance of payments surplus. “This flow of remittances also generated a real estate boom for land around Beirut. The literally “skyrocketing” prices, according to Ignatius, “testified to the Lebanese love of their land and to the lack of alternative investments” (1983:1153)! Other economic if not “developmental” benefits of civil war include its tendency to attract large amounts of covert payments to support local factions fighting the war. These subsidies turned into direct income and employment benefits for Lebanese teenagers. “The starting

152

The Politics of United States Foreign Aid

salary for part-time militiamen was 1000 Lebanese pounds or roughly $250 per month” (1983:1153). Conversely, and this is the point, durable peace and successful aid projects could create “economic problems” by encouraging Lebanese expatriates to return home with corresponding decreases in remittances, war subsidies, land prices, and income and employment. Legitimate US aid has a difficult time demonstrating the superiority of traditional income and employment benefits to a society that through “illegitimate aid” (not found on the LOGFRAME!) can continue politically as it has for centuries and reap significant rewards in the process! Under these conditions, why should the US taxpayer find greater benefit in supporting aid than the Lebanese recipient? Such paradoxes, in addition to US institutional and local Middle East political culture problems, prevent US policy-makers from properly planning and executing aid projects in the Middle East. A similar set of local issues has constrained US aid results in Turkey. After the “loss” of Iran in 1979 ($1.04 billion in total aid of which 81% or $844 million was military aid, 1962–84; USAID, 1984:16), and in particular the loss of intelligence stations for monitoring Soviet activities, Turkey became increasingly important to the US and NATO. But the economic situation is tenuous: 32% of its foreign exchange earnings are appied to foreign debt service, the Iranian war has curtailed Turkish oil suppplies, and as in Lebanon, many Turks (roughly one million) send remittances back from European jobs as a significant source of foreign exchange (USAID FY 82 CP, Annex IV: 115–116). Proposed US economic aid (all ESF) amounted to $300 million in balance of payments assistance, of which 66% would be in grants. Since FY 82, economic aid to Turkey has declined to $175 million in FY 85, while military aid has increased from $403 million to $700 million (FMSC and MAP). In FY 85 then, 79.6% of US aid to Turkey was military ($700 million out of a total $879 million; USAID, 1986:14). As military aid to Turkey overwhelms the total aid package, the question arises whether this will buy the US any more leverage over Turkish policies than in the past? For, the past history of US military aid here contains a large unintended consequence component. The “kickoff” of President Truman’s post World War II “Containment” policy was in Turkey. Substantial US economic and military aid brought Turkish participation in the Americanled UN action in Korea. This was brought to fruition later when Turkey was admitted to NATO along with Greece in 1952 (Gruen, 1980:368). Secretary of State John Foster Dulles gained Turkish participation in the “northern tier” strategy of the Baghdad Pact (later CENTO which ended in 1979) by large increases in aid, despite resistance from Egypt, Syria and Jordan. But the US attempt to establish another Israeli-type client state foundered on issues of Greek treatment of the Turkish minority on Cyprus. JFK first aroused Turkish suspicion of US insensitivity to local issues over global anti-Soviet concerns when he considered removing obsolete Jupiter missiles from Turkey in 1962 in exchange for removal of Soviet missiles in Cuba (Gruen, 1980:369). More importantly, the Turks perceived correctly that the US consistently sided with Greece (common Christian culture, etc.), In 1964, LBJ had framed the impending Turkish military intervention on Cyprus in East-West terms as a threat to NATO and defense against the USSR. In 1974, Congress imposed an arms embargo after the Turks used US military aid equipment to intervene on Cyprus. Turkey retaliated by abrogating its base rights military agreement with the US (Gruen, 1980:366).

US Aid to the Middle East 153 As in Central America, overriding US preoccupation with East-West issues tends to arouse local wrath and negate the positive effects of both economic and military aid in unintended ways. Contradictory foreign policy thrusts, such as concern for the USSR instead of Turkish minority treatment by Greek thugs, or concern for USSR proxy behavior in El Salvador instead of uncontrolled death squad military actions on the local population, makes it difficult to leverage real US interests via aid. These contradicitons, of course, relate to the disjointed process by which foreign aid policy is made. In Turkey, one US interest which Israel and Egypt also share, is containing the spread of Islamic fundamentalism and “communist-backed radicalism in the Middle East” (Gruen, 1980:78). But military and economic aid must be supported by consistent funding over time to achieve its intended effects on institution-building of the political center, strengthening the economy, sectoral activities, and welfare of the local poor. Regardless of which US aid goal prevails, the guiding notion is the global foreign policy notion that aid will lead to recipient behavior compatible with US interests. Burns (1985:3) suggests that among others, JFK viewed this connection in Egypt as a “rigged slot machine” into which economic aid is inserted to produce “congenial political results. “Because of simplistic notions like this operating according to the vagaries of US domestic power politics, many aid successes and failures have been unintended. This is particularly true in the cases of the major US recipients: Israel and Egypt. Where public policy can only “control” one of many complex variables affecting a program such as foreign aid, results, successful or not, are likely to be partly the product of chance. Such is the case in Egypt where US aid has been involved historically with a host of unexpected failures and successes. Proving that the dollar amount of aid may be more imporant than its programmatic substance, Egyptian aid and foreign policy ties with the US remain strong and the country is a pillar of stability in an otherwise turbulent region. A US “blunder” (Burns, 1985:105) that may have evolved into a belated success was the US decision to cancel its offer to finance the Aswan High Dam in 1956. The US loan was a prerequisite for further aid from the World Bank and Great Britain. President Eisenhower and Secretary of State Dulles, whom Burns describes in this affair as about as “subtle as temperance crusaders in a distillery” (1985:110), had mixed up Nasserism with international Communism. They imagined that anti-Communism in the Middle East could, by US pressures and alliances via the “Eisenhower Doctrine” (see Chapter 1) become a “potentially more powerful force in the Arab world than revolutionary Arab nationalism” (1985:110). They reacted with abrasive force to, among other things, Nasser’s attempt to “blackmail the administration into providing more money by hinting that the Soviets had offered to finance the entire project” (Spanier, 1983:88). The unexpected result of this short-term US capital savings, was that Gamal Nasser nationalized the Suez canal (ostensibly for revenues to finance the dam!). As the political blackmail potential of this act was high to countries like Israel and Great Britain (which relied on the canal), they joined forces with France (which disliked Nasser’s support for Algerian rebels) and attacked Egypt in 1956. In a feeble attempt to ingratiate itself with Nasser and the Third World, the US intervened to save Nasser and stop the fighting by withholding support from the three aggressors. Unfortunately, the Soviets upped the ante and threatened to annihilate all three unless they discontinued their warlike activities.

154

The Politics of United States Foreign Aid

Hence, the unexpected result of the US withdrawl of its Aswan offer: Soviet emergence as champion of Arab nationalism! In FY 87, the US will spend approximately $1.0 billion (ESF) and $189 million in PL 480 food aid in Egypt (USAID, FY 87 CP: 91). Ironically, the US will also add to the Egyptian “energy sector” next fiscal year by providing $15 million to replace the “turbine runners” of the Aswan High Dam (USAID, FY 87 CP: 103)! Maintenance of the capital originally provided by the Soviets may bring more “political returns” and leverage than the uncertainties of financing long-term projects in this regional institutional mazeway! Long-term DAprojects have been historically unpopular with Congress because they take a long time to produce uncertain results (such as the Aswan Dam to Soviet “taxpayers”!). The foreign aid program has attempted to stabilize the Egyptian economy and political system through ESF, DA and Food Aid projects. As in other countries, each type of aid produces measurable successes and failures. In 1975, the US was supporting Anwar Sadat to hold Islamic and other sectarian extremists in check. He had publicly played up the ability of US aid to reinvigorate the economy which would provide him a base of support for consistent political action. The Suez Cement Plant illustrates the unintended failures of USAID in Egypt and the problems such failures can bring for larger US foreign policy objectives. The Project Paper was finished by USAID in 1976 for the plant which was to “underpin housing projects already underway in the Suez Canal area and in Cairo” (Burns, 1985:196). Congress appropriated $90 million to cover foreign exchange costs in June 1976 for an expected completion by early 1980. 18 months after Sadat’s 1981 death the Suez project “had still not produced its first bag of cement” (1985:196). Burns suggests that inflated expectations were only part of the problem. Rather, “problems connected with the assistance program itself… diminished its political value to the US government” (1985:196). As noted, given its volatile budgetary context, AID has every incentive to stress planning over implementation. “Egyptian bureaucratic sloth and inefficiency served as a brake on many of AID’s undertakings, but the reams of red tape and the lengthy feasibility studies which (USAID) wrapped around its projects also delayed completion. As a result, an enormous bottleneck developed in the aid “pipeline” in the 1970s. By the end of 1980, nearly half of the more than $5 billion in economic assistance appropriated by Congress between 1974 and 1980 for use in Egypt remained unspent” (1985:196). In particular, the “inability of AID to bring its projects to quick fruition clearly hampered American efforts to produce the ‘peace dividend’ that Sadat needed” (1985:197). This was not just a local phenomenon. It should be noted that as of December 31, 1983, AID unliquidated balances reached $6. 4 billion world-wide, the highest level since FY 1978 (Senate Appropriations Bill, FY 1985:24–25). The aid generated expectations that structurally and institutionally could not be fulfilled and, in a now-familiar pattern, this jeopardized the credibility of both Sadat and his US allies. But the problem was less a function of Egyptian political culture than distrust among US bureaucratic actors, in particular congressional interest in budgetary policy control of foreign aid and AID interest in covering its moves from outside critics. Where the external recipient context demands action, the US cannot respond. It is instructive to note that the US moved from capital project aid to greater reliance on PL 480 aid in Egypt (1977), because

US Aid to the Middle East 155 it was more politically popular with Congress, tangible to Egyptians and bureaucratically simple (1985:197). Similarly, in 1975, Secretary Kissinger (Ford Administration) pledged $750 million in grants and concessionary loans, and $200 million in PL 480 food aid to Sadat as a reward for cooperation in the Israeli negotiations (Sinai II). This was purely political aid, based neither on Egyptian need nor absorptive capacity (Burns, 1985:184). Predictably, the Cairo embassy had trouble identifying enough projects on which to spend money. The infrastructure was crumbling, the public bureaucracy bloated and the agricultural sector stagnant.7 Nevertheless, US aid tended to “enrich a small number of upper middle class entrepreneurs (10% of the population) or “fat cats” (1985:188). By 1977, two-thirds of the $2. 3 billion in grants and loans committed to Egypt (1975–77) had still not been expended. “Bureaucratic inertia in both Cairo and Washington contributed to this ‘clogged pipeline’. AID had its own interminable feasibility studies to conduct…” (1985:189). Finally, in the curious world of foreign aid, project “success” can often lead to programmatic “failure” (as in Vietnam). AID efforts in Egypt contributed vitally to food supplies and infrastructure. This provided the basis of leverage to reach an Israeli-Egyptian accord in 1979. But these US-Egyptian successes produced growing backlashes. The US had not miraculously cured the economic problems with its projects, and this “fertilized” political unrest among Islamic fundamentalists, militant Copts and Moslem extremists (1985:195). These extremists criticized the tendency of US aid to feather the nests of the rich (which it did) at the expense of the poor (which it usually did not, but popular perceptions were contrary). These problems and the implicit attachment of Sadat to the US (by 1981 the Cairo AID Mission was the largest in the world—a symbol of peace but a convenient target for extremists), threatened Sadat’s efforts and culminated finally in his assassination by Islamic screamers in 1981. Under these volatile circumstances, aid successes in the region were mostly unexpected. For example, after the “disastrous chain of events” in 1956–58 related to Aswan, the US quietly agreed to food shipments to the Nasser regime distributed by CARE. This aid policy reduced Egyptian-US tensions in 1959–60 and later in 1961–62. But after African students protesting US policy in the Congo burned down the US Information Service library in Cairo (1964), many in Congress such as Robert Michel (R-Ill) sought a cutoff of US aid to Nasser. Nasser also reacted sharply to the US decision to provide Israel with M-48 tanks to balance out T-54s sent to Egypt in 1963 (1985:163). As in the Aswan affair (1956), Nasser was using the threat of a Soviet rapprochement, i. e. aid and arms, as a lever to gain more US aid (food, arms and infrastructure). Here, the US (President Johnson) was trying the “short leash” on Nasser. Based on this policy, the US suspended food aid in February, 1965. Given the record of US intransigence, Nasser moderated his anti-American behavior (apologized for the library burning and agreed to pay $500, 000 damages), and became conciliatory in the Yemeni civil war (a SaudiEgyptian agreement to end outside involvement). LBJ used short-term food aid agreement extensions (6 months) to keep Nasser in check and thus provide US leverage via aid. Similarly, the US used food, economic and military aid combinations to encourage Sadat to make concessions to Israel (1974), energize the private sector, and end dependency on Soviet markets and technology. Aid also encouraged Sadat to negotiate the 1979 Camp David peace accords in 1979. As indicated, translated across US and Egyptian institutions,

156

The Politics of United States Foreign Aid

US aid could not work short-term miracles for the economy and the generic flow of aid contributed to fertilization of large crops of anti-US extremists. But it cannot be doubted that US aid contributed significantly to Egyptian evolution to “political maturity” (Ignatius, 1983:1141). Though “unprofitable” in the narrow cost accounting sense, the aid stabilized Egypt, Israel and several other nations such as Saudi Arabia, and this enabled local political institutions to improve their respective economies and “buy time” in the classic sense.

Goals Of the three regions studied, the Middle East is least deserving of US aid. The countries are unique in that they are oil-rich and provide each other with foreign aid. Kuwait began the first concessional assistance by one developing country to another in 1961. By 1984, the authorized capital of the 10 major OPEC aid agencies stood at $28 billion. Of this sum, $24 billion is committed to such purposes as “government-to-government general support assistance” (long-term bilateral grants and loans) and multilateral project aid (Shihata and Sherbiny, 1986:17). OPEC has disbursed $16 billion in aid to its trade partners and allies (66% in Asia) since the mid-1970s oil price increases. It has moved from the administrative complexities of project aid (pre-1973) to increased reliance on more rapidly disbursed grants and long-term loans, and direct deposits under general government assistance (1986:19) which is like ESF. Hence, the US provides bilateral aid to a region that is itself a major bilateral donor! The obvious US motive in duplicating such aid is security (prooil, anti-USSR). As noted, if ESF is included in security assistance, from 1946–84, 67% of US aid to the Middle East has been for defense-related purposes; 33% for developmentalfood aid purposes (USAID, 1984:7). But profitability and development have also been important US goals in this region. AID funding for these goals tends to reflect the particular mix of local problems, defined consistently with US global-regional political interests (containment, markets, democratic development). Here again, foreign aid becomes the major vehicle for foreign policy expression. Within the US aid portfolio, DA, Food Aid and to a lesser extent, ESF, become the major tools of socioeconomic development. The responsibility attached to these tools is far beyond funding authority to change recipient policies or reinforce structures where necessary. In the Middle East, US aid tries to “develop” countries via the simple notion that economic growth and stability will lead to mutual recipient-US political and financial benefits. PL 480 aid (mainly wheat exports) to Egypt, for example, provides a rapid, congressionally popular and administratively simple means to leverage US interests. By 1960, PL 480 imports accounted for 66% of Egyptian grain imports (Burns, 1985:119) but it reduced US-Egyptian political tensions. This form of aid has served dramatically as a “leash” to guide Egypt toward reconciliation with Israel and toward political maturity as in no other recipient case. PL 48O supplemented direct transfers of ESF and military aid, over which the US has even less control. Nevertheless, US aid provides a major foreign exchange source for Egypt. Despite pipeline problems of the past,, a record $1.2 billion was expended in FY 1985 (USAID, CP, FY 87:90). By contrast, such countries as Israel and Jordan are given direct cash transfers (ESF) to “develop” themselves via agroindustry and infrastructural projects.

US Aid to the Middle East 157 Lebanon continues its “sectarian fighting” and will receive only $2 million in ESF in FY 87 for projects, such as medical care and relief assistance. Other countries such as India and Morocco have fewer security problems, permitting DA to be employed (in theory) more effectively. The AID FY 87 program in India, for instance, is entirely DA ($72 million) with $6 million for the Child Survival Program. Indicative of the diversity of programs in this region, AID is spending $28 million far: social forestry, fuelwood, and conservation programs, and is using PL 480 Title II local funds ($80.5 million) for PVO programs in cooperative development, rural development, health, education and supplementary nutrition (USAID, FY 87 CP: 132). Morocco’s $38 million in DA will finance agriculture, population, energy and human resource development projects. PL 480 funds ($40 million) will be used to alleviate grain import needs and provide local currency financing of development projects (USAID, FY87 CP: 208). These examples suggest that although “development” has often been defined in simplistic evolutionary traditionalmodern capitalist terms, the actual US aid program has been more diverse, taking into account local priorities and interest group configurations (in regional countries other than the US client states of Israel and Egypt). The US has also tried to make development “profitable” by 2 means: (1) congressional provisions to “tie” aid, ensure repayment, and maintain accounting controls, all of which contribute to finding more benefits than costs in providing aid, and (2) providing food aid which is visibly profitable for US farmers, congressional representatives and aid recipients. Food aid is now even more profitable in that local currencies are increasingly used for local costs of development projects in forestry, agricultural development and health. For example, Brady notes (1986:1O) that the majority of forestry (social, tropical, conservation) and tree-planting activities in LDCs funded by US aid are now accomplished with US food assistance. DA projects are funded in part through Food-ForWork PVO and World Food Program (WFP) projects and through use of local currencies generated under Titles I and II. The point is that, although occasionally at the expense of recipient wrath, at least in the case of DA, US institutions have combined to build consensus for foreign aid by making it appear profitable. We noted in the first chapter how Congress viewed aid as an investment to be safeguarded with many conditions and that the project was supposed to pay for itself directly (convertible currency, debt service), indirectly (local currency), or in kind (tied aid, base rights). This fiscally conservative attitude has largely prevailed with the possible exception of the New Directions and Alliance periods. So, consistent with frequent US insensitivity to culture and nationalistic forces, the Aswan High Dam, for example, was perceived to be merely another investment, for which the US would monitor Egyptian fiscal and monetary policies to ensure ability to repay. Nasser, of course, resented such blatant intrusions into Egyptian affairs and wanted the US and Britain to turn over its grants to the World Bank to reduce Western political manipulation of the aid (Burns, 1985:58) and to give him “one-stop shopping” on approvals. Further, the very idea behind PL 48O was profit. According to Burns, the program was a “businesslike attempt to ease storage costs” (for US commodity surpluses) “by allowing friendly food-deficit nations with limited dollar supplies to purchase American agricultural commodities with non-convertible currencies” (1985:116).

158

The Politics of United States Foreign Aid

A persistent problem of US aid has been the implicit notion that Third World leaders play by American profitability and political rules. As noted before, McGeorge Bundy observed that LBJ treated all Third World leaders like reasonable US politicians that could be persuaded to compromise on any issue by some combination of threats and promises (1985:152). This attitude is prevalent among US aid policy-makers in Congress, State, and DOD (less so in AID because of their frequent Peace Corps backgrounds and rotation into the field where contact with local counterparts is required). The presumption is that the Third World, like we, are committed primarily to the genie of financial profit; that this entirely reasonable goal is shared and that interference with profit-making is viewed by the Third World, again like ourselves, as a major problem to be resolved quickly. Those who view matters differently are deviants, commies and outside troublemakers. But particularly in the Middle East, the notion that everyone has a price (and a low one at that) stirs resentment easily from the many poor who are also members of religious sects. Still, the notion of profit in food aid has been “softer” than in loans and this (despite occasional complaints from countries such as Brazil, Canada, Australia, and New Zealand which argue reasonably that food aid interferes with their profits by dumping US surplus) has helped it gain more good will and leverage for the US. JFK sensed some of this resentment and had tried to establish better relations with Nasser by using economic aid without the alliance-building pressures or “pactomania” that accompanied most Middle East aid. JFK tried to use aid (DA and PL 480) less for security, profitability and leverage than for longterm “take-off” and economic development (Burns, 1985:125). But regardless of US foreign policies and foreign aid techniques, the states of the Middle East have been at war more or less constantly for the last 37 years. Whether or not the Soviets are about to pounce on the oil and extort the West by cutoff threats, the conflict between Arab-Zionist religious nationalists and related issues like Palestinian refugees constitute enough of a perceived threat that the US has consistently attempted to balance nationalism by some combination of force, diplomacy and aid. But foreign aid usually involves the other two tools! As noted by Ignatius (Washington Post, July 13, 1986), while the US April 1986 bombing of Gadhafi was successful in “raising the costs of terrorism” and changing the political balance away from radicalism, the US still needs to provide Middle East moderates with an alternative to radicalism. Foreign aid could do this if properly programmed, funded and executed. Until aid can build consensus behind reasonable solutions, the American foreign policy goals in this region continue to sound like the “liberal premises” and “containment” policies that have failed for decades: preserve territorial boundaries of friendly clients like Israel, contain the USSR and radical nationalism, and preserve US access to oil via aid geared toward political stability (Gurtov, 1974:10). On the other hand, the US showers economic and military aid to its clients (Israel and Egypt) and prospective clients (Saudi Arabia and Jordan) far beyond recipient absorptive capacities. The bulk of all US aid goes to the Middle East for security purposes. But aid is a two-edged sword. It is a tool of stability and development. But US aid has also been used by recipients with occasional US consent as a tool for oppression and torture (Shah of Iran). This feature is easily used by opponents of authoritarian regimes to arouse anti-US resentment. The perceived “reasonableness” of US goals is countered by the Middle East view that the US backs corrupt and unresponsive regimes with its aid and seeks to impose

US Aid to the Middle East 159 client states like Israel on them (Graham, Washington Post, December 1, 1985). Conversely, radical nationalist and Soviet threats (real or not) provide US client regimes with readymade budget requests for more military aid that are highly successful in Congress. In 1953, Nasser even told the US that more military aid would be needed to keep the armed forces loyal and pro-Western (Burns, 1985:15)! The usually positive results of such pleas in Congress generate similar requests from neighboring clients. Shaw notes that Egyptian aid, for instance, tends to be driven up with each increase in Israeli aid (Wall Street Journal, May 29, 1985). Saudi Arabia feels slighted because it has to pay cash for its US weapons (via the Special Defense Acquisition Fund) while the rest of the region relies on loans and grants (FMSC and MAP). King Feisal of Saudi Arabia also wants President Reagan to support the “just cause of the Palestinians”, an issue which still remains unresolved after it surfaced in 1955 over the Aswan Dam project. When Israel and Egypt receive more arms, Jordan wants more too (though it recently withdrew its request for hand-held anti-aircraft weapons) (CQ, February 16, 1986). So US foreign aid policy is not viewed as evenhanded by the recipients who have to live with each other as well as the US and Soviets. The interesting fact about US aid to the Middle East is that it has taken on a political life of its own. For example the FMSC program requires US procurement, which accounts for its powerful support among defense contractors and congressional allies (it is not simply a consumer-responsive program as DOD maintains). The FY 87 FMSC request alone amounts to 37. 8% of entire foreign aid request ($6.1 billion out of $16.1 billion; USAID, 1986:6). Curiously, the FY 87 appropriations bill provided “up to $250 million for the procurement and up to $150 million for research and development” of the Israeli Lavi jet fighter. The question then is whether the American-Israeli Political Action Committee (AIPAC) could be so strong that it actually overpowers the big military-industrial complex lobbies and sends US procurement overseas? If so, the notion of leverage and purely political aid may have taken on new meaning! The security and “governability” goals have resulted in a classical pattern of US “dependency” on preserving a Middle East balance of terror at a large and growing price. Can aid flexibly and effectively build stable institutions before the volatile mixture of armed warlords, and their international supporters, produces enough oil blackmail to ignite the region? If so, how much aid and in what programmatic proportions should it be provided? Will the resource needs of the rest of the US (and Third World aid recipients) eventually prevail over intense and powerful Middle East lobbies?

Constraints The conflicting goals and pressures over Middle East foreign aid policy, as well as the complexity and volatility of the local political context, virtually assures that the foreign aid program will not achieve the results which programmers intend. For a variety of reasons related to the structure and functioning of the aid bureaucracy and the American political system, the foreign aid program emerges each year from the distortions produced by multiple and conflicting pressures. The program is the result of a power game in which AID is the weakest actor and DOD-State controls its scope and direction. The roles of these actors, as well as Congress and the Presidency tend to be redefined with each issue-they are not consistently guardian or spender. This produces a relatively incoherent foreign aid

160

The Politics of United States Foreign Aid

policy which tends to throw large amounts of money at problems with only marginal results in other than abstract terms like “conflicts deterred”. Under these conditions, USAID reacts defensively to try and preserve its budgetary base by “institutionalizing outside criticism” (Tendler, 1975: Chapter 4). This tends to give aid critics a free hand and renders the foreign aid program even more unpopular. Few actors can defend the program substantively and aggressively before US publics without getting mixed up in the jargon of foreign policy. We have noted how structural and rule constraints operating on the aid machinery affects goal achievement. Historically and presently, Congress imposes conditions in appropriations bills which reflect individual member, regional interest, and commercial concerns. Sometimes AID may be helped by this, such as accounting controls and requirements that contra aid not be funded out of CIA’s “contingency funds which are considered all but unlimited” (Omang, Washington Post, July 13, 1986). Elsewhere, the congressional tendency to view foreign aid as waste and porkbarrel combines in the tendency to “micromanage” the program, with restrictions for example, to serve environmental protection and women in development. This intrudes on the work of AID field managers and generates resentment by local counterparts. However, the best current example of a constraint that may damage foreign aid credibility even further is the Gramm-Rudman-Hollings legislation noted earlier. The effect of the law is to require across-the-board cuts (4.3% for FY 86) in outlays. But obligations and outlays for foreign aid, like all public policies, are as much a product of technical factors as congressional decisions, such as the type of program (DA with slow outlays; MAP with fast outlays), weather (DA agricultural development programs) and the economy (a 1% error in the forecast of economic growth can throw off FY 87 revenue and deficit projections by $15–$20 billion). In other words, according to Gray (Washington Post, July 13, 1986) G-R-H is “inherently flawed”. G-R-H acts as a powerful constraint on managerial flexibility for a program that must face political uncertainties at home and in the recipient country. The aid delays and corresponding waste of budgetary resources caused by US rules and repertoires, such as G-R-H, are legion. The Eisenhower attempts to use military aid as leverage with Nasser (1952–55) were unsuccessful because the “complexity of America’s aid machinery” was an “anchor” on efforts to use the lure of military assistance to control Egyptian behavior (Burns, 1985:34). While the US spent 3 years in unsuccessful talks with Egypt, the latter received rapidly disbursed Soviet-bloc aid in 1955 (Czech arms) which broke the Western monopoly on arms supplies to the Middle East (1985:32). Though Eisenhower-Dulles (and Ronald Reagan now in similar decision contexts) viewed the crisis as a Soviet opportunity to leap over the “Northern Tier” and establish a base at the heart of the Arab world (the East-West context), the more significant issue for foreign aid has been the tendency for suspicious neighbors to cash in on the US with each crisis. During this period, in which the US was trying to increase leverage via military aid, the US imposed counter-productive conditions on local behavior. Despite Nasser’s obvious nationalistic leanings, Congress required through the Mutual Security Act of 1954 that arms assistance be accompanied by a “Military Assistance Advisory Group (MAAG) “to supervise the use of weapons and “ensure that recipients kept their promises” (1985:16). The Egyptians predictably withdrew their request for military aid. Eisenhower tried to diffuse the situation with more economic aid “thus freeing funds in Cairo for commercial

US Aid to the Middle East 161 military purchases” (1985:17) which were not subject to MSA restrictions. The complexity and sluggishness of this system (almost totally a product of Congressional-Executive distrust and check and balance zeal) generated impatience in the past among Third World recipients such as Nasser, and continues to do so in the delicate post-Marcos Philippines where Congress recently slashed the executive request out of budget balancing selfrighteousness, ostensibly for “profitability” (of the penny-wise pound-foolish variety). Despite real successes by the economic aid program in some parts of the world, Former US Egyptian Ambassador Eilts notes that “no single aspect of current US-Egyptian relations receives as much excoriation from Egyptian officials as the economic aid program” (Burns, 1985: xvii). He traces this to red tape, conditions attached to restructure government machinery, limited visibility of US-aid financed projects, unwillingness of the US to turn over aid monies to Egypt (as it does in Israel) and, excessive Congressional interference (1985: xvii). This doesn’t even touch the resentment generated by aid conditions requiring peace with Israel and anti-Soviet behavior (such as the 1955 Baghdad Pact when no Soviet threat was apparent) (1985:24). Similarly, in 1963, the Senate passed the “Gruening Amendment” to Section 620 of the Foreign Assistance Act. This prohibited US aid to any country “engaging in or preparing for aggressive military efforts” against any other country securing US aid (Burns, 1985:145). Here was a righteous expression of Congressional displeasure against US interests by reducing managerial flexibility in the field. Later policy-makers tried to inject considerations into aid such as the vague notion of “even-handedness” that only produced backlash, clogging the pipeline even further. President Eisenhower’s attempts to build a close relationship with Egypt in 1955 “proved deeply unsettling in Israel” and feeling isolated, proceeded to attack the Egyptian-“controlled” Gaza strip in retaliation for Moslem attacks on Israelis there. The Egyptians then resented US largesse in providing them with only limited and tied aid! (Burns, 1985:26). Thirty years later, US aid is still used as a leash on several warring countries/groups in this region and with about the same level of uncertain results. The current attempt to impose structural conditions “evenhandedly”, such as governmental reorganization, and changed fiscal and monetary policies, only brings further resentment. US aid policy-makers know that Israel will permit neither conditions nor foreign technicians on their soil. Egypt feels justifiably slighted at US “acquiescence” in this, since the current three-digit inflation in Israel is partly a product of an uncontrolled budget process that allows cabinets complete freedom to spend funds (CQ, March 9, 1985:435), many of which have been provided by the US (nearly 20% of their budget). Currently, G-R-H legislation will act as a constraint on US aid to the Middle East. In addition to learning complex explanations about the US authorizations and appropriations processes, earmarking patterns, continuing appropriations resolutions, reprogramming and even reconciliation, Middle Easterners can now learn about “sequestered” funds. Since many recipient policy-makers have been trained in US public administration programs with USAID funding, such lessons may have already been learned and can now be applied to maximum advantage. For instance, ESF to Egypt will be reduced by 4.3% or $35 million (leaving $815 million) for FY 87 to “save outlays” and help balance the budget. G-R-H cuts will be “uniform” in name only. As noted in Chapter 3, programs whose funds are obligated in the first year (capital and military aid) will be penalized more severely than those with

162

The Politics of United States Foreign Aid

slower rates of obligation/outlays (DA). Both types of programs would suffer, but given the need for capital programming, G-R-H would result in a paper savings only. Further, rumors of G-R-H also intensified the politics of aid funding. In the Middle East, a $1.2 billion Israeli ESF package was miraculously transferred “at the discretion of Congress” before it was subject to FY 86 “sequestration” (also termed “castration” by Aaron Wildaysky) by G-R-H (Maroni, Nowels, and Woldman, 1986:27). The aggregate effect of such rule constraints is to displace the goals of security, profitability and development with particular agency agendas covered by Congressional legislation which, in turn, damages the coherence of the entire program.

Policy Actors It was noted earlier that administrative organizations tend to acquire the operational characteristics of their clients. The thesis that realistic conflict produces rival trust and leads to improved foreign aid policy-making and results is most difficult to apply here. An academic might even term it “counter-intuitive”. The fact that tribes, clans, sects and states of the Middle East have been locked in conflict for centuries somehow does not seem to be increasing mutual trust. Similarly, US institutions charged with formulating aid policies for this region seem to suffer from the same disease—ongoing conflict, contentiousness, and distrust with only occasional lapses into substantive consensus and bipartisanship. In an Egyptian folk tale about the frog and the scorpion (Hedrick Smith cited by Burns, 1985:167), “the scorpion, unable to swim, asked the frog to carry it across a stream. The frog refused, fearing that the scorpion would inflict a fatal wound in midstream. The scorpion pleaded with the frog, promising not to sting it. Reluctantly, the frog acquiesced, put the scorpion on its back and proceeded across the stream. Halfway across the stream, the scorpion suddenly stung its benefactor. As both frog and scorpion began to drown, the frog asked in confusion ‘Why did you do that?” I couldn’t help it’ was the scorpion’s reply.” The problem of nation-building via foreign aid is, at its core, one of trust. But in the Middle East, many feudal “advocates” or “scorpions” in Lebanon, Iran, Egypt (under Nasser), Syria, and especially Libya, clash in complex and lengthy family feud-like wars. This immense centrifugal force (made worse now because classy and more influential scorpions like Nasser don’t exist anymore!) is not countered by authoritative and representative institutions that seek to cultivate shared interests (guardian “frogs”). The tragedy of sharing this kind of political “commons” is that, with sufficient armaments willingly supplied by a variety of private/public sources, all drown because they “can’t help it”. Into this fray, for largely self-interested reasons, the US has entered as “guardian”. But as a representative democracy whose institutions respond to the powerful, its mission is tainted by conflicting pressures and institutional constraints of its own. Distrust and information distortion, partly home-grown in the bureaucratic culture and from interest group pressures, and partly imported from its Middle East clients by what Foster Dulles once termed “localitis”, prevent US foreign aid from providing a viable alternative to the radicalism and perennial distrust of the region. It is argued here that US problems in intelligently guarding spenders both on the US side of foreign aid and in the Middle East are not simply absence of knowledge or sensitivity. These weaknesses exist on occasion. But systemically the problem is that institutional

US Aid to the Middle East 163 incentives distort information and power to further narrow, often inappropriate ends such as personal agendas. Part of the consensus problem is that a single solution (the “one best way” of apportioning aid among subprograms for a recipient) cannot be proffered to please all participants in the policy-making process. The very structure of US foreign aid policy-making encourages fragmentation, distrust and an emphasis on form (planning and programming) over substance (implementation). Much of these dynamics are evident in conflicts over annual program budget authority. Let us examine how Congress and other US institutions play mostly “mixed” budgetary roles and suggest why this tends not to produce needed “rival trust” except in special circumstances such as Egypt. The Aswan Dam affair was noted briefly as a foreign policy issue. It is also an excellent case study of the difficulty in obtaining consensus on Middle East aid issues. Subsequent projects and programs have been much smaller. But the same counter-productive dynamics are evident today in the roles played by US foreign aid institutions. The US formally cancelled the offer to finance the High Dam in July, 1956. Before then, the key supporters of aid were ICA (USAID’s predecessor), the CIA, State, Ambassador Byroade and Dulles-Eisenhower. Secretary Dulles supported the aid proposal but feared the effects of a degenerating aid relationship with Egypt, the volatile Palestinian refugee issue, and possible Congressional indifference to economic aid later on. Congress became the focal point of opposition. Though British Prime Minister Eden had already withdrawn support, the Shah of Iran was opposed (viewing this as a reward to Nasser for bad behavior against the US), and Treasury Secretary Humphrey’s opposition influenced Eisenhower’s behavior, the bulk of criticism came from Congress. However, it cannot be said that Congress “guarded” the public purse here, but rather the interests of a few powerful groups. The Israeli lobby (now AIPAC) swayed many in the House and Senate to vote against it in Israel’s interest (Burns, 1985:50). The cotton lobby even feared that arable land reclaimed by the dam project would weaken the US hold on world cotton markets. This lobby argued that the High Dam was a subsidy for destruction of the southern cotton industry (1985:69), i. e. that the loan was “unprofitable”! But differences between the structures of the House and Senate also affected the decision to oppose US aid to Aswan. Before the advent of budgetary conservatism as a Congressional pastime, the House delegated its foreign aid decision-making to a “few selfstyled experts”, rather than spend lots of time on an issue that had little home district impact or visibility for re-election. One such expert was Otto Passman (D-LA), then Chair of the House Appropriations Subcommittee on Foreign Affairs. The level of effective guardianship in the House on this and other foreign aid issues is revealed by the fact that he voted against every foreign aid authorization and appropriation for 20 years following World War II. (Congress cut executive aid requests by an average of 18% during the 1950s). “Son”, he told a State Department official, “I don’t smoke and I don’t drink. My only pleasure in life is kicking the shit out of the foreign aid program of the USA” (Burns, 1985:49). This attitude contrasts sharply with Passman’s current counterpart, David Obey (D-Wisconsin) who actively pursues a more sensible balance in economic-military aid via educational consensus-building exercises such as as open appropriations hearings and panel discussions which serve to question the broader rationale for foreign aid (CQ, March 16, 1985:498).

164

The Politics of United States Foreign Aid

It was noted that State Department and Ambassador Byroade also supported the dam proposal (here they were the guardians and Congress advocated special interests—a peculiar but not uncommon role reversal). “Arabists” in the State Department had been pushing for good relations with the Arab world and viewed US involvement with the creation of Israel in the late 1940s and early 1950s with alarm. Hence, for them the dam was an obvious expression of US goodwill. Similarly, DOD also advocated closer US-Egyptian ties to decrease US vulnerability to Arabs in the Middle East (Burns, 1985:12). Ambassador Byroade rebuked Eisenhower’s “punitive thinking” and over-concern with East-West and American-Israeli issues. For this perspective he was accused of “localitis” by Dulles and transferred to South Africa in 1956 (1985:92). On the opposition side, the Israeli lobby coalesced widespread US support for Israel in Congress. But Ike had received only 36% of the Jewish vote in 1952 and thus had no fear of losing what he never had via a favorable decision for Egypt. His decision, like many in foreign aid, was not directly-related to the substance of the issue at hand. Rather, he was deciding to punish Nasser for contumacious, scorpion-like behavior, and to preserve his Yugoslavian aid package in Congress (1985:4). AIPAC continued to reign supreme, stimulating LBJ to cut off Egyptian aid after the 6 Day War in 1967. They also protested resumption of aid in 1974. Where 2 countries (Egypt and Israel) now account for 32% of the entire US foreign aid budget, Shaw (Wall Street Journal, May 29, 1985) wonders whether Congress has not failed as a restraining influence and key interagency processes have not broken down? Guardianship has been weak in the absence of evaluation of base rights and other security aid in this region. Nor has Congress demonstrated “evenhandedness” in analysis of the rest of the foreign aid program not covered by powerful lobbies. In this context, Congress responds rationally to powerful organized interests, such as AIPAC, State Department Arabists, defense contractors, commodity lobbies, and elsewhere to special concerns— base rights, NATO, conservative views on family planning, freedom fighters, and even aid to Northern Ireland. The advocacy vacuum for the program as a whole is filled by situation-specific rules promulgated in response to this collection of single interests. This creates distortions and hampers both programming and implementation. It also clogs the pipeline and weakens the credibility of US clients such as Sadat and his replacement Hosni Mubarak (1985:207). Furthermore, where the US administration is intent on militarization of foreign aid (currently and historically in policy toward Israel 1969–76), Congress often reacts to Executive constitutional power as “commander-in-chief” (Fuerwerger, 1979:180). The guardian and spender roles in foreign aid are mixed and situational, placing the burden for programming and execution on USAID which has little authority over the program as a whole. Such policy-making by default works against a coherent foreign aid program that achieves intended results. One example of distortions in the process leading to unsatisfactory results was US aid to Iran. Local uncertainties, coupled with US bureaucratic rigidities led to lengthy support of the Shah with military aid and loss of leverage in ability to control its use. The long history of aid to Iran ($1.0 billion from 1962–78 of which 80% was military aid; USAID, 1984:16) solidified the Shah’s position among US foreign aid policy-makers. The Shah also purchased $3 billion in US arms which made him a powerful customer in US policy circles. The senior State-CIA staff as well as President Carter remained behind the Shah despite growing

US Aid to the Middle East 165 evidence to junior staffers and field personnel that aid was being misused (given to the “secret police”, SAVAK, and death squads working to torture and oppress Iranians) and that official information provided by the Shah’s institutions to the US was inaccurate (Armstrong, cited in Stillman, 1983:433). These actors assumed varying degrees of advocacy and guardianship on the issue of continued US support for the Shah. But it was clear that supporters of the Shah held firm not from eagerness to find political “reality” in Iran. Realistic conflict did not take place because advocates of aid to the Shah were not required to face available information and the remainder was distorted to make him appear in control. Senior staffers held to the view that small groups of “outside agitators” were opposed to the Shah and created security problems for him. Their position was not based on critical examination of the evidence at the time. By contrast, as indicated in the last chapter, President Reagan’s response to massive evidence available from all levels that Marcos was insupportable demonstrates that realistic role conflict can eventually produce the correct decision. Examination of the Iranian aid decision as a dynamic problem of role conflict enables one to see the consequences of mixed or ill-defined roles. Distrust among US policy-makers distorted the information needed to make the appropriate decision. However, this is only one interpretation. Seitz (1980), for example, viewed the problem of failed USAID technical assistance efforts to Iranian ministries for general public administration (1956–61) and the secret police (SAVAK) (1955–63) in a different light. He suggested that these projects failed because “American public administration advisers were generally ignorant of the political implications of their work” (1980:409). He was incredulous that the US would undertake such assistance where “high-level political support for administrative reform did not exist” (1980:409). He suggested also that general advisor ignorance was based on the “explicit AID assumption that technical assistance was nonpolitical” (1980:410). One the basis of the Iranian case, he concluded that “we do not know how to help developing nations reform their administrative structures” (1980:411). Much of this analysis is plausible. Many AID contractors are indifferent to local political complexities and often seek only their check and a plane ticket out after doing their “reports”. But the conclusion that ignorance caused the result is based on the “know nothing” idea proffered in Chapter 2 that more knowledge could have prevented failure— but American institutions are simply immune to such knowledge. Unfortunately, local knowledge does not always lead to positive results. One of the most effective Mission Directors in Ecuador did not speak Spanish but he knew the US foreign aid budget process and maximized resources for local projects. Conversely, most Central American “political basket cases” are filled with US experts on every conceivable detail of culture, language and political economy. Many of these people have worked for AID, were in the Peace Corps, and were given short-term contracts by AID Missions to try out innovative schemes in these countries. Lack of knowledge, experience, and sensitivity was simply not the cause of the Iranian problem, or any other foreign aid failure. Rather the problem has been organization of knowledge and its effective use in programming and implementing projects. Our thesis has been that “we do know how” but conflict over what we do must be realistic, or fed by clear guardian and spender roles. This did not happen in Iran or in many other regional examples. Each institutional actor believed that he/she knew the “reality”. But only some

166

The Politics of United States Foreign Aid

of them could base their position on substantive facts. Hence, the failure that could not be prevented by training or institution-building strategies was, in final analysis, a problem of information distortion by actors of differential power and role clarity.

Conclusion In sum, successful technical assistance projects, food aid, balance of payment support, and trade depends not merely on the “terms of trade” (bases for aid) or more US knowledge and sensitivity to local cultures. The ultimate effectiveness of aid to achieve security, profitability, and development objectives in this region depends on mutual agreement and trust. This mutuality is established not by laws and regulations or obedience to US bureaucratic routines, but rather from intense and realistic sparring about the issues over time. For this reason, Egyptian-US relations have evolved to an all-time high, and have been since Kissinger-Sadat trust was achieved in 1974 following the Syrian-Israeli disengagement accord (Burns, 1985:181). Nasser and his successors have been sparring with the US over food aid, military aid, refugee issues and the scope and purpose of economic aid projects for over 30 years. By contrast, aid to Israel has been based less on realistic conflict than its “special relationship” with the US and historical sympathy for the Jewish plight. This means that aid provides only marginal leverage and, of course, is not programmed or executed in the traditional sense by AID. The US has even less realistic relationships with the rest of the Middle East, in part because of its Israeli client. This has assured that regional aid policy is the result of intense political rivalries and bureaucratic in-fighting. Intense conflict can lead to improved role definition, trust and better results (as in Egypt) or maintenance of the status quo (and more Irans).

Chapter Seven CONCLUSION

Introduction In this final chapter we summarize what has been said so far, review the major problems with foreign aid policy as revealed through the BRC model, and make recommendations for improving both the process and results. Foreign aid is planned and funded like most US domestic policies, but it is unique in that it has both a domestic and international component. Since foreign aid (military, food, development) is the most tangible vehicle of US foreign policy, it receives most of the blame (as a perennially unpopular congressional issue) but little of the praise because its results lack visibility and attribution. Lacking a political base of its own (most of its clients are abroad; powerful domestic clients of its subproducts distort results), foreign aid has been moved along by conflicting pressures from all directions, mostly in response to some actor’s (DOD, State, Presidency, Congress) perception of military threats or developmental need. Decisions have not generally been informed by realistic field assessments, largely because the institutions making foreign aid policy are primarily interested in other issues (DOD-defense; Congress-domestic policy; USDA-US agriculture, etc.). Foreign aid becomes an unstable byproduct of other policies. Nearly 35 years ago, political theorist Verne Lewis (cited in Lyden and Miller, 1982:264) was hopeful that improving the objectivity of budgetary analysis would permit decisionmakers to compare relative effectiveness of economic and military aid programs to the achievement of “national security”. In the context of intense budgetary politicking by all foreign aid actors, several attempts along the “rationality” front were made. AID introduced PPBS in the early 1970s. The language and routines of program budgeting still permeate AID decision-making, contributing to better understanding of input-output relationships, and permitting more incisive inquiries. At a higher level of aggregation, State Department developed the Integrated Foreign Aid Budget Request that included the SAPWRG process which coordinated the development of requests for all foreign aid programs. But State Department sets most of the funding ceilings and program proportions in advance. Consensus and trust may have increased among foreign aid actors, but results may not have improved because communication and conflict is still unrealistic—the product of “frozen pluralism” and immense power differentials among actors. The foreign aid budget is made, then, in true “incremental” fashion by mixed role actors that jockey for marginal gains while “guardians” examine the controversial aid issues only, and “advocates” rely on Presidential, DOD and Congressional support for budget authority. But, given the mixed role conflict, the process tends to exclude the value of implementation efficiency and effectiveness. Such problems can be best resolved by structural adjustments from Congress to give USAID greater input and control over results. DOI: 10.4324/9780203840184-7

168

The Politics of United States Foreign Aid

Summary and Conclusion From World War II to the present, the cluster of programs called foreign aid have moved all over the board, emphasizing profitability, welfare, and security objectives at different times and places almost without pattern. Since the path by which aid can lead to “development” is debatable, objective assessment of the results of foreign aid over its 5 evolutionary historical phases has been difficult. The first phase of US aid, Postwar Relief, was unusual in that a fiscally conservative Congress (fearing gold drainage) multilateralized US aid funding which financed predominantly humanitarian assistance to Europe. Aid policy was defined by Allied needs to rebuild war-torn economies on which massive political support and consensus existed in the US. Only on contemporary defense aid to Israel has the popularity of a foreign aid issue ever peaked this high with Congress and the US public. From 1949–1960, foreign aid was primarily the vehicle for establishing military alliances against the Soviets. The twin incentives of military (stick) and economic (carrot) aid were used to “contain” the Soviets along “frontiers” established largely in the minds of several US presidents. Given the palpable failure of this simplistic security approach to developing capitalist countries and building true military-trade alliances, the third phase (1961–72) stressed exportation of US liberal political ideas via the Alliance for Progress. The ambivalent application of liberal principles where revolutionary changes actually occurred resulted in return to more traditional uses of aid as a cold war leverage instrument. By 1973, the Alliance and prior cold war aid had been lumped together as the traditional model—top down, capital intensive, with benefits “trickling down” to the poor. The 1973 New Directions effort of the 4th phase stressed “bottom up” development via aid programmed to the poor in smaller projects, executed often beyond the reach of large donor or recipient bureaucracies in PVOs. Given the fact that crisis-driven foreign policies governed the overall uses of foreign aid (evident in State Department budgetary ceilings) in both traditional and New Directions periods, the comparative results of specific aid expenditures have been difficult to isolate and attribute to aid policies. The contemporary Reagan era rhetorically stresses the private sector and PVOs, and views government as a constraint on economic development (capitalist) in the Third World. In practice, aid is still programmed by crisis to countries and regions. Though ESF and MAP expenditures remain large, DA projects seem to be decreasing in size and focusing more on the profitabilty and entrepreneurship of both private and public sectors than before. Though profound disagreement exists on the results of even a single aid project, it was suggested that general results could be lumped into three categories: (1) recipient dependency, (2) technocratic and complex products transferred, and (3) multiple unexpected consequences. Occasionally aid has not produced recipient dependency but rather, interdependence; nor complexity but rather, appropriate scale “user friendly” projects, and fewer unintended or unexpected results. On the other hand, dependency, complexity and unexpected results have occasionally been positive and led to growth and development! A major purpose of the book was to explain some of these field differences by stressing US policy-making as the major determinant. In Guatemala, for instance, US military aid reinforced US ties to unelected military regimes which depended largely on military aid for their legitimacy and continued existence. Further, despite enormous amounts of DA to Nicaragua, an unintended consequence was

Conclusion

169

that the landless rural labor force was 1000% higher in 1977 than in 1950. The red tape and complexity of US aid, requiring full recipient understanding of the US budget process, often produces capital intensive glamor projects that fragment the local bureaucracies into state enterprises and autonomous institutions. This creates a public sector coordination problem for recipients as US-financed projects push ahead in isolated sectors. Nevertheless, aid to Costa Rica, Brazil and Colombia produced many more successes and fewer instances of dependency, complexity or unintended consequences. Aid to Haiti even produced the leverage that led to Duvalier’s demise! In Asia, aid produced dependency and little leverage in Vietnam. But aid to Taiwan and South Korea was quite successful in contributing to economic growth. Aid to the Philippines contributed to Marcos’ downfall as well. In the Middle East, US aid often “delegitimized” regimes in India, Egypt, and Turkey, making it hard for them to govern. The US sought dramatic and complex projects like the Aswan High dam, and through aid, sought to spread the effects of US-Israeli interdependence around other states like Saudi Arabia, Jordan, and Egypt with success in the latter case only. US food aid provided leverage and nutritional well-being in Egypt and encouraged Egyptian concessions to Israel, contributing to the peace necessary for foreign aid to achieve any kind of success. The BRC model (Figure 3.1) was proffered as a conceptual vehicle by which more precise explanations of these aid results might be advanced. The model suggests that certain goal combinations drive aid in contradictory directions. Aid programs are translated into policy by (1) legislative and bureaucratic rules and repertoires which constrain action, and (2) role conflict among institutional actors of widespread differences in power. The efficacy of security, profitability, or developmental aid goals depends upon the utility of the operational rules and the degree of rival trust that exists between actors. Unfortunately, the rules have been accumulated according to specific congressional concerns over time (human rights, AIPAC interests), contributing to an almost permanent power imbalance where the level of distrust among actors is higher than ever (US versus Latin America on trade protection and debt repayment; Congress versus President over G-R-H; authorizations versus appropriations committees where the latter now makes legislation; and Congressional earmarks versus USAID needs on the use of aid funds). Foreign aid actor conflict over these kinds of issues can evolve into realistic conflict and rival trust, with more positive aid results. Conflict tends to be incremental, which is useful in avoiding big errors and assuring continued communication, but costly in excluding key values such as implementation. Where properly managed for long-term political considerations, aid can facilitate rival trust and contribute to development. Though more precise research is needed on the interaction of these variables in comparative project perspective, application of the BRC perspective to Latin American, Asian and Middle Eastern instances suggested general verification of the variables and provided a broad explanation of differences between success and failure. Based on the BRC model, conclusions can be reached and recommendations proffered for (1) achieving better aid results, (2) clarifying goals appropriate to differing country problems, (3) modifying internal and external operating rules, and (4) improving the relationship between institutional actors and policy results. For example, from the regional aid reviews several propositions can be formulated on the results of economic and military aid. First, we have seen that military aid can be a two-edged sword, increasing or decreasing

170

The Politics of United States Foreign Aid

US leverage and recipient regime legitimacy. Pouring military aid into personalistic and authoritarian regimes such as Haiti and the Philippines for vague and doctrinaire security objectives, will provide carte blanche for misappropriation of funds with much greater political than financial (unprofitability) costs to the US. Though increasing military aid to a greater percentage of the local budget, and to the mythical “center” of the military should bring leverage, where consensus between US actors and recipient political/ economic objectives is initially lacking or unclear, the aid becomes a “sunk cost”, useful to the recipient and regime critics as a counter-control device against US pressures. For instance, decades of military aid to Somoza for stability purposes, ignored the need to develop effective opposition to his regime. Failure to direct aid beyond the ruling family or junta creates a political vacuum that is always filled by extremists of the left and/or right. Further, “aid” to Duvalier increased his repressive capability. Though many of the DA projects were extremely successful by narrow technical criteria, the aid lesson from Haiti is that authoritarian regimes negate these successes. In Haiti, no amount of successful DA projects could compensate for the terror and overt corruption of the regime which locals perceived correctly to be US-financed. Here, the US paid twice for Haitian foreign aid— once for DA projects, then for political refugees in Miami! Conversely, military aid for specific purposes, such as stabilizing a coalition, on which the recipients can at least agree to disagree clearly and rationally, can stabilize a political system and prevent extremist sniping and blackmail at centrist programs. A lesson of aid to El Salvador seems to be that properly targeted and monitored, military aid can be effective in achieving “soft” objectives, despite decades of civil war. Additionally, despite accounting norms, net political results can be positive even if gross aid exceeds local financial and institutional absorptive capacity and produces waste and more corruption. The key seems to be prior mutual consensus among US actors, and between the US and recipient on goals and implementation. Second, results can be improved by redefining the role of foreign aid in the foreign policy process. Though total foreign aid amounts to less than 2% of the annual US federal budget, its consequences are far greater for US “national security” and global trade objectives. If a few billion dollars in PL 480 and ESF aid to Egypt and Israel can contribute to holding the Middle East together, this keeps the oil flowing and prevents inflation, unemployment and political unrest in the US as well. However, aid policy-making inputs are distorted and tend to interfere with its real responsibilities around the world. While the congressional foreign aid authorizations process may be enhanced by new leadership, e. g. Fascell and Lugar on the respective committees, other congressional oddities such as G-R-H remove policy-making power from Congress and shift more of it to the President where foreign policy power already existed. This increases the disjointed tendency of the foreign aid program to move ahead according to the politics of supplemental appropriations, earmarks and continuing resolutions, all of which de-emphasize debate and analysis. These practices inhibit the development of rival trust among the relevant US institutions which could improve foreign aid results. Such newsworthy issues as “aid” to the Contras absorb hours of debate, despite the fact this has little to do with foreign aid. Rather, it is an issue of war powers, with management and funding by DOD and CIA, distribution only by USAID. Nevertheless, the public confuses Reagan hyperbole on aid to the Contras (like the French resistance; conflict only 2 days drive from Texas border, etc.) with foreign aid generally. The foreign aid process should be able to defend itself against being used for such transparent

Conclusion

171

purposes. US aid was discredited for a time by the Mitrione affair in Uruguay in the late 1960s, where a CIA operative used USAID cover. The foreign aid policy process should be able to debate the issue squarely before funding and responsibility follow. Where were the “freedom-loving” Contras during Somoza’s reign? How did these exiles, many of whom were officers in Somoza’s National Guard, support the Chamorro opposition then? Since they were clearly co-opted and did nothing to aid freedom then, why support them now in the cause of creating yet another regime in Nicaragua? Such issues need to be debated by foreign aid actors if funds are to be termed “aid”. Otherwise, the support should be classified as Defense or Foreign policy, not foreign aid. Thus, aid results could be improved by its formal dissociation from foreign/defense policy. Third, aid results can be enhanced by increased emphasis on implementation of projects. Since projects are often learning experiences in themselves with many unexpected consequences, it would be more useful if AID were directly responsible for both programming and execution. Currently, the knowledge that private contractors implement aid with only periodic official oversight tends to weaken AID incentives to develop innovative project implementation approaches based on lessons learned. While much has been written on this, little finds its way into institutional routines and repertoires. For example, roles might be reversed with private contractors overseeing and evaluating AID project implementation. AID would gain corporate memory and experience; contractors could tear into AID activities with a vengeance that would ultimately enhance results. Both parties would have every incentive to perform; greater trust and a better working relationship would result. The field results of US aid are a product of goals translated by institutional rules and roles. The predominant goal of the foreign aid program has been security, and this definition is largely the product of inordinate DOD, Armed Services Committee and State influence over the scope of the program. DA has its advocates in the process: USAID and multilateral agencies, such as OAS and World Bank. Early postwar US aid to Europe was both developmental and multilateral. In a stable world, unsullied by petty tyrants and terrorists, total DA emphasis would be appropriate. The US could provide capital grants and technical assistance for projects similar to the US Office of Economic Opportunity (OEO), Economic Development Administration (EDA), and urban redevelopment. Security problems, of course, have had little to do with the success or failure of federal poverty and development programs, other than where a housing project is overrun with crime. But in the Third World, projects are not implemented in a stable federal system where roles are relatively clear and conflict eventually leads to rival trust (or program termination before it occurs). Aid projects in El Salvador, for example, must also contend with left (FMLN) and right (ARENA) extremists who fear loss of support from successful US aid projects. The “outside agitator” premise on communist infiltration is obviously exaggerated and produces an almost total concern with East-West issues when the substantive issues are local (Turkey versus Greece over Cyprus). Nevertheless, outside agitators seem to make unscheduled appearances to fill US-created political vacuums in places like Nicaragua. Hence, the solution to what become regional security problems, after years of neglect and indifference to developmental issues, has to be something more than disbursement of more DA funds. While this dilemma should not lead to support of additional outside agitators, such as aid to roving bands of exiles from the last coup, security assistance to the military “centers” of neighboring countries to build

172

The Politics of United States Foreign Aid

up political centers as in El Salvador, cannot be resonably opposed. Even more important than the appropriate mix of development-security goals behind aid is the degree of prior mutual consensus between recipient and US aid planners. Where the recipient agrees on the purposes of aid, improved results usually follow (Taiwan, South Korea, Egypt and Brazil). Whatever form the aid takes (ESF, PL 480, FMSC, MAP or DA), success depends less on the right mixture of aid subprograms than on mutual agreement on how the funds should be spent properly. What this suggests is that even abstract “liberal” premises of aid may be appropriate if both recipient and US agree to their translation into programs meeting local needs. Aid to Costa Rica, Venezuela, and Colombia for “liberal” political purposes such as electoral democracy worked reasonably well because both parties to the transaction agreed, i. e. aid was not perceived to be a mere quid pro quo for US leverage. In theory then, all aid need not be either DA or military aid to achieve success. On the other hand, under current usage of the term, “aid” means all things to all people—it builds clinics, supports Contras, trains teachers, mines ports, and provides arms to petty tyrants. To improve issue analysis and program marketability to the US public, foreign aid should be divided into clearer responsibility centers. USAID should be responsible for PL 480 and DA aid. DOD should be responsible, i. e pay for, all military aid. ESF funds fall in the gray area between the two. AID manages them with State guidance, and many are spent for defense-related purposes. Hence, it may be more appropriate to draw ESF funds from the DOD budget and let AID manage them for agreed upon purposes. This would leverage executive policy-making control away from DOD-CIA-OMB technocrats and State senior staffers. With greater authority to influence development of the Integrated Request each year, those in AID who work daily on both foreign aid and its political implications would have more voice in policy-making. This additional authority would remove some of the incentive to institutionalize criticism that detracts from program implementation. As noted, AID receives most of the responsibility and criticism for the total foreign aid program with little effective authority to control its scope or direction. For the late 198.s, the goals of US foreign aid are likely to remain the same except that the notion of “profitability” will lead to even greater unexpected political results. Increasing evidence that rightist authoritarian regimes have neither more rational economic policies nor greater interest in US-style political democracy have led to the logical question: For what was all the US aid used then? In Haiti and the Philippines the answer is clearly: personal profit for the ruling kleptocracy and oppression of legitimate opposition to the regime. That such uses of foreign aid are both financially and politically unprofitable for the US is evident now to even the Reagan Administration. Other Republican leaders such as Undersecretary of Treasury for International Affairs David Mulford, have recently opposed loans to Chile on human rights grounds. The point is that viewed by key policy actors as a high risk political and financial proposition, aid can achieve liberal developmental ends. To this end, more policy analysis and issue debate is required to avoid incrementally recycling the mistakes of the past. One can expect to see increased emphasis in time, resources, and media interest in foreign aid authorizations and appropriations which can only benefit future aid results. Even if initial mutual agreement could be reached between US and recipients on the scope, purpose and pace of aid, the current rules by which foreign aid is planned and implemented would continue to constrain effective action. We have seen that externally,

Conclusion

173

foreign aid must deal with GR-H “outlay savings” zeal and that this penalizes early year, rapid paced outlays like military aid or capital projects. To subject an already confused program, promulgated annually from the conflicting pressures of institutional actors with entirely different objectives and grasps on reality, to arbitrary across-the-board cuts for the “abstract” objective of a balanced budget is both inefficient, and naive. For, in the long run, such “savings” are likely to contribute to greater deficits as required military expenditures increase in world trouble spots to put out fires that earlier aid might have remedied by providing leverage and construction of stable political centers. A lesson of El Salvador is that even under textbook conditions, building the center is a costly and longterm commitment. Similarly, Congress tries to make foreign aid policy via continuing resolutions, earmarks, supplementals, conditions and prohibitions on executive requests. The executive responds by “counter-control” techniques such as reprogramming, transfer, deferral and rescission to make ends meet and preserve managerial flexibility. Though the executive usually “wins” such bureaucratic wars, they waste time and resources that could be spent on effective implementation. These kinds of “wargames”, made even more deadly by resource cutbacks likely to result from G-R-H, also diminish time available for debate on regions and countries that do not have current “hot spot” status. Earmarks and other constraints are expressions of congressional distrust of the President, learned from such historical crises as Vietnam; transfers and deferrals are expressions of executive distrust for a Congress perceived to be sluggish and ill-informed on international issues. This tension cannot lead to healthier debate and better policy results until adversary roles and responsibilities are clarified. The arbitrary interventions by Congress then encourage AID to create a defensive superstructure of rules. Such rules stress elaborate programming over innovative implementation and this damages foreign aid results. However, stronger leadership on the authorizing committees and appropriations subcommittees is already enhancing the role of congressional guardianship in foreign aid. Needed is a more objective standard for aid disbursement than the political visissitudes of the moment. For example, the appropriations process could be strengthened by use of agreed -upon guidelines or a “formula” for disbursement of aid. Simon (Washington Post, June 14, 1975) argues that foreign aid is disbursed largely on “whim or casual conversations of a Secretary of State with a foreign aid official.” He suggests that “no other government program provides as unrestrained an opportunity for executive expenditure as does our foreign economic aid.” Use of a “needs based” formula that also stresses “restraint on military spending” and “respect for civil liberties”, however, might increase congressional propensity to earmark. But it could also provide a mutual focus for bureaucratic rivals, reducing fragmentation and distrust. The rules of the game for the recipient side have also been discussed briefly. ESF aid, for example, is deposited in recipient central banks for withdrawl according to agreedupon purposes. Usually, as in the case of Marcos officials using aid funds for his re-election campaign, there is no specific mutual agreement, especially where base rights are the quid pro quo. Hence, cronies of the regime can spend millions in US funds for personal preferences (often consumer and military goods) and the US usually does not dramatically intercede from fear of neo-imperialist backlash.

174

The Politics of United States Foreign Aid

In the case of Taiwan, US foreign aid was placed in a separate extrabudgetary recipient’s fund. In contrast with AID policy in many countries where US dollar aid or aid-generated local currency becomes part of the governmental budget, aid to Taiwan was administered outside the budget by semi-autonomous institutions. During the 1951–65 period, Chiang Kai-shek wanted to increase military expenditures to recover Mainland China from the Communists (a personal agenda). Through the extrabudgetary fund mechanism, the US was successful in both keeping the lid on Taiwanese military expenditures and preventing fungibility of funds for development purposes (Jacoby, 1966:221–222). This mechanism prevented substitution of military funds demanded by the powerful military establishment for its “development” priorities. The key to effectiveness of US aid to Taiwan was also the high degree of competence and personnel continuity of local oversight institutions— the Council on International Economic Cooperation and Development, and the Joint Commission on Rural Reconstruction (1966:59). Where large amounts of US aid tax the absorptive capacities of small regimes often besieged by a variety of opponents, opportunities for fund diversion are enormous. Since this costs the US leverage as well as out of pocket sums, both AID and Congress should ensure that aid is not disbursed before: (1) mutual agreement on its uses (meaning more recipient input into project shape and on the participation of US contractors) and (2) agreement on extrabudgetary funds and oversight commissions are reached. The most serious constraint and the greatest opportunity for improving foreign aid results lies in role redefinition and clarification. Currently, foreign aid policy is formulated by a host of bureaucratic spenders with different primary missions and different sets of assumptions. Though such diversity can be good, differences in institutional power prevent translation of differing premises into a coherent policy product. Guardians such as Congress act significantly during crises but do not exercise consistent programmatic oversight. Congressional foreign aid policy-making thus lacks coherence. Operating in piecemeal fashion, this permits the most powerful actors in the executive branch to exercise primary control. AID has the least influence on practically every issue, meaning that despite AID’s elaborate programming and the apparent consensus generated recently by the SAPWRG process for an Integrated Request, final policy is still a product of power politics between DOD-Armed Services Committees-contractor lobbies and softer perspectives exercised by weaker actors such as AID and State. Agency positions on aid cannot always be predicted. But AID’s influence is marginal, limited often to siding with the agency likely to carry the day in order to reduce external criticism and ensure future funding (continuing resolutions). Shifting patterns of dominance between the executive generally and Congress on particular issues leads to many unexpected consequences and the establishment of rigid ties with client regimes that can produce trouble later.

Recommendations Clearer definition of guardian and spender roles can be accomplished by relatively minor structural changes in the foreign aid policy process. Currently, guardian and spender conflict-communication is impeded by almost permanent power imbalances. DOD access to the President is greater than either AID or State. This is due in part to powerful contractors (companies like RCA and Lockheed through the Armed Services Committees) as well as

Conclusion

175

the institutional access given to the Presidential National Security Advisor but not to the Secretary of State or AID Administrator. Policy is determined less according to the merits of economic or defense needs than interest group-driven machinery—as one would expect in American politics. The balance of power should therefore be formally shifted to give the “soft” or economic side greater access which would improve the substance of guardianspender conflict. The dividends would be better foreign aid programming and perhaps fewer superficial aid policies, such as aid to Nicaraguan Contras, Marcos, Duvalier, and every Middle Eastern country that threatens war. First, congressional guardian and foreign aid advocacy roles can be clarified by reducing the noise produced by the system and sharpening the analysis prior to decision. Authorizations and appropriations hearings should “zero base” foreign aid for each country and review it in five year cycles, rather than build from last year’s base that represents the unexamined accumulated funds of the past. This would reduce the incremental tendency to examine only the “hot spots”, leaving the others for future crises. The change of leadership and formats of both appropriations and authorizing committees, together with gradual development of executive-congressional consensus on termination of aid to Marcos and Duvalier, may herald a new era of trust between foreign aid actors. The intense lobbying that would occur for each country would, of course, consume lots of time and make it more difficult to spend time on other matters such as the AID CDSS process. But we noted that the latter is partly a defensive institutional response to uncertain external signals. By focusing congressional attention on countries, the CDSS-ABS process could be directly integrated or “crosswalked” to the congressional policy process, with net time saved. Public attention would be focused on foreign aid, which would threaten some in Congress. But overall this new structure would help clarify the roles of both the actors and the types of aid in generating or reducing larger foreign policy crises later. Second, AID should have increased responsibility, authority and visibility in the foreign aid process. While AID is an advocate of more aid appropriations, it should also have more authority to oversee development of the entire aid program, military and economic. In short, as intimated above, AID should have the “guardian” role of co-ordinating foreign aid programming instead of State Department. Given its extensive field experience in technical assistance as well as developing CDSSs on larger policy issues, AID should be the best—equipped actor to gauge the scope and direction of the program. Accordingly, State should not be able to set budget ceilings for AID. Further, AID should be able to defer budget authority (deobligate and reobligate) by Mission where evidence exists that funds are or will be improperly spent. Such authority would strike terror into the heart of every US contractor and recipient counterpart agency with plans to lock AID into expenditures that it must make but would rather not. This kind of authority in Vietnam might have made a significant difference to South Vietnamese government insensitivity to US concerns and could only have improved what was to come. In short, AID should be more influential in programming ESF, FMSC, MAP, DA and PL 480 to each country. This added authority would lead to greater emphasis by AID on implementation. Finally, clearer role definition can be accomplished by development of a decision process that stimulates realistic communication and trust. It was noted in Chapter 3 that State Department developed the Integrated Budget Request process beginning in FY 82 that consists of two subprocesses. First, the Secretary’s Assistant Policy Review Working

176

The Politics of United States Foreign Aid

Group or SAPRWG develops the International Security Assistance (ISA) side of foreign aid, co-ordinating requests by the Defense Security Assistance Agency and the ISA Bureau of DOD and State Regional Bureaus. Second, and simultaneously, USAID develops its request for DA and PL 480 aid. The processes dovetail on programming and use of ESF. State Department, as noted, takes final responsibility for ESF. The Integrated Foreign Aid Budget Request consists of major inputs from the SAPRWG, USAID and to a lesser extent USDA (on PL 480) and Treasury. Despite evidence of increasing consensus from narrowing differences between requests and appropriations (Table 3.1), it is unclear whether the new decision process has improved aid policy-making by shifting resources to areas of greater need, or simply serves as a vehicle for generating consensus behind the AdministrationDOD program for foreign aid. Despite enormous differences in agency power and therefore influence over results, the concept of an improved forum has much merit. The idea behind the Integrated Foreign Aid Budget Request and SAPRWG should be expanded to include Congress as well. To this end, recently Senator John Warner (R-VA), and Representatives Robert McEwen (R-Ohio) and Richard Cheney (R-WYO) sponsored the “National Security Act” attached to the recent DOD reorganization act. It would require the administration to submit a comprehensive written report on the national security strategy of the US, including a rundown on vital interests, goals, and objectives of foreign policy. The act would mandate a process by which the Secretaries of State and Defense would testify on the report each year at the beginning of congressional sessions to joint meetings of the Foreign Affairs and Armed Services Committees of both House and Senate. To the extent such a process focused on all countries in a cyclical review, it could provide Congress: “a basic yardstick against which to measure the myriad individual decisions it makes on foreign aid, military assistance, and defense appropriations. Although foreign policy goals logically should determine defense strategy, the Armed Services and Foreign Affairs committees almost never meet or do business with each other” (Broder, Washington Post, July 6, 1986). Such a process could clarify roles, forcing a more coherent debate over differences within the administration and perhaps a reconciliation of viewpoints. For example, the process could reduce the congressional tendency to throw money at all Middle East actors in a continuing extortion game abetted by DOD-contractor interests in arming the region. It is difficult to recommend the proper combination of Middle East aid before the fact, because like all public policy, the program is a hypothesis with many unexpected consequences. But improved policy-making via this forum might also prevent the wastage of funds that occurs in Central America (El Salvador, Belize) and the Middle East (Egypt) when larger foreign policies, e. g. the Caribbean Basin Initiative, push funds on a country before AID or the recipient can properly plan needed projects. It could prevent the use of “aid” for covert purposes where the CIA “manages” the Contra program using DOD and CIA inputs as well as Congressionally-appropriated non-lethal funds, while USAID “distributes” $300 million in economic aid (Omang, Washington Post, July 13, 1986). The advantage of such a process is that it would place information directly into the power machinery that now distorts the incentives for improved policy-making and execution. Such a process could end the incremental tendency to go from “crisis to crisis and parochial interest to parochial interest”, which may serve to stabilize the rest of the budget process, but is extremely dangerous when dealing with security and development issues. Though

Conclusion

177

billed as a means of forcing “a degree of rationality and planning into the foreign policydefense field” (Broder, Washington Post, July 6, 1986), its major advantage would be to allow foreign aid policy actors to play power politics more realistically, less secretively, and with better information. As Cassen suggested recently (1986:12), we should appreciate what aid can and cannot do to reduce poverty and affect development. The aggregate programs of US foreign aid have always had broader goals which makes attribution of results to them difficult. Nevertheless, clarification of guardian and spender roles by such means as the proposed National Security Act can enhance aid results by reducing the range of distrust over key issues. Both information and incentives for its use in decision-making would be improved which can only help the foreign aid program do what it is supposed to do.

BIBLIOGRAPHY

Adler, J.H. ABSORPTIVE CAPACITY: THE CONCEPT AND ITS DETERMINANTS (Brookings Institution, Washington, D.C., 1965) Alexander, A. “US Withholds $13 Million In Aid From Zimbabwe Over ‘Insult’” THE ATLANTA CONSTITUTION (July 11, 1986), p. 11A. Albright, J. “Effort To Free Sudan of Thirst Resulted In ‘Circle of Death’” THE ATLANTA CONSTITUTION (December 9, 1985), p. 1A. Allison, G.T. THE ESSENCE OF DECISION: EXPLAINING THE CUBAN MISSILE CRISIS (Little, Brown, Boston, 1971). Anthony, R.N. and Young, D.W. MANAGEMENT CONTROL IN NONPROFIT ORGANIZATIONS (Irwin, Homewood, Ill., 1984). Apple, R.W. “For The US The Toughest Philippine Policy Decisions May Yet Lie Ahead” THE NEW YORK TIMES (February 26, 1986), p. 9. Armstrong, N. “The Fall of the Shah”, in Stillman, R.J. (ed.) PUBLIC ADMINISTRATION, CONCEPTS AND CASES, 3rd edn (Houghton Mifflin, Boston, 1984), pp. 426–448. Arrow, K.J. SOCIAL CHOICE AND INDIVIDUAL VALUES (John Wiley, New York, 1951). Asher, R.E. GRANTS, LOANS, AND LOCAL CURRENCIES, THEIR ROLE IN FOREIGN AID (Brookings Institution, Washington, D.C., 1961). Atlanta Constitution, “Young Entrepreneur Spins New Business In Mill City” (December 25, 1985), p. 4A. _____ “Military Aid Cost US $120 Billion Since 1950, And The Price Is Rising” (December 29, 1985), p. 8A. _____ “Korea’s Chun Hears Whisper Of The Axe” (February 26, 1986), p. 15A. Auerbach, S. “Farm Subsidies Have Global Impact, Developing Nations Say US, EC Policies Undermine Their Efforts To Compete In World Markets” THE WASHINGTON POST, July 13, 1986, Gl. ______ “South Korea: New Trade Threat” THE WASHINGTON POST (February 9, 1986), p. B1. Bauer, P.T. and Yamey, B.S. “Foreign Aid: What Is At Stake?” in Thompson, W.S. (ed) THE THIRD WORLD, PREMISES OF US POLICY (Institute For Contemporary Studies, San Francisco, 1983), pp. 75–115. Becker, E. “Hanoi Tries Profits In Bid To Boost Economy” THE WASHINGTON POST (February 23, 1983), p. A20. Boston Globe “Brazil Resists IMF Policies” (November 23, 1985), p. 11. Brady, N.C., “Statement of Nyle C.Brady Senior Assistant Administrator For Science and Technology, USAID, Before the Committee on Foreign Relations, US Senate” (March 19, 1986). Branigin, W. “Insurgency, Economic Crisis Threaten US Ally” THE WASHINGTON POST (August 12, 1984), p. A1. ______ “Manila Military Embattled” THE WASHINGTON POST (March 25, 1984), p. A1. ______ ‘Statistical Improbability’Aquino Blanked in Town in Marcos Country” THE WASHINGTON POST (February 9, 1986), p. A1. Broder, D.S. “Symptoms and A Strategy” THE WASHINGTON POST (July 13, 1986).

Bibliography 179 Brown, L.R. “Reshaping Economic Policies” in Brown, L.R. (ed) STATE OF THE WORLD 1984 (Norton, New York, 1984), pp. 194–210. Burgess, J. “Philippine Economy Is Election Issue, Marcos Widely Cited As Cause of Slump” THE WASHINGTON POST (December 8, 1985), p. A30. Burns, W.J. ECONOMIC AID AND AMERICAN POLICY TOWARD EGYPT, 1955–1981 (State University of New York Press, Albany, 1985). Caiden, N. “Comparing Budget Systems: Budgeting in ASEAN Countries” PUBLIC BUDGETING AND FINANCE, Vol. 5, No. 4 (Winter, 1985), pp. 23–39. _____ and Wildaysky, A. PLANNING AND BUDGETING IN POOR COUNTRIES (John Wiley, New York, 1974). Cannon, L and Hoffman, D. “Hill Warnings Sounded On Contra Aid Package” THE WASHINGTON POST (February 9, 1986), p. A22. Carlucci, F.C. “The Commission on Security And Economic Assistance, Report To The Secretary of State” (Department of State, Washington, D.C., 1983). Carnoy, A. THE GREATEST FAILURE OF AMERICAN FOREIGN AID (Larum, San Francisco, 1972). Cassen, R. “The Effectiveness of Aid, The Record and the Lessons” FINANCE AND DEVELOPMENT, Vol. 23, No. 1 (March, 1986), pp. 11–15. Chardy, A. “Price Tag, Lost Cash Imperil Central American Aid Package” THE MIAMI HERALD (February 25, 1984), p. 18A. Cline, W.R. (ed), POLICY ALTERNATIVES FOR A NEW INTERNATIONAL ECONOMIC ORDER: AN ECONOMIC ANALYSIS (Praeger, New York, 1979). Coffin, F.M. WITNESS FOR AID (Houghton Mifflin, Boston, 1964). Cohen, S.D. THE MAKING OF US INTERNATIONAL ECONOMIC POLICY, PRINCIPLES, PROBLEMS AND PROPOSALS FOR REFORM (Praeger, New York, 1977). Collier, D. (ed.) THE NEW AUTHORITARIANISM IN LATIN AMERICA (Princeton, 1979). Congressional Budget Office of the US, “Assisting The Developing Countries: Foreign Aid and Trade Policies of the US” (Washington, D.C., September, 1980). Congressional Quarterly, “Budget Cuts Leave Mark On Foreign Assistance Arms Aid” (CQ, Washington, D.C., December 21, 1985). _____“Obey: Reagan’s Aid Priorities Spur Doubts About Program” (CQ, Washington, D.C., March 16, 1985). ______“Central American Aid Request” (CQ, Washington, D.C., August 24, 1985). ______“Some Strings On Nicaraguan Aid Eased By Intelligence Measure” (CQ, Washington, D.C., November 23, 1985). _____ “House Votes Amendment-Laden Foreign Aid Bill” (CQ, Washington, D.C., July 13, 1985). _____ “Divided House Panel Approves Spending Bill” (CQ, Washington, D.C., July 20, 1985). _____ “Hill, Administration Keep Heat On Marcos” (CQ, Washington, D.C., November 16, 1986). _____ “Marcos, Under US Pressure, Heads For Test” (CQ, Washington, D.C., November 9, 1986). _____ “Administration Begins Drive For Boost In Middle East Aid” (CQ, Washington, D.C., February 2, 1985). _____ “Israel Returns $51. 6 Million In Foreign Aid For Graham-Rudman Cut Equity” (CQ, Washington, D.C., January 25, 1986). _____ “Schultz Resists Members’Push To Speed Up Extra Israel Aid” (CQ, Washington, D.C., March 9, 1985). Cotler, J. and Fagen, R.R. (eds) LATIN AMERICAAND THE UNITED STATES: THE CHANGING POLITICAL REALITIES (Stanford, 1974). Cowell, A. “Africa Norm: Rule By One” THE NEW YORK TIMES (August 14, 1984), p. 1. Currie, L. THE ROLE OF ECONOMIC ADVISERS IN DEVELOPING COUNTRIES (Greenwood, Westport, 1981).

180

Bibliography

Deely, D. and Guess, G. “A Public Mixed-Enterprise Approach To Forestry and Natural Resource Management For Integrated Rural Development in Ecuador” (USAID, Quito, October, 1980). Deutsch, K.W. “Theories of Imperialism and Neocolonialism” in Rosen, S.J. and Kurth, J.R. (eds) TESTING THEORIES OF ECONOMIC IMPERIALISM (Heath, Lexington, MA, 1974). Dickey, C. “The Gang That Blew Vietnam Goes Latin “THE WASHINGTON POST (November 28, 1982), p. D1. _____ “Heavy US Aid Brings Little Progress” THE WASHINGTON POST (October 2, 1983), p. A1. Diskin, M. “Elections Alone Are No Solution” THE MIAMI HERALD (March 18, 1983), p. 25A. Downs, A. AN ECONOMIC THEORY OF DEMOCRACY (Harper and Row, New York, 1957). _____ INSIDE BUREAUCRACY (Little, Brown, Boston, 1967). Easterbrook, G. “Making Sense of Agriculture, A Revisionist Look At Farm Policy” THE ATLANTIC MONTHLY, Vol. 256, No. 1 (July, 1985), pp. 63–78. Eberstadt, N. and Lewis, C.M. “How Many Are Hungry?” THE ATLANTIC MONTHLY, Vol. 257, No. 5 (May, 1986), pp. 34–38. Eckholm, E. LOSING GROUND, ENVIRONMENTAL STRESS AND WORLD FOOD PROSPECTS (Norton, 1975). Fagen, R.R. (ed) CAPITALISM AND THE STATE IN US-LATIN AMERICAN RELATIONS (Stanford, 1979). Fallows, J. “The Case Against Credentialism” THE ATLANTIC MONTHLY, Vol. 256, No. 6 (December, 1985), pp. 49–74. ______“Defense: The Spend-Up” THE ATLANTIC MONTHLY, Vol. 258, No. l (July, 1986), pp. 27–31. Farnsworth, C. “The Annual Foreign Aid Dance” THE NEW YORK TIMES (May 24, 1985), p. 15. Feinberg, R.E. THE INTEMPERATE ZONE, THE THIRD WORLD CHALLENGE TO US FOREIGN POLICY (Norton, 1983). Feuerwerger, M.C. CONGRESS AND ISRAEL, FOREIGN AID DECISION-MAKING IN THE HOUSE OF REPRESENTATIVES, 1969–1976 (Greenwood, Westport, 1979). Fisher, L. “Annual Authorizations: Durable Roadblocks to Biennial Budgeting” PUBLIC. BUDGETING AND FINANCE, Vol. 3, No. 1 (Spring, 1983), pp. 23–41. Frankel, G. “Mozambique Dam: Death of a Dream” THE WASHINGTON POST (May 18, 1986), p. A25. Fuerbringer, J. “Many In Congress Voice Support For Increase In Aid To Philippines” THE NEW YORK TIMES (February 26, 1986), p. 1. Gelb, L. “10 Years after Vietnam, US A Power In Asia” THE NEW YORK TIMES (April 18, 1985), p. 1. Goldman, M.I. SOVIET FOREIGN AID (Praeger, New York, 1967) Gorman. R.F. (ed.) PRIVATE VOLUNTARYORGANIZATIONS AS AGENTS OF DEVELOPMENT (Westview, Boulder, 1984). Goulet, D. and Hudson, M. THE MYTH OF AID, THE HIDDEN AGENDAOF THE DEVELOPMENT REPORTS (IDOC North America, New York, 1971). Graham, B. “Communists Gain In Chile, Pinochet’s Policies Said To Help Leftists Organize” THE WASHINGTON POST (May 18, 1986), p. A21. Graham, R. “The Rise Of Islamic Fundamentalism” THE WASHINGTON POST BOOK WORLD (December 1, 1985), p. 5 Gray, W.H. “Tough Votes Still Lie Ahead” THE WASHINGTON POST (July 13, 1986), p. B8. Greve, F. “Pentagon Considers A Central American War Winnable’” THE MIAMI HERALD (August. 3, 1983), p. 5A. Gruen, G.E. “Ambivalence In The Alliance: US Interests In The Middle East and the Evolution of Turkish Foreign Policy” ORBIS, Vol. 24 (Summer, 1980), pp. 363–378.

Bibliography 181 Guess, G. “Institution-Building For Development Forestry In Latin America” PUBLIC ADMINISTRATION AND DEVELOPMENT, Vol. 2, No. 4 (October-December, 1982), pp. 309–325. ______“The Political Economy of American Rural Development” PUBLIC ADMINISTRATION REVIEW, Vol. 43, No. 3, May/June, 1983, pp. 277–285. _____“US Is Sucker For An Election” THE MIAMI HERALD (June 7, 1983), p. 17A. ______“Foreign Assistance and The Roots of US Policy Problems in Central America”, unpublished paper presented at International Studies Association meeting (Atlanta, 1984). ______“Bureaucratic-Authoritarianism and The Forest Sector in Latin America”, Technical Paper N. 43 (University of Texas, Institute of Latin American Studies, Austin, 1984). ______“Profitability Guardians and Service Advocates: The Evolution of Amtrak Training” PUBLIC ADMINISTRATION REVIEW, Vol. 44, No. 5, September/October, 1984, pp. 384–393. ______“Role Conflict In Capital Project Implementation: The Case Of Dade County Metrorail” PUBLIC ADMINISTRATION REVIEW, Vol. 45, No. 5, September/October, 1985, pp. 576–585. Gurtov, M. THE UNITED STATES AGAINST THE THIRD WORLD, ANTINATIONALISM AND INTERVENTION (Praeger, New York, 1974). Gwertzman, B. “Schultz Denounces Philippine Fraud” THE NEW YORK TIMES (February 20, 1986), p. A8. Harden, B. “Despite Vast US Aid, Ethiopia Stays Hostile” THE WASHINGTON POST (April 14, 1985), p. A10. Hardin, G. “The Tragedy of the Commons” SCIENCE, Vol. 162 (1968), pp. 1243–1248. Harrison, L.E. “Latin America’s Trouble Isn’t Our Fault” THE WASHINGTON POST (March 30, 1986), p. D7. Herman, E.S. and Brodhead, F. DEMONSTRATION ELECTIONS, US-STAGED ELECTIONS IN THE DOMINICAN REPUBLIC, VIETNAM, AND EL SALVADOR (South End, Boston, 1984). Hirschman, A.O. THE STRATEGY OF ECONOMIC DEVELOPMENT (Yale, New Haven, 1958). _____ A BIAS FOR HOPE, ESSAYS ON DEVELOPMENT AND LATIN AMERICA (Yale, 1971). _____ “The Turn to Authoritarianism in Latin America and the Search for Its Economic Determinants” in Collier, D (ed) THE NEW AUTHORITARIANISM IN LATIN AMERICA (Princeton, 1979). _____ ESSAYS IN TRESPASSING, ECONOMICS TO POLITICS AND BEYOND (Cambridge, 1981). Ho, S.P. S. ECONOMIC DEVELOPMENT OF TAIWAN 1860–1970 (Yale, New Haven, 1978). Hoang Trong, L. “Survival And Self-Reliance: A Vietnamese Viewpoint” ASIAN SURVEY (Apri1, 1975), pp. 281–300. Hoffmann, S. “Realism and Its Discontents” THE ATLANTIC MONTHLY, Vol. 256, N. 5 (November, 1985), pp. 131–136. Hofheinz, R. and Calder, K.E. THE EASTASIA EDGE (Basic Books, New York, 1982). Hough, R.L. ECONO (April, 1975), pp. 281–300. ITY, RETHINKING US POLICY (National Defense University, Washington, D.C., 1982). Hudson, M. “The Political Economy of Foreign Aid” in Goulet, D. and Hudson, M. THE MYTH OF AID, THE HIDDEN AGENDA OF THE DEVELOPMENT REPORTS (IDOC, New York, 1971), pp. 73–121. Ignatius, D. “How To Rebuild Lebanon” FOREIGN AFFAIRS, Vol. 61 (1983), pp. 1140–1156. “Bombing Gadhafi Worked, We Destroyed More Than His House; We Exploded His Myth” THE WASHINGTON POST, July 13, 1986, B5. International Monetary Fund, “Uruguay-Recent Economic Developments” (Washington, D.C., March 20, 1985). Ippolito, D.S. HIDDEN SPENDING, THE POLITICS OF FEDERAL CREDIT PROGRAMS (North Carolina, Chapel Hill, 1984).

182

Bibliography

Jacoby, N.H. US AID TO TAIWAN, A STUDY OF FOREIGN AID, SELF-HELP, AND DEVELOPMENT (Praeger, New York, 1966). Jickling, D. “Development Management in USAID: A Baseline Review”, paper presented at American Society For Public Administration Conference, Anaheim, CA., April 14, 1986. Jones, G. “Military Chiefs Rebel Against Marcos” THE ATLANTA CONSTITUTION (February 23, 1986), p. 1A. Kaiser, R.G. “Is Dependency on U.S. Aid Doing Israel Any Good?” THE MIAMI HERALD (June 3, 1984), p. lE. Kanter, A. DEFENSE POLITICS, A BUDGETARY PERSPECTIVE (University of Chicago, 1979). Kaplan, J.J. THE CHALLENGE OF FOREIGN AID, POLICIES, PROBLEMS, AND POSSIBILITIES (Praeger, New York, 1967). Kemp, J. “Policy Fusion For Central America” THE MIAMI HERALD (August 1, 1983), p. 11A. Kim, K.S. and Roemer, M. GROWTH AND STRUCTURAL TRANSFORMATION, STUDIES IN THE MODERNIZATION OF THE REPUBLIC OF KOREA 1945–1975 (Harvard, Cambridge, 1979). Klare, M.T. and Aronson, C. “Exporting Repression:US Support For Authoritarianism In Latin America” in Fagen, R. (ed) CAPITALISM AND THE STATE IN US-LATIN AMERICAN RELATIONS (Stanford, 1979), pp. 138–168. Krueger, A.O. THE DEVELOPMENTAL ROLE OF THE FOREIGN SECTOR AND AID, STUDIES IN THE MODERNIZATION OF THE REPUBLIC OF KOREA, 1945–1975 (Harvard, Cambridge, 1982). _____ “Import Substitution Versus Export Promotion, Why Outward-Oriented Policies Are Better” FINANCE AND DEVELOPMENT, Vol. 22, No. 2 (June, 1985), pp. 20–24. Kurth, J.R. “Testing Theories of Economic Imperialism” in Rosen, S.J. and Kurth, J.R. TESTING THEORIES OF ECONOMIC IMPERIALISM (Heath, Lexington, MA, 1974). LaFaber, W. INEVITABLE REVOLUTIONS, THE UNITED STATES IN CENTRAL AMERICA (Norton, New York, 1984). Lal, D. “The Misconceptions of’Development Economics’, The Many Shortcomings of the Dirigiste Dogma” FINANCE AND DEVELOPMENT, Vol. 22, No. 2 (June, 1985), pp. 10–14. Lamm, R.D. “Attach Strings To Emergency Relief” THE NEW YORK TIMES (April 17, 1985), p. 15. Lappe, F.M., Collins, J., andKinley, D. AID AS OBSTACLE, TWENTY QUESTIONS ABOUT OUR FOREIGN AID AND THE HUNGRY (Institute for Food and Development Policy, San Francisco, 1980). Leigh, C. and Alexander, A. “‘Progress’ In Developing Countries Leads To Disappearance of Forests” THE ATLANTA CONSTITUTION (December 10, 1985), p. 1A. Lewis, J.P. and Kallab, V (eds) US FOREIGN POLICY AND THE THIRD WORLD, AGENDA 1983 (Praeger, New York, 1983). Lewis, V.A. “Toward A Theory of Budgeting” in Lyden, F.J. and Miller, E.G. (eds) PUBLIC BUDGETING, PROGRAM PLANNING AND IMPLEMENTATION, 4th edn (Prentice-Hall Englewood Cliffs, N.J., 1982). Lieber, J. “Coping With Cocaine” THE ATLANTIC MONTHLY, Vol. 257, No. 1 (January, 1986), pp. 39–48. Liska, G. THE NEW STATECRAFT, FOREIGN AID IN AMERICAN FOREIGN POLICY (University of Chicago, 1960). Lowenthal, A.F. “Ronald Reagan and Latin America: Coping With Hegemony in Decline” in Oye, K.A., Lieber, R.J., and Rothchild, D. EAGLE DEFIANT: US FOREIGN POLICY IN THE 1980s (Little, Brown, Boston, 1983), pp. 311–335. Lyden, F.J. and Miller, E.G. (eds.) PUBLIC BUDGETING, PROGRAM PLANNING AND IMPLEMENTATION, 4th edn (Prentice-Hall, Englewood Cliffs, N.J., 1982).

Bibliography 183 Maroni, A.C., Nowels, L.Q., and Woldman, J.M. “The Gramm-Rudman-Hollings Deficit Reduction Process (PL 99–177): Implications For Programs Under The Jurisdiction of the Senate Foreign Relations Committee” (Congressional Research Service, Washington, D.C., 1986) Maynes, C.W. “If The Poor Countries Go Under, We’ll Sink With Them” THE WASHINGTON POST (September 18, 1983), p. C1. McCarthy, C. “Facing The Facts Of Contra Fund Abuse” THE WASHINGTON POST (July 6, 1986). McDonald, R.H. “Confrontation and Transition in Uruguay” CURRENT HISTORY (February, 1985), pp. 57–88. McManus, D. “Latin Turmoil Frays Liberals’ Solidarity, Handful of Activists Throwing Support Behind US Aid To Contras” THE WASHINGTON POST (January 5, 1986), p. A14. McNeil, D. THE CONTRADICTIONS OF FOREIGN AID (Croom Helm, London, 1981). McPherson, M.P. “The Success of Land Reform” THE WASHINGTON POST (March 20, 1983), p. C8. _____ “Development Issues, US Actions Affecting Developing Countries” (Development Coordination Committee, Washington, D.C., 1985). Meier, G. EMERGING FROM POVERTY, THE ECONOMICS THAT REALLY MATTERS (Oxford, New York, 1984). Merton, R.K. “Bureaucratic Structure and Personality” in Merton, R.K., Hockey, B. and Selvin, H.C. (eds) READER IN BUREAUCRACY (Free Press, New York, 1952). Meyer, L. “Israel As A Public Works Project” THE WASHINGTON POST (December 15, 1985), p. D1. Mickelwait, D.R., Sweet, C.F., andMorss, E.R. NEW DIRECTIONS IN DEVELOPMENT: A STUDY OF USAID (Westview, Boulder, 1979). Mikesell, R.F. PUBLIC INTERNATIONAL LENDING FOR DEVELOPMENT (Random House, New York, 1966). Molotsky, I. “Relief Official Accuses Ethiopia Of Biting The Hand’ That Feeds It” THE NEW YORK TIMES (December 13, 1984), p. 1. Montgomery, J.D. THE POLITICS OF FOREIGN AID: AMERICAN EXPERIENCE IN SOUTHEAST ASIA (Praeger, New York, 1962). _____ FOREIGN AID IN INTERNATIONAL POLITICS (Prentice-Hall, Englewood Cliffs, NewJersey, 1967). _____“Administering To The Poor (Or, If We Can’t Help Rich Dictators What Can We Do For The Poor?) “PUBLIC ADMINISTRATION REVIEW, Vol. 40, No. 5 (September/October, 1980), pp. 421–425. _____ AFTERMATH: TARNISHED OUTCOMES OF AMERICAN FOREIGN POLICY (Auburn, Dover, MA, 1986). Morawetz, D. TWENTY-FIVE YEARS OF ECONOMIC DEVELOPMENT, 1950 TO 1975 (World Bank, Washington, D.C., 1977). Morley, L. and Morley, F. THE PATCHWORK HISTORY OF FOREIGN AID (American Enterprise Institute, Washington, D.C., 1961). Morss, E.R. and Morss, V.A. US FOREIGN AID, AN ASSESSMENT OF NEWAND TRADITIONAL DEVELOPMENT STRATEGIES (Westview, Boulder, CO31982). Nathan, J.A. and Oliver, J.K. FOREIGN POLICY-MAKING AND THE AMERICAN POLITICAL SYSTEM (Little, Brown, Boston, 1983). Nelson, L.E. “Send Accountants To El Salvador” THE MIAMI HERALD (November 8, 1983), p. 19A. Nesmith, J. “Mexico, Central America Robbing The Future” THE ATLANTA CONSTITUTION (December 13, 1985), p. 1A.

184

Bibliography

Newfarmer, R.S. “The Private Sector and Development” in Lewis, J.P. and Kallab, V. US FOREIGN POLICY AND THE THIRD WORLD AGENDA 1983 (Praeger, New York, 1983), pp. 117–139. New York Times, “The Center Grows In Salvador” (April 3, 1985), p. 24. ______“Seoul’s Model: Tokyo or Manila?” (February 26, 1986), p. 26. Nowels, L.Q. “US Foreign Assistance In An Era Of Declining Resources: Issues For Congress in 1986” (Congressional Research Service, Washington, D. C., 1986). Odell, J.S. “Correlates of US Military Assistance and Military Intervention” in Rosen, S.J. and Kurth, J.R. (eds) TESTING THEORIES OF ECONOMIC IMPERIALISM (Heath, Lexington, MA, 1974). O’Donnell, G. “Commentary on May” in Cotler, J. and Fagen, R.R. LATIN AMERICA AND THE US, THE CHANGING POLITICAL REALITIES (Stanford, 1974), pp. 164–176. O’Leary, M.K. THE POLITICS OF AMERICAN FOREIGN AID (Atherton, New York, 1967). Omang, J. “Administration Mobilizes To Direct Aid To Rebels, US To Advise, Equip, Critique Contras” THE WASHINGTON POST, July 13, 1986, Al. Oppenheimer, A. “Andean Countries Loosen Barriers To Investment” THE MIAMI HERALD (February 18, 1985) 1B. Ottaway, D.B. “US Helps Ethiopian Rebel Area” THE WASHINGTON POST (April 14, 1985), p. A1. ______and Tyler, P.E. “US Sends New Arms To Rebels, Afghans, Angolans Get Stinger Missiles In Change Of Policy” THE WASHINGTON POST (March 30, 1986), Al. Packenham, R.A. LIBERALAMERICAAND THE THIRD WORLD, POLITICALDEVELOPMENT IDEAS IN FOREIGN AID AND SOCIAL SCIENCE (Princeton, 1973). Paddock, W. and Paddock, E. WE DON’T KNOW HOW, AN INDEPENDENT AUDIT OF WHAT THEY CALL SUCCESS IN FOREIGN ASSISTANCE (Iowa State, Ames, 1973). Popkin, S.L. “Pacification Won’t Fly In Salvador, You Have To Protect Lives To Win Hearts And Minds” THE WASHINGTON POST (April 17, 1983), p. D5. Ranis, G. “Foreign Aid: Euthanasia or Reform?” WORLDVIEW (March, 1980), pp. 19–20. Richburg, K.B. “Our Haiti Meddling Worked, America Helped Topple A Government And Everybody Cheered” THE WASHINGTON POST (February 23, 1986), p. F5. Roberts, S.V. “Senate Condemns Philippine Result As Rife With Fraud” THE NEW YORK TIMES (February 20, 1986), p. A1. Rowe, J.L. “Mexico Again On The Brink In Debt Crisis” THE WASHINGTON POST (June 8, 1986), p. D1. Rowen, H. “Philippine Debt Stunned Lenders” THE WASHINGTON POST (January 29, 1984), p. G1. ——“Debt Doubts” THE WASHINGTON POST (December 15, 1985), p. G1. ——“Can He Turn The World Bank Around?” THE WASHINGTON POST (June 12, 1986), p. A21. Rudolph, L.I. and Rudolph, S.H. “The US, India, and South Asia” in Lewis, J.P. and Kallab, V. US FOREIGN POLICY AND THE THIRD WORLD AGENDA 1983 (Praeger, New York, 1983), pp. 86–117. Rumpf, R. “The Real Taiwan Problem” THE WASHINGTON POST (April 29, 1984), p. D6. Rupert, J. “Auditor Says Marcos Diverted Interest From AID Account” THE WASHINGTON POST (March 20, 1986), p. A1. Russell, C.S. and Nicholson, N.K. PUBLIC CHOICE AND RURAL DEVELOPMENT (Resources for the Future, Washington, D.C., 1981). Schoultz, L. “US Foreign Policy and Human Rights Violations in LatinAmerica,AComparativeAnalysis of Foreign Aid Distributions” COMPARATIVE POLITICS (January, 1981), pp. 149–170.

Bibliography 185 Schultz, G.P. “Foreign Aid and US Policy Objectives” Current Policy No. 58 (Department of State, Washington, D.C., February 9, 1984). Seitz, J.L. “The Failure of US Technical Assistance In Public Administration: The Iranian Case” PUBLIC ADMINISTRATION REVIEW, Vol. 40, No. 5 (1980), pp. 407–413. Shaw, H.J. “Buttering Up Client States With Guns” THE WALL STREET JOURNAL (May 29, 1985), p. 17. Shepherd, J. “When Foreign Aid Fails” THE ATLANTIC MONTHLY, Vol. 255, No. 4, (April, 1985), pp. 41–46. Shihata, I. and Sherbiny, N. “A Review of OPEC Aid Efforts” FINANCE AND DEVELOPMENT, Vol. 23, No. 1 (March, 1986), pp. 17–21. Simon, P. “We Need A Foreign Aid Formula” THE WASHINGTON POST (June 14, 1975), p. A12. Simons, M. “In Haiti, The Land Is Worn Nearly To The Bone” THE NEW YORK TIMES (June 15, 1986), p. E3. Sinclair, W. “The World Doesn’t Need Our Farmers, More and More The Starving Billions Are Feeding Themselves” THE WASHINGTON POST (December 29, 1985), p. B1. Solarz, S.J. “Press For Philippine Reforms” THE NEW YORK TIMES (August 8, 1984), p. 25. Spanier, J. AMERICAN FOREIGN POLICY SINCE WORLD WAR 11, 9th edn (Holt, Rinehart and Winston, 1983). Stallings, B. “Peru and the US Banks: Privitization of Financial Relations” in Fagen, R. (ed) CAPITALISM AND THE STATE IN US-LATIN AMERICAN RELATIONS (Stanford, 1979), pp. 217–254, Streeten, P. “A Problem To Every Solution, Development Economics Has Not Failed” FINANCE AND DEVELOPMENT, Vol. 22, No. 2 (June, 1985), pp. 14–17. Sylvan, D.A. “Consequences Of Sharp Military Assistance Increases For International Conflict and Cooperation” JOURNAL OF CONFLICT RESOLUTION, Vol. 20, No. 4 (December, 1976), pp. 609–636. Tait, A.A. and Heller, P.S. “International Comparisons of Government Expenditures”, Occasional Paper No. 10 (Washington, D.C., International Monetary Fund, 1982). Tendler, J. INSIDE FOREIGN AID (Johns Hopkins, Baltimore, 1975). Thompson, V.A. MODERN ORGANIZATION, 2nd edn (Alabama, 1977). Thompson, W.S. (ed.) THE THIRD WORLD, PREMISES OF US POLICY (Institute for Contemporary Studies, San Francisco, 1983). Todaro, M.P. ECONOMIC DEVELOPMENT IN THE THIRD WORLD, AN INTRODUCTION TO PROBLEMS AND POLICIES IN A GLOBAL PERSPECTIVE (Longman, New York, 1977). Tonelson, A. “What Central American Democracy?” THE NEW YORK TIMES (May 22, 1985), p. 29. Trebat, T.J. BRAZIL’S STATE-OWNED ENTERPRISES: A CASE STUDY (Cambridge, 1983). Tyler, P.E. and Ottaway, D.B. “The Selling of Jonas Savimbi, Public Relations Firm Paved Guerrilla’s Way” THE WASHINGTON POST (February 9, 1986), p. A8. United States Agency For International Development, “Design and Evaluation of AID-Assisted Projects” (USAID, Washington, D.C., 1980). ______“Congressional Presentation, Fiscal Year 1982, Main Volume, Annex II, Asia, Annex IV, Near East” (USAID, Washington, D.C., 1981). ______“US Overseas Loans and Grants and Assistance From International Organizations, Obligations and Loan Authorizations July l, 1945-September 30, 1984” (USAID, Washington, D.C., 1984) ______ “Congressional Presentation, Fiscal Year 1986, Annex 11I, Latin America and the Caribbean, Volumes I and II (USAID, Washington, D. C., 1985). ______“Congressional Presentation, Fiscal Year 1987, Main Volume, Annex 11, Asia and the Near East, Annex 11I, Latin America and the Caribbean” (USAID, Washington, D.C., 1986). ______“Fiscal Year 1987 Summary Tables” (USAID, Washington, D.C., 1986).

186

Bibliography

United States Department of Defense, “Congressional Presentation For Security Assistance Programs, Fiscal Year 1987, Volumes I and II” (Defense Security Assistance Agency, Washington, D.C., 1986). United States General Accounting Office, “Review of Inspector General Functions In Agency For International Development” (Washington, D.C., 1982), GAO/ID-82–9. ______, “Assistance To Haiti: Barriers, Recent Program Changes and Future Options” (Washington, D.C., 1982), GAO/ID-82–13. ______, ”Donor Approaches To Development Assistance: Implications For The US” (Washington, D.C., 1983), GAO/ID-83–23. ______,“AID Recognizes Need To Improve The Foreign Assistance Planning and Programming Process” (Washington, D.C., 1985), GAO/NSIAD-85–110. United States House of Representatives, “Congress and Foreign Policy-1984” (Committee on Foreign Affairs, Washington, D.C., 1985). United States Office of Management and Budget, “Budget of the US Government, Fiscal Year 1986” (OMB, Washington, D.C., 1985). United States Senate, “Foreign Assistance and Related Programs, Appropriation Bill, 1985” (Committee on Appropriations, Washington, D.C., 1984). Warwick, D.P. A THEORY OF PUBLIC BUREAUCRACY, POLITICS, PERSONALITY AND ORGANIZATION IN THE STATE DEPARTMENT (Harvard, Cambridge, 1975). White, R.E. “Cut Loose from Chun” THE NEW YORK TIMES (March 5, 1985), p. 21. Wiesner, E. “Domestic and External Causes of the Latin American Debt Crisis, The Crucial Impact of Inadequate Domestic Policies” FINANCE AND DEVELOPMENT, Vol. 22, No. 1 (March, 1985), pp. 24–27. Wildaysky, A. THE POLITICS OF THE BUDGETARY PROCESS, 4th edn (Little, Brown, Boston, 1984). Wilford, W.T. “The Congressional Mandate On U.S. Foreign Aid: Can It Break The Vicious Circle Of Poverty?” JOURNAL OF SOCIAL AND POLITICAL STUDIES, Vol. 15, No. 4 (Fall, 1979), pp. 261–268. Witcher, S.K. and Walsh, M.W. “With Mexico Focusing On Debt Repayments, Ports and Roads Suffer” THE WALL STREET JOURNAL (June 11, 1986), p. 1. Wolf, C. FOREIGN AID: THEORY AND PRACTICE IN SOUTHEAST ASIA (Princeton, 1960). Wolfson, D.J. PUBLIC FINANCE AND DEVELOPMENT STRATEGY (Johns Hopkins, Baltimore, 1979). Woll, P.CONGRESS (Little, Brown, Boston, 1985). Wood, R.E. “Foreign Aid And The Capitalist State In Underdeveloped Countries” POLITICS AND SOCIETY, Vol. 10, No. l (1980), pp. 1–34. Wynia, G.W. POLITICS AND PLANNERS, ECONOMIC DEVELOPMENT POLICY IN CENTRAL AMERICA (Wisconsin, 1972). ______ THE POLITICS OF LATIN AMERICAN DEVELOPMENT (Cambridge, 1984).

INDEX

actor conflict 260 aid meaning of 264 Alliance for Progress 37, 38, 40–41 “country programming” 41, 42–43 project rationale 42 Aquino, Benigno 218, 219 Asia 179–221 aid conditions 215 aid statistics 179–80 budgetary systems 210 centralized power 193, 195 conflict resolution 215–16 constitutional democracy 184 constraints 214–21 decentralized economic power 194 democratic reform leverage 187 dependency 198 analyses 199–200 distribution of power 193 economic aid successes 196, 203–4 economic dependency 201–2 effects of aid 196–7 excessive optimism 186 foreign policy goals 179 goals 182–95 high staff to project ratios 197 military dependency 201–202 mutuality of interests with aid recipients 192 overconfidence 186 political development doctrines 183 programming practices 206 projects to increase farmer productivity 197–198 radicalism 190 recipient and donor conflicts 214 recipient growth aid profitability, and 189 results 195–204 revolution 190 role conflict 212–220 security as guiding premise 191–2 Soviet pressure 182–3

stability-democracy presumption 186 transfer of excess stocks 211 US domestic political concerns 214–15 unintended effects 189–203 Aswan High Dam 236, 241–2, 250, 251 support for proposal 251–2 authoritarian regimes 261 Bangladesh 70 base rights commitments 228 bilateral foreign aid allocation by major region 181 Brazil 141 British Guyana 40 budget authority requests statistics 1975–1986 114 budget cycle 81–2 budgetary expense 2–3 Bundy, McGeorge 242 bureaucratic problems informal subsystems 80 bureaucratic role conflict model 79–132 aid results 87, 259–60 adversary relationships 111 alternative models 94–5 annual reauthorization 103 budget format 105 budget process as pressuring device 104–5 budget request 98 budgeting 102 cash flow emphasis 106 civil servants 127 complexity 90 conclusions 260–1 conflict between guardian and spender roles 109–11, figure 112 Congress see Congress constraints on achievements of goals 99 constraints 95–108 contractors see contractors

188

Index

creativity, discouragement of 101 culture of bureaucracy 95–6 decision premises 90–5 dependence on economic aid 87–8 dependency 90 development 98–9 deviations from past plans 100 domestic US interest groups 130 evaluation routines 107–8 execution phase 105–6 existing rules 85 feedback routines 107–8 figure 86 financial benefits to exceed costs 93 focus of budget submissions 103–4 foreign aid budget process 85 foreign aid policy analysis 111–13 goals 90–5 “guardians” 113–25 host-donor relations 126–7 inattention to implementation 89–90 incentive structure 108 independent variables 84–5 institutional control emphasis 107 military aid 93, 94 opportunities 109–132 output measures 87 overly complex projects 88 pattern of advocacy 131 planning 96–7 policy process 110 programming 96–7 routines 100–1 proportion of economic-military assistance 128 rational choice 83 realistic conflict 119 recommendations 260–1 recipient line ministries 131–2 recurrent cost items 101 repertoires 95–108 resource allocation 102 reverse exploitation of foreign aid program 106–7 review of projects 98–9 role conflict 109–32 rules 95–108 security-related funding 127–8 self-criticism 108 “spenders” 127–32

translation of goals into policy 109 unintended consequences 88–9, 92 weapons, sale of 101–2 Byroade, Ambassador 250–2 Cambodia 211–12 Carroll, Rear Admiral Eugene, Jr. 154 Carter, President 143, 168, 169 Chile 92–3 China 183–4 aid from North Vietnam 202 Colombia 70 productivity growth 140–1 Congo 238 Congress 115–25 aid preferences 115 budgeting of foreign aid political pressures 208–9 cash flow analysis 116–17 change of role 125 continuing resolution 124 defense orientation 116 distrust among foreign aid advocates 120–2 distrust of executive foreign aid policy-making 119–20 earmarking of funds for specific uses 118 earmarks for Economic Support Fund 1986 121 improvements of relations with executive aid actors 124 making foreign aid policy 265–6 overfunding 118–19 policy making 117 politically unpopular issues 117 predominant positions taken by 115–16 rules for foreign aid 208 security assistance appropriations 176–7 superficial treatment of foreign aid 176 view of foreign aid 175–6, 207 Consultative Group on International Agricultural Research (CGIAR) 61–2 contemporary foreign aid programs 14–26 contractors 129–30 profit-seeking 129–30 control of organizations 80–1 Costa Rica 60 level of aid 170 critical perspectives 52–78 avoidence of critical information 58 “bureaucratics” 60

Index conservative 73–7 definition of foreign aid 74–5 drain on US balance of payments 76–7 “inappropriate regime” thesis 76 policies tending to impede development 75 reduced governmental presence 73–4 donor jargon 58 evaluations from political spectrum 53–78 “know-nothing” 57–8, 61–2 land reform 57 “leave them alone” 55–6 liberal critique 68–73 agribusiness 69–70 confusion about means and ends 71 contradictions 71–3 cultural practices 72–3 discouragement of local production 70 “interventionist” programs 73 large farmers 69–70 low-income countries 69 mutuality of interest 68–9 policy fads 72 “trickle-down” theories 71 weaknesses 71–3 radical left 62–8 market-price orientation 65–6 neo-classical economic theory 65 real development policymaking, inhibition of 67–8 state capacity 67 summary 77 Cuba 60, 61 Cyprus 234 D’Aubisson, Roberto 156 de la Torre, Haya 170 Development Loan Fund (DLF) 38 Development Assistance (DA) 14, 15–16 technical assistance transfers 16 developmental objectives 5–6 Diem, Ngo Dinh 185, 200–1 domestic context 5 domestic programs 9–10 Duarte, President 156 Dulles, John Foster 234, 250 Duvalier, Baby Doc 69, 173, 174

189

Economic and Military Assitance Programs request for fiscal year 1986 149 economic and trade policies 5 Economic Support Fund (ESF) 13, 14, 16–19 balance of payments support, rationale 18–19 Egypt 18 military objectives 19 security objectives 19 Ecuador 59–60, 138 Ecuadorian Forest Service 62 Eden, Anthony 250 Egypt 13, 235 bureaucracy 236–7 ESF program 18 food supply 238 negotiations with Israel 237 permanently dependent client state 227 resentment generated by aid conditions 247 stabilization of economy 236 Eisenhower, President 33–4, 15–23 El Salvador 11, 141–2, 143, 151, 156–7, 162, 168, 207, 263 Ethiopia marxist government 10 European Recovery Program 32–3 evolution of US foreign aid 26–51 cost considerations 27 five overlapping stages 26 historical policy, and 51 issue determinants 26–7 liberal political ideas, export of 37–8 local currency requirements 35 low-interest loans 36–7 “New Directions” mandate see also New Directions 45–50 “punitive peace” policies 31 self-help 39 self-interest 44–5 social change thrust 39–40 social justice considerations 30–1 statistics 28–9 Excess Defense Articles (EDA) program 70 expenses 146 federal system 194 Figueres, Jose 190 Fishel, Wesley 154–5

190

Index

five programs 1–2, 14–26 Food Aid 14, 21–2 P.L. 480 program 21–2 Food For Peace 70 Food For Work (FFW) 9, 10 foreign aid blurred lines of policy 3–4 confusion over objectives and evaluation criteria 2 criticism of 3 evaluation dilemma 2 military dominance 12 objective assessment of results 257 past and present 1–51 political economy of 1–51 presidential request for fiscal year 1986 budget authority 17 unpopularity of 212–13 foreign aid appropriations 7–8 foreign aid foreign policy linkage 6 foreign exchange cost financing 145–6 Foreign Military Sales (FMS) 19 statistics 1946–1981 20 foreign policy foreign aid as instrument of 7 process 201–2 France 27 Fuentes, Ydigoras 155 Government Relief in Occupied and Liberated Areas (GARIOA) 30 Gramm-Rudman-Hollings (G-RH) 122–3 Middle East 246, 248 outlay savings 265 Greece 32 Gruening Amendment 247–8 Guatemala 55–6, 60, 155 land and tax reforms 138–9 Haiti 162, 173–5, 261 clear goals 174–5 development projects 174 guardian and spender roles 173–4 Harding, President 27 Hoover, President 27 Humberto, General Carlos 143 Hwan, Chun Doo 187 implementation field problems 6–7

increased emphasis on 262–3 improved forum concept of 270–1 India 230 independent sources of government power 25–6 International Cooperation Administration (ICA) 38 International Paper Company 68 Iran 11–12, 13, 61, 70, 253–4 Israel 13, 224 negotiations with Egypt 237 permanently dependent client state 227 US commitment to 227–8 Jagan, Cheddi 40 Johnson President 40, 43 Kennedy, President 37, 38, 39, 40–1, 43, 143, 151 Kennedy, Senator Ted 73 Kissinger, Henry 237 Kissinger Report 1984 154 Kuwait 239 Latin America 43, 66–7, 133–178 agencies programming aid 161–2 AID budget process 162 aid results 167 Alliance for Progress 142–3 amount of funding 133–4 annual reauthorization by Congress 163 appropriation of funds by Congress 163 assistance over one million dollars 159–60 Congress, role of 175–6 see also Congress constraints 158–66 contradictory sociopolitical context 136–7 counterpart dependency 144 criteria for evaluations 165–6 Department of Defense (DOD) 170–3 aid, size of 171–2 Armed Service Committee relationship 177 defense advocacy 172 revolutionary governments 172–3 “dependency” 137–8 determinants of foreign aid results 164 economic aid 135 enrichment of military leaders 140 evaluation phase 164–5 evaluation of aid programs 169–70

Index financial evaluations 165–6 foreign policy 134–5 goals 147–58 democratically authoritarian structures 155–6 Department of State 168 developmental need 147–8 hemispheric hegemony 152 incentives to left and right 156–7 military and economic programs 157 military, support of 151 mixed guardian-spender roles 168 negotiations among warring factions 156–7 nondevelopmental 147 “outside agitator” premise 150 profitable aid, fiction of 157–8 rhetoric of developmental activities 148–9 running security premise 154 security assistance 148, 153 security-profitability 151 “trickle-down approach” 150–1 welfare need 147–8 harmful features of US economic aid 138 items which must be programmed into project 160 local institutional dependency on foreign aid process 138 military assistance 141–2 political cultures 136 population 156 potentially revoluntary threats 168–9 power differences in policy-making 167 presumption of results in one fiscal year 163–4 project development programming phase 158–9 reinforced dependency 145–6 results 135–47 role conflict 166–77 security crises 134 statutory “checklists” 160–1 straight financial issues, emphasis on 162–3 “subsidies” 141 unintended consequences 145–6 US military assistance 139–40 use of aid as leverage 144–5 Lebanon 230–2 amount of aid 240

191

national income 232–3 sectarian factions 232 Lend-Lease program 30 level of results 4–5 Lewis, Verne 256–7 Lon Nol 211–12 McNamara, Robert 43, 45 Mali Livestock project 55 Mann Doctrine 40 Marcos, President Ferdinand 13, 69, 187–9, 220, 266 business opposition 194–5 procedural irregularities 188–9 rejection of 218–19 Marshall Plan 32, 33, 36, 42, 75 Middle East 222–55 aid successes 226 Arab-Zionist conflict 243 base rights commitments 228 breakdown of interagency processes 252 conflict 248–9 Congressional conditions 245 consensus problem 250 constraints 245–8 criticism of aid policy 245 delegitirnization of regimes 229–30 dependency 226–7, 228–9 East-West issues 234 financial profit 242 goals 239–44 Gramm-Rudman-Hollings, legislation 246 “guardian” role of US 249–50 Islamic fundamentalism 234–5 legitimate political authority 223 local political complexities 254 militarization of foreign aid 252 mutual agreement and trust 255 need for aid 224 policy actors 248–54 political aspects of aid 244 “profitable” development 241 results 225–39 security 243–4 US aid efficacy 222 US global-regional political interests 240 volatility 229 war 242–3 military alliances 257–8 Mill, J.S. 73

192

Index

Monroe Doctrine 31 Morgenthau Plan 31 Morocco amount of aid 240–1 multilateral development banks (MDBs) 22–3 Mutual Security Act 1951 35 Mutual Security Act 1954 246–7 Nasser, Gamal 231, 235–6, 238–9, 246 negative controversy examples 9 New Directions 45–50 “appropriate technology” 48 “bottlenecks first” approach 47 concepts 50 distributional objective 45–6 “expert flip-flop” as 49 meaning 46–7 obstacles to success 48–9 New Directions legislation 1973 206–7 Nicaragua 13, 213 National Guard 142 political vacuum 263 Nixon, President 37, 44, 152 North Atlantic Treaty Organization (NATO) 33 North Korea 184–5 North Vietnam aid from China 202 Obey, David 251 OPEC 239 Panama 60 Paraguay 70 Passman, Otto 251 permanently dependent client states 12 Peru 41 security assistance to 153 Peterson Report 44–5 Petkoff, Teodoro 68 Philippines 13, 61, 196 aims of aid 217–18 continuity of purpose 221 decline in institutions 192–3 distrust of institutions 219–20 “genuine cooperation” 221 Pinochet, Augusto 92–3 Poland 27 policy determination of 268–9 policy models

need for 79–90 Political Action Committee (PAC) 12 political development 24 power imbalance 268 power-seeking 8 private development strategies 25 private sector Reagan approach 24 Private Voluntary Organizations (PVOs) 70 program budgeting 256–7 Reagan, President 11, 50–1, 146, 187, 217, 220 recipient dependency 258–9 Rockefeller Report 152 recommendations 268–72 resource conflict model weakness of 82 role clarification 267–8, 271–2 role definition 267–8, 270 Russia 27 Sadat, President Anwar assassination 238 SAVAK 61 Sandinistas 156 Scott Paper Company 60 Secretary’s Assistant Policy Working Group (SAPRWG) 122 Security goal of foreign aid program 263 Security Assistance 14 security dimension 77–8 Somoza 69, 169, 170 South EastAsia Treaty Organization (SEATO) 185 South Korea 183, 184–5, 187 State Department funding ceilings, setting of 257 statistics aid distribution 15 Stinger missiles 111 Stroessner, General 70 Sudan “Freedom from Thirst” Projects 4–5 Suez invasion 34 Suharto, President 69, 191 Sukarno 190–1 Syria 230 Taiwan 186–7, 189–90, 196, 267 Tanzania 49–50 Terry, President Belaunde 155 Trong, Le Hoang 199, 200

Index Truman, President 31, 35, 184 Turkey 233–4 military aid 233–4 United Fruit Company 60 United Nations Relief and Rehabilitation Administration (UNRRA) 30 USAID 15, 38 evaluations 53–4 field missions 59 guardian and spender 125–6 increased influence, need for 269–70 major foreign aid advocate 128 personnel, use of 61 planning and budgeting process fiscal year 1986 97 “political development” objectives 40 Uruguay 24–5 Venezuela 68 Accion Democratica (AD) Party 155

193

earthquake 30 Vietnam 26, 43, 45, 73, 185 absorptive capacity 211 Civic Action Program 216 conditions on use of American aid 207–8 consumption of aid 199 dependency 198–9, 200–1 escalation 205–6 foreign policy institutions 205 Land Development Program 210, 216–7 military grants 206 Tilapia fish project 198 unintended consequences 200–1 US aid-financed imports 199 War on Poverty 73 Yugoslavia 32 Zia Ul-Haq 69 Zimbabwe 41

THE ECONOMICS OF AID J.M. Healey

ROUTLEDGE LIBRARY EDITIONS: DEVELOPMENT

ROUTLEDGE LIBRARY EDITIONS: DEVELOPMENT

THE ECONOMICS OF AID

THE ECONOMICS OF AID

J.M.HEALEY

Volume 8

I~ ~~o~f~;n~~~up

LONDON AND NEW YORK

First published in 1971 This edition first published in 2011 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OXl4 4RN Simultaneously published in the USA and Canada by Routledge 270 MadisonAvenue, New York, NY 10016 This edition published in the Taylor & Francis e-Library, 2011. To purchase your own copy of this or any of Taylor & Francis or Routledge's collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.

Routledge is an imprint ofthe Taylor & Francis Group, an informa business © 1971 J.M.Healey

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers.

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 0-203-84017-8 Master e-book ISBN DOI: 10.4324/9780203840177 ISBN 13:978-0-415-58414-2 (Set) eISBN 13:978-0-203-84035-l (Set) ISBN 13:978-0-415-59267-3 (Volume 8) eISBN 13:978-0-203-84017-7 (Volume 8)

Publisher's Note The publisher has gone to great lengths to ensure the quality of this reprint but points out that some imperfections in the original copies may be apparent. Disclaimer The publisher has made every effort to trace copyright holders and welcomes correspondence from those they have been unable to contact.

The Economics of Aid byJ.MHealey

LONDON

ROUTLEDGE & KEGAN PAUL

First published 1971 by Routledge and Kegan Paul Ltd. Broadway House, 68-74 Carter Lane, London, EC4V 5EL This edition published in the Taylor & Francis e-Library, 2011. To purchase your own copy ofthis or any of Taylor & Francis or Routledge collection ofthousands ofeBooks please go to www.eBookstore.tandfco.uk.

s

© J.MHealey, 1971

No part ofthis book may be reproduced in any form without permission from the publisher, exceptfor the quotation of briefpassages in criticism ISBN 0-203-84017-8 Master e-book ISBN

ISBN0 7100 6954 5 (c) ISBN0 7100 6955 3 (p)

General editor's introduction

The plight ofthe underdeveloped or emerging countries is one ofthe most pressing problems in the world today. It is often difficult for the ordinary inhabitant of an advanced country or for the student, for that matter, to appreciate the great gap between his position and that of his counter-part elsewhere in the world. The size of the gap and whether it is widening or not will be a major issue for international discussion in the years ahead. It is very closely connected with the question of economic aid which is the subject of this book. Curiously enough, although there is a great deal of talk about economic aid, about its value, of what sort it ought to be, whether it is wasted, and so on, there is no good elementary treatment of it available. This is especially curious when it is appreciated that the subject is ameoable to elementary economic analysis. The sort of techniques which the student learns are most useful in clarifying the problem of aid. At the same time, a study of these problems indicates the value of the theoretical principles and techniques that have been acquired in other courses. Thus, such matters as technical progress, the consequences of an increase in a particular factor of production, and the balance of payments are all placed into a focus which they otherwise might not get. Dr Healey is eminently well qualified to write a book of this kind. He is an expert on both the theoretical and applied side of this and related fields. He has direct experience of underdeveloped countries and has seen for himself the kind of problems they face. The book was written while Dr Healey was a lecturer at Queen Mary College, and does not, of course, necessarily reflect the views of the Overseas Development Administration of the Foreign and Commonwealth Office, where he is currently employed. M.H.P.

Contents

Page

General editor's introduction Tables Figures Acknowledgments Introduction 1 Aid: motives

vi viii ix X

xi 1

2 Aid: allocation principles

11

3 Aid and economic growth I

21

4 Aid and economic growth II

34

5 Debt and the terms of aid

43

6 Aid tying: trade and resource allocation effects

49

7 Some questions and further issues

59

Appendix: Resource gap models: statistical estimation of the parameters

64

Further reading

68

Tables

page

1

Official and private flow of financial resources from D.A.C. and Communist Bloc countries to less developed countries and multilateral institutions ($m) 2 Per capita G.N.P. and receipts of net official assistance by selected developing countries from O.E.C.D./D.A.C. countries and multilateral agencies, 1964-6 annual average 3 Relationship between aid receipts per capita, 1964-6 and gross domestic product per head, 1965 4 Grant element in loans with different interest rates and maturities 5 Grant element and gross official disbursements to developing countries by D.A.C. countries in 1965 6 Performance of developing countries 1957-62 7 Estimates of aggregate savings and trade gaps of developing countries 8 Recipient's imports with untied aid 9 Recipient's imports-A's aid tied to marginal project 10 Recipient's imports-with tied non-project aid

4

5 8 12

13 28 37 50 51 52

Figures

page 1 Effect of capital inflow on growth rate

2 3 4 5 6 7

Effect of capital inflow on growth rate Ex-ante 'saving' and 'trade' gaps and growth rates

Demand and supply of exports Hypothetical debt profile Aid tying Aid tying

24 25 36 39 45 53 56

Acknowledgments

This book lays little claim to originality and owes much to the writing of many economists in recent years. The analysis in Chapter 3 is based largely on the approach pioneered by Professor Chenery and his associates and draws particularly on R.McKinnon's exposition in the Economic Journal, 1964. The analysis of aid tying in Chapter 6 draws on an unpublished paper by Mr J.P.Hayes and a published report by J.Bhagwati prepared for UNCTAD. I have benefited from discussions with colleague economists in the Ministry of Overseas Development and at Queen Mary College, London. I am particularly grateful to Maurice Peston, Michael Lipton, Alf Vanags and Nick Baigent for reading the script and making critical comments and suggestions. None of these is responsible for the errors and inadequacies remaining. The opinions expressed in the book are entirely personal and do not in any way reflect the views of my employer, the Overseas Development Administration of the Foreign and Commonwealth Office.

Introduction

Many interesting books have been written on aid but most of these have been concerned with aid as an aspect of foreign policy, giving economic aspects a secondary place. Moreover there has been a tendency for writing on these subjects to be directed towards the needs of the politicians and administrators who are faced with day-to-day judgments and decisions which constitute aid policy. For this reason the literature on aid is generally rather pragmatic in its approach. For the undergraduate and post-graduate economist a more analytical approach to aid and aid policy is desirable. This means defining aid, examining the role of aid and its allocation and assessing the effects of particular forms of aid with the concepts and techniques of economic theory, which are familiar to the student. This short study examines some of the main aspects of foreign assistance to developing countries in terms of economic principles. The book does not pretend that foreign assistance policy is, or can be, decided on purely economic criteria, but it is based on the belief that policy issues in this field are greatly clarified if they are examined from the economist's viewpoint. It is hoped that this book will be helpful to students and teachers of undergraduate and post-graduate courses in development economics and international economics. The level of the analysis at no point should present any difficulty for a third-year economics undergraduate.

1 Aid: motives

What is aid? Aid is an ambiguous word and there is no conunon agreement on its definition and hence its measurement. Different organizations and different countries include or exclude a variety of items. Thus, the Development Assistance Committee of O.E.C.D. records the 'flow of longterm financial resources to less developed countries and multilateral agencies' and this includes both official flows and private investment, private lending and export credits. Some bodies do not record private capital flows. But it is not merely a question of whether aid should include flows of private resources, for official flows are very heterogeneous in character; including outright grants, loans on different terms of interest and repayment, funds in convertible currency and funds which are tied to spending in one country and which may also be tied to specific uses. Simple aggregation is therefore almost meaningless and certainly misleading. One suggestion for reducing these heterogeneous categories to a conunon measure is to define aid as the real cost or sacrifice to the countries providing the nominal flows of resources. This would give aid an economic meaning. The real cost or 'grant element' in a loan would be measured by the difference between its nominal value and the future interest payments and amortization discounted at a rate which reflects the return on the funds alternatively open to the country supplying them. In the case of a grant the cost would obviously be identical with its nominal value. This approach to aid measurement in terms of the economic sacrifice made by aiding countries is discussed more fully in the next chapter. A satisfactory concept of aid is particularly important for assessing the magnitude of each developed country's true aid effort and the problem of sharing the burden of assistance equitably between countries. In the rest of this chapter, however, the term 'aid' or 'assistance' will be used in the loose sense and will normally refer to the flow of long-term official financial resources between developed and developing countries. This includes 'bilateral' or direct government to government finance and 'multi-lateral' flows which are channelled through the United Nations' specialized agencies and regional bodies like the Inter-American Development Bank. Although a measure of the cost of aid is essential for any form of burden sharing among the aiding countries, philanthropy is not necessarily required to assist growth in the developing countries. Capital flows on terms which involve no gift element at all can nevertheless benefit the recipient economy, and Chapter 3 is devoted to the process by which a net capital inflow can influence the growth of domestic product.

Motives Official net flows of financial resources have been transferred from rich to poor countries on a large and growing scale in the last fifteen years (see Table 1). This is very much a

DOI: 10.4324/9780203840177-l

2

The economics ofaid

post-war phenomenon and it raises the question of why it has occurred. What interests, considerations, motives have generated such a large official flow of funds and their particular geographical distribution? Governments usually state that they provide assistance to other countries to assist their economic development. (The meaning of' development' is examined below.) An interest in development per se, of course, implies a purely humanitarian motive but governments are also motivated by national self-interest. What are the self-interest motives of governments likely to be in the provision of official aid? The following pages review some of the direct commercial political and strategic interests in aid-giving and the self-interests of the aid-givers served indirectly by assisting economic development. Statistical evidence is then considered to see what light it throws on the factors that seem to influence official assistance policies. Commercial motives Capital flows in the form of loans at rates of interest which exceed the rate of return on investment in the lending country clearly benefit the lender commercially, but capital flows of this kind do not constitute 'aid' in the real sense. Where capital flows have some concessionary element in them, they may nevertheless, in principle, commercially benefit lenders as a whole and in the long run. This would be so if the development, resulting from aid, increased the scope for specialization and international trade. The significance of this effect is likely to be small given the size of already existing international markets for goods and given the alternative policy of reducing barriers to trade such as tariffs and quotas. It is unlikely that these broad commercial considerations have influenced assistance policy. On the other hand there is no doubt that individual countries have provided loans on concessionary terms as a weapon to gain a competitive advantage for their own exports in international markets. Although national export promotion seems to be a motive, in practice it only constitutes a genuine national advantage if export earnings are worth more than their nominal value and this would imply the existence of a disequilibrium official exchange rate. Political and strategic motives Some attempt is usually made to separate military assistance from total recorded official assistance figures, since the motive for this expenditure is clear. (Nevertheless military aid can in part be developmental in effect if the recipient would have expended some of his own resources on defence anyway.) The offer offinance to 'keep friends and influence people' is a long-standing international practice and the desire to preserve friendly links and historical ties by offers of assistance partly accounts for (say) the concentration ofU.K. assistance on Commonwealth countries. Aid becomes merely an instrument of foreign policy when its intent is to buy support for the policies of the 'assisting' country in (say) U.N. and other international forums. Aid which is effectively used to 'buy a base' for the air or naval operations of the aiding country is basically a payment for a service. However, the political and strategic intentions of aiding governments are neither as simple nor as limited as this. If friendship or support could be merely a financial purchase, why do they usually show concern that the finance provided be used for development in the recipient country? Since there is unlikely to be significant humanitarian concern for

Aid: motives

3

development, the aiding governments must assume that 'aid for development' will sub serve their political and strategic interests more effectively than straight subsidies or bribes. There are several ways of questionable plausibility by which the objective of development may be considered to serve the self-interest of the aid-givers. a The regimes receiving the assistance and whose friendship or support is sought are more likely to survive if their economy is developing. b The provision of aid for development may appear more respectable than straight subsidies and hence may gain more goodwill. c The political stability of an area may be considered to vary directly with its level or pace of development.

What light do aid statistics throw on the relevance or importance of these different motives? Statistical evidence of two kinds is available: trends in the volume of assistance and cross-section data on the geographical distribution of aid at a particular time. These will be examined in tum

Statistical trends Table 1 shows the trend of total assistance between 1956 and 1967. The figures cover net official flows of financial resources from D.A.C. countries 1 to less developed countries and multilateral agencies (i.e. gross disbursements minus return amortization payments). Financial flows from the major individual D.A.C. countries are shown as well as private net investment, lending and export credits. The figures for the Communist countries are estimates and are gross figures. During the 1950s the U.S.A. was the major contributor to total assistance but in the early 1960s a large jump in assistance levels is observable. A new plateau was reached in the 1960s with no significant upward trend and total assistance of the D.A.C. countries tended to fall as a proportion of their aggregate G.N.P. The sharp rise in U.S. and aggregate assistance about 1960 seems to reflect various influences. The U.S. Government publicly stated its strategic objective to be assistance for neutral and uncommitted countries to prevent them becoming dependent on Soviet aid. Although Table 1 shows no sharp rise in Communist aid disbursements at this time, their aid commitments were stepped up. Federal Germany also had cold war motives for embarking on larger assistance programmes in the 1960s as it sought sympathy for its position on Berlin and East Germany among uncommitted countries. A further major reason for the leap in assistance in the early 1960s was the emergence of many countries from colonial status who, admitted to the United Nations, pressed for more multilateral aid and, on independence, gained increased bilateral financial support from their former colonial rulers. Commercial motives also played their part. West Germany, Italy and Japan had made their economic recoveries by the late 1950s and to some extent began using loan finance as a weapon to compete in export markets.

4

The economics ofaid

Table 1 Official and private flow offinancial resources from D.A. C. and Communist Bloc countries to less developed countries and multilateral institutions ($m) D.A.C. countries*

Communist bloc Private netflow

Official netfiow ($m)

U.S.

UK. Germany Japan

1956 2,006 205

142

70

Total Grosst Private disbursements and official

Total %of official G.N.P 3,288

2,878

6,166

107

57 2,091 234

275

71

3,832

3,697

7,529

81

58 2,410 276

268

262

4,411

2,825

7,236

205

59 2,322 377

332

92

4,398

2,649

7,047

161

60 2,776 407

351

109

4,919

(0·54)

3,007

7,926

178

61 3,447 457

618

107

6,011

(0·63)

3,076

9,087

275

62 3,536 421

469

89

5,961

(0·57)

2,431

8,392

390

63 3,699 415

437

140

6,081

(0·54)

2,382

8,462

575

64 3,445 493

423

116

5,856

(0-48)

3,287

9,143

500 (?)

65 3,627 481

471

244

6,200

(0-47)

4,293

10,493

66 3,660 526

486

285

6,498

(0-46)

3,973

10,471

67 3,723 498

547

391

6,977

(0-46)

4,329

11,306

* D.A.C. countries are listed in a footnote in the text. The major source of financial flow statistics on these countries is the annual O.E.C.D. Development Assistance Review. These figures are taken from the 1967 and 1968 Reviews. The O.E.C.D. and U.N. both publish. detailed figures oflongterm capital aid to poor countries. t The figures for the Communist Bloc are estimates ofeconomic grants and credits to non-Communist less developed countries. They are gross figures unlike the net figures for D.A.C. countries. These estimates are largely derived from calculations by the Bureau oflntelligence and Research of the U.S. Department of State. (See H.J.Amold,Aidfor Development, 1966, p. 162.)

Total D.A.C. official flows have continued to rise somewhat in real as well as money terms but it is noticeable that U.S. assistance has not risen between 1963 and 1967. There is little doubt that this reflects a weakening of some of the dominant motives behind U.S. assistance programmes. Assistance has not always purchased friendship or goodwill. Assistance has not always ensured political stability and it has generally failed to achieve 'quick' results economically.

Geographical distribution The geographical distribution of assistance rather than time trends is more likely to yield some indication of the motives or interests behind it. Thus, it is possible to statistically

Aid: motives

5

correlate assistance per capita to different countries with various country characteristics, some of which may form approximate measures of the various economic and political interests of those countries providing the assistance. For example, interest in the economic development of aided countries may be represented prima facie by their per capita income since their degree of poverty may be viewed as inversely related to their need for development and the adequacy of their domestic resources for development. There is, however, some ambiguity in this approach because low income levels may indicate a low capacity for using foreign capital effectively and hence the inadvisability of supplying large capital flows. Table 2 shows net official financial assistance received and gross national product for different developing countries (1964-6) on a per capita basis. Table 3 shows the class distribution. The data in both cases refer to bilateral and multilateral assistance from D .A. C. countries. Both these tables suggest, and systematic statistical analysis confirms, that there is no significant correlation between per capita assistance and per capita income level of Table 2 Per capita G.NP and receipts of net official assistance by selected developing countries* from O.E.C.D.ID.A.C. countries and multilateral agencies, 1964-66 annual average U.S.$

Recipient area/countries

G.NP

Europe:

Assistance Bilateral

Multilateral

Total

3·8

0·8

4·7

of which: Greece

510

3·6

l ·6

5·3

Spain

530

0·9

0-4

1-4

Turkey

240

5-4

0-4

5·7

Yugoslavia

390

5·0

l ·5

6-5

5-4

0·7

6-1

Africa: of which: Algeria

230

14·0

0·2

14·3

Libya

210

5·3

-0·6

4·7

Morocco

170

7· l

0·6

7·7

Tunisia

180

16-7

l ·2

17·9

U.A.R. (Egypt)

150

4·6

0· l

4·7

50

l ·7

0·5

2·3

140

7· l

0·6

7·7

50

0·9

0-4

l ·2

Burundi Congo (Kinshasa) Ethiopia

6

The economics ofaid

Ghana

230

6-0

l ·6

7·6

Guinea

70

3·7

l ·l

4·9

Kenya

90

6-4

0-4

6-8

Liberia

180

30-4

l ·2

31 ·6

Malawi

40

8·3

l

8·3

Nigeria

100

l ·5

0·5

2·0

Rwanda

50

2-4

0·6

3·0

120

5·3

0·6

5·9

Somalia

50

7-4

l ·5

8·8

Tanzania

70

3·3

0·3

3·6

Uganda

80

2·8

0·3

3· l

Zambia

160

6-1

-0·7

5-4

§

5·6

2·2

7·8

3-4

l ·0

4-4

Sierra Leone

African & Malagasy States America: of which: Costa Rica

360

8·0

5·9

13-9

Dominican Republic

210

14·2

0· l

14·2

El Salvador

260

4-4

l ·3

5·7

Guatamala

290

2·3

-0·1

2·2

75

0·9

0·2

l ·0

Honduras

360

3·9

l ·2

5· l

Jamaica

430

4·2

0·7

4·9

Mexico

430

0·6

l ·2

l ·8

Nicaragua

320

5· l

2·9

8·0

Panama

450

16-9

l ·7

18·6

Trinidad & Tobago

590

12·5

3·3

15·9

Argentina

650

-1·7

0·8

---0·9

Bolivia

140

8·2

1-4

9·6

Brazil

220

2·9

*

3·0

Chile

450

12·9

2·8

15·7

Colombia

270

2·5

2·3

4·8

Ecuador

190

3·6

0·7

4·3

Guyana

260

10·5

§

10·5

Haiti

Aid: motives Paraguay

200

2·8

2· l

5·0

Peru

270

3-4

l ·7

5· l

Uruguay

540

l ·l

0·6

l ·7

Venezuela

780

0· l

3·6

3·7

2·8

0·3

3·2

Asia: of which: Iran

210

0-4

0·3

0·7

Iraq

240

10

0·2

l ·2

1,070

41·8

8·5

50·3

Jordan

220

25·9

10·7

36-6

Kuwait

3,290

t

--4·0

-3·9

Lebanon

390

0-4

3-4

3·8

Saudi Arabia

190

*

-1 ·l

-1 ·l

Syria

180

-0·3

0·9

0·6

Yemen

90

l ·l

0· l

l ·l

Afghanistan

85

3·2

0·2

3·3

Burma

60

0·7

*

0·7

Ceylon

130

l ·6

0·2

l ·8

India

90

2·2

0·3

2·5

Nepal

70

l ·6

0· l

l ·7

Pakistan

90

3·7

0·5

4·2

Cambodia

120

l ·6

0·3

l ·9

China (Taiwan)

190

3·9

0·8

4·8

Hong Kong

320

0·5

0·6

70

0·6

* *

120

7·0

0· l

7· l

60

22·8

0·3

23· l

Malaysia

260

2·2

l ·3

3·5

Philippines

140

2·2

0-4

2·7

Thailand

110

1-4

1-4

Vietnam (South)

110

22·1

l l

Israel

Indonesia Korea (South) Laos

Oceania: of which:

0·6

22·1

7

8

The economics ofaid

Papua & New Guinea

§

Total recipient countries

37·3

l

37·3

3·6

0·5

4· l

* Major recipients of Sino-Soviet aid conunitments in recent years include India, Indonesia, Pakistan, Iran, U.A.R., Syria,Afghanistan, Brazil and Algeria. Source: Development Assistance Review, 1968, p. 271 t = negligible l = incomplete § = not available

the recipient. If a recipient's capacity for development is indicated by its recent growth rate then, once again, statistics do not reveal that the providers of assistance systematically allocated their aid in this way. Ofcourse, 'political' influences are least likely to show up statistically in aggregate data on foreign aid since the political and strategic interests and ties ofdifferent aid-giving countries Table 3 Relationship between aid receipts per capita, 1964-6 and gross domestic product per head, 1965 Number ofcountries Gross domestic product Average net official per capita aid receipts (Bilateral & multilateral) Under$2

Under $100

$100 to under $200

$200 to under $300

$300 to under $500

4

4

4

2

19

9

$10 to under $20

3

$20 to under $50

2

$2 to under $4· l *

$4· l * to under $10

4 5

$500 Total

4

18

4

3

40

2

3

8

3

10

3

$50 and more Total

* Average of all recipient countries

13

25

19

11

12

80

Source: Development Assistance Review, 1968, p. 147 are likely to offset one another. However, data in Table 2 reveal a few countries which receive extremely high levels of bilateral aid per capita which clearly reflects their strategic position on the fringe of the Communist Bloc or in areas of vital concern to the Western powers (e.g. Vietnam, Laos, Jordan and Israel). A systematic relationship between the scale of aid and size of receiving country is observed for total D.A.C. aid distribution and also for individual aid-giving countries. There appears to be some minimum quantum of aid regardless of size of country, which means that small countries receive disproportionately large amounts of aid. This result may reflect development considerations in so far as small countries require assistance for minimum irreducible administrative overheads. It may also reflect political influences-the relatively favourable treatment which has been extended by major powers

Aid: motives

9

(especially U.K. and France) to small colonies which have recently become independent. This 'dependency' effect reflects the obligation imposed on the assisting country by historical ties and also the desire to preserve friendly future ties with ex-colonies. After this brief review of influences on the actual pattern of the allocation of net flows of official financial resources, it is necessary to look more carefully at the aid objective which is of most interest to the economist-aid for development.

Aidfor development Official aid is often provided ostensibly for the development of the receiving country although it has already been pointed out that this may be only one objective among many of aid-giving countries. It is now necessary to consider more carefully the meaning of development and how aid is expected to assist it. The term 'development' covers many different types of change. From various academic viewpoints it includes notions of industrialization, urbanization, modernization, westernization or changes in value systems from those which are tradition-based to those which are achievement-orientated. From an economic perspective development can be measured (though not defined) by the market value of the quantity of goods and services available per person. Development can then be said to occur when there is an observed increase in gross national product per capita, i.e. when G.N.P. grows faster than population. The inadequacies of this statistical measure of develop-ment are all too obvious. It cannot be identified with an improvement in economic welfare for an arithmetic mean does not yield even the crudest indication of the distribution of output (income) between people, groups or regions. It cannot be identified with an improvement in total welfare for it takes no account of the displacement of traditional crafts, the dislocation of kinship groups and the tensions that arise in a society where new values and aspirations (required for increased production for the market) conflict with traditional ones. 2 However, this crucial issue of what really constitutes development and what would be an adequate index of development takes us beyond the realms of economics into sociology, psychology, anthropology, etc. The reader who wishes to pursue these much wider and deeper matters is referred to a short bibliography at the end of this book. In order not to be grounded from the start the following economic analysis will use G.N.P. per capita as a measure of development despite its obvious deficiencies. With development measured this way the contribution of aid to development will depend on the growth of population and the growth of G.N.P. Population growth may be viewed as given and independent of aid and G.N.P. growth, since historically, accelerated population growth appears to have resulted largely from falling death rates which in tum have been due to exogenous influences. However, population growth is clearly not completely independent of aid flows or growth of output since aid in the form of technical assistance (e.g. medical assistance) may reduce death rates while increased output of domestic goods and services affects the health and the nutrition of the population and hence mortality rates. These relationships and their importance have not been fully examined or estimated by social scientists. For this reason attention here will be focused only on the contribution of aid to growth of G.N.P. of the recipient economy.

10

The economics ofaid

Economists often take a formal approach to the aid/growth relationship. A model of an economy is built with a number of variables which have technological or behavioural relationships and which are often fixed coefficients. The growth of G.N.P., the key variable, is then seen as limited by certain domestic constraints in the economy. These constraints on growth may be classified into three types, all of which may be relieved by external resources. First, there is the savings constraint on growth which is familiar from the Harrod-Do mar type model of economic growth. Secondly, there is the constraint imposed by the domestic supply of certain strategic goods. This arises if all outputs require inputs in fixed coefficients as in a Leontief type of input-output model of an economy. Thirdly, there may be a constraint on growth imposed by the domestic availability of certain types of skills. The role of external assistance can then be seen as supplementing those domestic resources which constitute the dominant bottlenecks on growth. This view of the different roles of aid in growth is still somewhat controversial. Not all economists agree that these are separable constraints on growth, and this aspect of economic theory is discussed more fully in Chapter 3. It must be stressed at this stage, however, that the economists' approach, via models and bottlenecks, which is spelt out later in this book, is a somewhat mechanical one. Growth is seen essentially as something which can be programmed by feeding in the appropriate quantities of the appropriate resources. It largely ignores the qualitative factors, the underlying value systems, human responses to financial and other stimuli, institutions and the policies of the domestic government, or else it assumes that these will be favourable to growth. Although there is little possibility that inflows of foreign assistance will have any direct and significant effect on local values, attitudes and institutions, the provision of aid may be made conditional on the domestic government following policies which seem likely to favour growth. Whether this potential 'leverage' effect of aid constitutes an addition to the purely 'resource' effect will depend on whether such a strategy is 'politically' acceptable to the aid-recipients and whether the aid-givers know more than the aid-recipients about the types of policy which are most appropriate in the particular conditions of the country concerned. The 'leverage' effect of aid on growth is clearly something which cannot be subjected to the formal analysis which is later applied to the role of assistance as supplementation of specific domestic resources. Notes

2

The D.A.C. countries includeAustralia,Austria, Belgium, Canada, Denmark, France, Germany, Italy, Japan, Netherlands, Norway, Portugal, Sweden, Switzerland, United Kingdom, and United States. Development would seem to involve a basic paradox that if people want to get more of what they value then changing their values is a prerequisite for satisfying this desire.

2

Aid: allocation principles

The description in the previous chapter does not represent, in any sense, an ideal situation. Aid has not been clearly defined in practice. The provision of development finance has been a largely nationalistic and often competitive activity and only limited movement has been made towards viewing and organizing assistance as a co-operative international activity in which the burden of aid might be shared equitably between countries and the allocation of aid might be based on clear common objectives. This chapter will examine some of the important conceptual issues that would arise in a rather more ideal world of international co-operation for foreign assistance to less developed countries. This will cover (a) the definition and measurement of genuine aid; (b) the international sharing of the cost of assistance; (c) the optimum allocation of assistance among different countries.

The definition and measurement ofaid What is aid? Aid is a gift. However, only a part of the financial resources transferred from rich to poor countries is actually in the form of outright grants, while loans clearly cannot be entirely gifts if they have ultimately to be repaid with or without interest. There may be a gift element in loans and if this can be measured then grants and loans on different terms can all be reduced to one comparable standard. From the point of view of the lender, 'aid' is the real cost of providing the loan, and the real cost of the resources lent or given is the benefit forgone by not using them in their best alternative use. In this calculation of the opportunity cost of assistance there are two major considerations. First, the cost will depend on the terms of the loan, interest rate, grace period and repayment schedules. Second, it will depend on the conditions of the loan or grant, i.e. whether the goods which can be obtained by the recipient are valued at their opportunity cost. In other words the real cost of loans depends on the gift element in money terms but also the extent to which the value of commodities made available under these loans differs from their value if the donor disposed of them in a different way. It is useful to begin by ignoring the second complication and take it up later. The next three sections may prove difficult for students who have not been introduced to the concept of the discount rate and its use in investment economics. They might consult a companion volume in this series, J.L. Carr's Investment Economics.

The terms ofassistance The real cost of aid will depend on the terms on which the monetary flows are provided. If they are in the form of grants then the nominal value and real cost will be identical. The donor

DOI: 10.4324/9780203840177-2

12

The economics ofaid

has forgone the use of these financial resources forever. lf they are in the form of a loan the lender loses the use of these resources initially but their use is not completely forgone but only postponed since the loan will be repaid in future instalments and the lender may also receive interest payments. Hence the real cost is the difference between the present value to the lender of the future flow of repayments and interest and the magnitude of the initial loan. In assessing the present value of future return flows from the borrower it is necessary to consider the lender's alternative policy of investing the loan funds at home at a positive rate of return. To put the question another way, what sum invested now in the donor country at the current rate of return would yield a flow of income equal to the repayments and interest expected from the borrower? The answer lies in discounting the future repayments and interest at the rate of return which could have been earned by the lender if, alternatively, the funds had been invested at home. The present value of an aid loan will then depend on the rate of interest charged relative to the rate of return on investment in the donor country and the period of repayment of the aid loan. The higher the rate of return in the donor country (given the rate of interest on the loan) the greater will be the opportunity cost to the donor. Also the longer the period of repayment of the loan, other things being equal, the greater will be the cost. Thus an interest-free loan of £100 repaid in equal annual instalments over 25 years will cost the donor £30·4 if discounted at 3 % but £48·9 if discounted at 6 %. lf it is repaid over 50 years the cost will be £48·5 and £68·5 respectively. The loan is costless to the lender if the rate of interest is identical with the lender's alternative rate of return. lfthe rate of interest on the loan exceeds the lender's opportunity cost the lender will, of course, benefit from aid on these conditions. Table 4 sets out the cost or grant element of loans on different terms. When real aid flows are distinguished they emerge as considerably smaller than foreign assistance flows conventionally defined. Table 5 shows statistics for 1965. It is noticeable that the level of genuine aid was on average three-quarters of conventionally measured official assistance from D.A.C. countries in 1965. There was also considerable intercountry variation in the grant element of assistance although the use of this measure does not significantly change the country 'ranking' in aid performance for 1965.

Table 4 Grant element in loans with different interest rates and maturities Lenders' rate ofdiscount Terms ofloan Rate ofinterest 2%

3%

4%

5%

6%

7%

8%

9%

10%

20

22·1

27·8

32·8

37·6

41·7

45·5

Maturity* (yrs.) 30

28·9

35·7

41 ·5

46-6

51 ·0

54·7

40

34·2

41 ·5

47·5

52·5

56-9

60·5

20

14·7

20·8

26-3

31 ·3

35·8

39·8

30

19·3

26-8

33·2

38·8

43·7

47·8

40

22·8

31 ·l

38·0

43·9

48·7

52·9

20

7-4

13-9

19·8

25·0

29·8

34· l

30

9·6

17·8

24·9

31 ·l

36-4

41·0

40

11-4

20·7

28·6

35·0

40·5

45·3

Aid: allocation principles 6-9

13-1

18·6

23·8

28-4

8·9

16-6

23·3

29·2

34·2

10-4

19·0

26-3

32·5

37·7

6-6

12·5

17·9

22·7

8-4

15·5

21·8

27-4

9·6

17·5

24·3

30· l

6-3

11·9

17· l

7·8

14·5

20·5

8·8

16-2

22·6

20 5%

30

0

40 20 6%

30

0

0

40 20 7%

30

13

0

0

0

40 * No grace period in repayment is assumed.

Source: Adapted from G.Ohlin, Foreign Aid Policies Reconsidered, O.E.C.D., 1966, pp. 111-12. Table 5 Grant element and gross official disbursements to developing countries by D.A. C. countries in 1965 (]) (2) (3) 1965 Figures As% ofnational income Grant element as % of Total gross Grant element total commitments official flows in gross official flows 0·67 0·25 0·34 0·17 0·64 l ·08 0·50 0·22 0·61 0·90

U.S. Sweden Canada Denmark Australia France Germany Norway U.K. Belgium Netherlands Austria Italy Japan Portugal

81 95 79 78 100 87 62 98 75 99 84 29 33 53 59

0-41 0-49 0·22 0·37 0·75

TotalD.A.C .

77

0·61

(4)

Rank according to (3)

0·60 0·24 0·31 0·16 0·64 0·98

4 10 9 15 3

0-44 0·20 0·52 0·83 0·32 0· 18 0· 19 0·22

7 12 5 2 8 14 13 11 6

0-48 0·55

Countries are ranked in descending order of national income per head. In calculating the grantelement of gross official flows (disbursements) a 10% discount rate was used. It should be noted that there is often inconsistency between the ratio of ~~:::

g~

and colunm (1) because colunm

( 1) refers to commitments of funds and colunm (2) to disbursements.

Source: Development Assistance Efforts and Policies, 1966 and 1967 Reviews.

14

The economics ofaid

From the borrower's viewpoint the real value of a loan (or grant) may differ from the real cost to the lender. The value of a loan will depend on the rate of return on investment in the borrowing country and this may differ from that in the lending country. Thus, if the borrower's rate ofreturn (or rate of discount) is greater than the lender's, the real value of a loan to the borrower will be greater than the real cost to the lender and vice versa. If the official exchange rate of the borrowing country is a 'disequilibrium' rate, the value of the foreign exchange resources provided by the loan needs to be adjusted by an appropriate 'shadow' exchange rate. This may alter the recipient's rate of return on aid resources compared to domestic resources.

'Tying' and the value ofaid In addition to the terms of aid and discount rate, the extent to which transfers of funds are 'tied' to procurement of goods in the lending country may also affect their value to the recipient and their cost to the lender. Loans tied to use in the lending country may reduce their value to the borrower compared to a situation in which the funds could be used to buy goods from the cheapest or most preferred international source. Although tying will normally reduce the value of loans to the recipient they will not necessarily reduce the real cost to the lender. If resources are fully employed in the lending economy the real cost of the tied loan is not reduced unless the particular industries supplying the aid goods are able to exploit a monopolistic advantage over the borrower by charging prices which exceed the opportunity cost of the goods supplied. This is possible but not inevitable and assumes collusion among firms in an industry when meeting orders on tied aid. If aid purchases are tied to industries where resources are unemployed, on the other hand, the opportunity cost of the goods supplied may be less than the price actually charged. In the short-run, capital and skilled labour in an industry (e.g. shipbuilding) may be immobile and their opportunity cost zero, even if materials used in production have alternative uses. Tying aid to 'surplus capacity' industries of this kind clearly reduces the real cost of aid on the assumption that the government would not otherwise undertake spending to activate the idle capacity for domestic use. Aid funds tend, in practice, to be spent on 'growth' industries where there is normally little idle capacity. The countries receiving aid often have their own capacity to produce the more 'traditional' products and these tend to be the 'declining' industries in the lending countries where excess capacity may well exist. Hence, except where aid is tied to specific industries, it would be somewhat coincidental if aid expenditure generally fell on industries with spare capacity. When tying takes the form of resources provided in kind, the 'gift' element of these resources will depend on the valuation placed on the goods, even when they are provided nominally on a grant basis. Food (wheat) surpluses provided under U.S. Public Law 480 1 are an important case in point. In practice these food smpluses have been partly valued at a price originally paid to U.S. farmers under the price support programme and partly valued at current world market prices which are lower. In both cases the wheat is overvalued. An extreme view would be that the resource cost of surplus food output to the U.S. is zero because the factors producing it are immobile-they have no alternative employment. The value to the recipient of this food grant, however, would be positive depending on

Aid: allocation principles

15

the international price that it would have had to pay for food. However, even the current world market price may overstate the alternative value of the wheat to the U.S. because the quantum of wheat supplied on U.S. aid has often constituted a large proportion of total world trade in wheat (often 25%). Hence, the alternative course of selling the 'aid' wheat in the world market would depress prices unless the elasticity of world import demand were infinite. More plausibly, assuming the elasticity of demand not to exceed unity, the alternative value of P.L. 480 wheat would have been at least 25% less than its value at current world prices. This would reduce the real cost of food aid to the U.S. but would not alter its value to the recipient. To sum up, therefore, it has been shown that the nominal value of 'aid' flows normally recorded in the official statistics of donor countries, do not necessarily represent the real cost to the donor country (or the real value to the recipient). The true aid element in socalled 'aid' programmes can be derived only when allowance has been made for the terms on which the funds are provided and adjustments made for goods supplied on 'aid' whose recorded value exceeds their opportunity cost. Efforts to share the cost of aid equitably among donors should clearly be based on this real cost concept. This issue of burden sharing will be taken up a little later.

Minimization ofthe cost ofaid The preceding analysis provides a measure of the cost of foreign assistance to the countries providing the assistance; it also provides criteria for minimizing the burden of assistance to the donor. Of course, one way of minimizing the cost of foreign assistance is to provide none. However if the provider of assistance aims to achieve a given real present value of benefit to the recipient, then it is possible to minimize the cost of doing this by varying the terms of assistance. The following formal propositions can then be established. 1 To achieve a given present value of benefit to the recipient, grants cost the donor less than loans if the yield on capital is higher in the donor country than the recipient country and vice versa. Although the general proof of this proposition requires mathematics the reason is apparent. Resources are being shifted from an economy which uses them more effectively to one which uses them less effectively. To achieve the same present value to the recipient, a loan must transfer more resources sooner than a grant because it has to be serviced. The earlier the resources are transferred the greater the cost to the provider of these resources. This can be illustrated by an example. Suppose the rate of return on investment is 7% in the recipient economy and 8% in the donor economy. Let the choice lie between a £100 loan for 50 years at 6% or a grant equal to the present discounted value to the recipient of the flow of benefits from the loan. The grant to provide the recipient with the same benefit would have to be £13·8.2 What is the cost to the donor of these alternatives? In the case of the loan the cost is £2 per annum for 50 years (i.e. difference between £6 annual interest received and £8 yearly alternative return if £100 were invested at home). The present value of this annual cost, discounted at 8%, is £24·5. The cost of the grant is £13·8 and this is clearly more economical.

16

The economics ofaid

On the other hand, if the rate of return on capital is higher in the recipient country than the donor country, a loan would cost the donor less than a grant for the same aid benefit to the receiver. It is worth noting that grants always cost the donor something but a loan need not. If the yield on capital is larger in the recipient than in the donor economy it is clearly possible to have an interest rate which provides a net gain to the recipient and the donor. 2 To achieve a given present value of benefit to the recipient when the yield on capital is higher in the donor than in the recipient economy: (a) Smaller loans at lower rates of interest cost the donor less than larger loans at higher interest rates, and (b) smaller loans for longer periods cost less than larger loans for shorter periods. These formal propositions are another application of the principle set out in 1 above. The policy implications of these criteria seem clear but their application in practice is somewhat qualified. First, since the rules assume that donors and recipients assess grants and loans by their present value it requires the recipient to be capable of investing and disinvesting freely at the yields on capital prevailing. Since projects financed by foreign assistance are usually large and lumpy the yields at any stage in the project may not reflect yields elsewhere. Further, yields are rarely known with accuracy in different economies even if it is assumed that they are equated at the margin in all investments. Thirdly, recipients may default on loan repayment. The existence of this uncertainty may oblige the donor, who would otherwise choose loans, to provide grants to ensure that recipients with low yields on capital receive any benefits at all (i.e. to cover the risk of default by a higher interest rate may raise the interest rate above the rate of return in the recipient country). Finally, there are also non-economic reasons for deciding the terms of assistance. For example, psychologically and politically it may be desirable to provide loans rather than grants. The notion of 'charity' can have a demoralizing effect on the recipient.

International cost sharing The measure ofaid described above provides a standard for comparison ofthe true contribution of different nations to the development process in poor countries. It would, therefore, form the basis for any scheme for sharing the cost or burden of assistance between countries. So far no comprehensive scheme of this kind has been developed, although there have been some ad hoc moves towards greater joint effort in the provision of aid. If foreign assistance did become entirely an area of co-operative international activity there would have to be some method of cost sharing devised. This would be an exercise in international public finance, and it could not be solved purely on the basis of economic theory. However, the literature of public finance has produced two main principles for the sharing of costs of certain public services between different groups within national boundaries. These two principles-the Benefit Principle and the Ability to Pay Principle-will now be examined in the context of international assistance for development of the poorer nations. a Bene.fit principle

The essence of the benefit principle is that each contributor makes a payment that is a quid pro quo for the benefit received. It is a principle that is unlikely to provide a guide for

Aid: allocation principles

17

international aid sharing because many of the benefits to the donors will be 'collective'. Indeed, it might be argued that cost-sharing arises only when the benefits are diffused and not directly related to the individual donor's aid. There are some benefits from aid-giving such as immediate commercial advantages (export promotion) or the 'moral glow' from providing aid from humanitarian motives, which accrue directly to the nation providing the aid and to no one else. However, some of the benefits (real or imaginary) are collectively enjoyed. In the case of the economic gains 3 of aid-giving it is possible, in principle, to distinguish the gains accruing to individual donors but virtually impossible in practice. When one of the benefits of aid is believed to be improved world 'security' it is clearly a collective benefit which cannot be allocated among the contributors.

b Ability to pay principle The 'ability to pay' principle is based on the idea of equal sacrifice, but in determining criteria for sharing the cost of 'aid' several important conceptual and empirical problems arise.

a First, should ability to pay be formulated in terms of 'nations' or 'individuals' as units? On grounds of convenience nations would be chosen, but in an integrated world community it may be contended that the burden of international activities for the common welfare should be related to individual capacities to pay rather than the capacities of nations. b Second, there is no objective basis for inteipreting the 'ability to pay' principle as a proportional or progressive 'tax' system (or for determining the actual progressive rate structure to be used internationally). If the 'nation' is to be the unit, then equality of sovereign states might suggest a proportional basis for burden sharing. On the other hand, most nations follow a progressive system of taxation internally, and this would suggest a similar system internationally. 4 If a progressive system is chosen, assessment of the aid contributions on the basis of per capita national income ignores the problem of equity between individuals in different nations. For this reason some economists have suggested that international taxation should be based on a 'world' distribution of income among individuals with each country paying a part ofthe total equal to the proportion its nationals would pay under a worldwide progressive personal tax system. This would still leave the finance to be raised by national governments and there would remain scope for inequity depending on the particular tax structure used by each government or whether the provision of aid involves incremental taxation which may not necessarily be progressive at the margin. c Thirdly, there are problems in comparing the national income of different countries owing to differences in concept and coverage of national accounts statistics and the problem of using official exchange rates which do not always provide an accurate guide to the relative purchasing powers of national currencies. Similar problems would apply to data used for a 'world income distribution'.

Allocation ofaid The assumption of international co-operation to share the burden of aid-giving does not necessarily affect the distribution of aid, since individual donors could distribute their

18

The economics ofaid

share of aid according to their own national objectives. However, let us now assume that aid distribution is handed over entirely to some multilateral agency. It might then be assumed, somewhat unrealistically,5 that such a body would be 'politically disinterested' in aid-giving, unlike most national donors whose country-wise aid allocations are influenced by historical ties, ideology, national foreign policy purposes, etc. If this were the case, on what principles would the multilateral agency distribute its aid resources? Can any purely objective criteria be devised? Many criteria or aid strategies have been suggested. Three are discussed below: (a) The redistribution principle. (b) Achieving 'self-sustained' growth. (c) Maximizing the development effect of aid.

I Pure redistribution ofincomes Given the inequality in per capita incomes between nations there is a case on ethical grounds for poor relief-an international redistribution of income from rich to poor involving some form of 'utility' weighting. This might take the form of a progressive tax system on the rich countries and subsidies to the poor countries which would vary inversely with their poverty. On this principle there would be no obligation for the subsidy receivers to use the resources for development purposes and they could use them to make a net addition to their own consumption. The donors would be obliged to provide assistance entirely in grants. There are some obvious objections to this principle in its pure form. Consumption subsidies would have no lasting effect,6 in situations where population growth impels some development effort. Moreover, in practice, inadequate resources are likely to be redistributed to make much impact on the degree of international inequality.

2 Achievement ofself-sustained growth Another possible strategy is to concentrate limited aid resources on those countries which are likely to achieve 'self-sustained growth' within a given period of time. The idea of self-sustained growth is that a judgment can be made about the possibility of a country achieving and maintaining a specified growth rate eventually without external assistance. Given a growth rate considered 'satisfactory' it is possible, in principle, to predict the time required for an economy to become independent of external assistance while achieving the desired growth rate. One method of doing so is to assume that growth is largely a function of capital accumulation and to construct a model of an economy which places plausible empirical values on the main parameters like the yield on capital, the savings rates, etc. (The following chapter is devoted to this particular problem.) Since some economies would have greater growth potential than others it is possible to have a strategy of concentrating aid on all those countries likely to sustain the required growth themselves within (say) 10 years or 20 years. This is the strategy of getting 'as many horses past the finishing post as possible ina given time' .7 The aid distribution which would result from this principle would be very different from that based on the pure redistribution principle since the 'horses' likely to finish quickest are not likely to be the poorest ones. There would seem to be little or no ethical or political justification for such a strategy. Is there an economic justification?

Aid: allocation principles

19

Some economists have put forward the hypothesis of a 'population trap' to economic development. One crucial proposition in this hypothesis is that up to a point an increase in income per capita will induce (with a lag) an increase in the rate of population growth which offsets the increase in the rate of growth of income feasible at the higher standard of living. Aid which raises income per capita will only lift a country clear of this trap and allow income growth to exceed population growth if it is available in sufficiently large doses. If limited aid were concentrated on a few countries therefore, it might allow them to break out of the trap. If spread over many it might lift none out of the trap. The hypothesis is somewhat extreme: it is doubtful if population growth is such a rigid function of income standards and there needs to be a particular combination of empirical values of the important growth variables for the hypothesis to hold in practice in many countries. 8

3 Maximization ofdevelopment effects ofaid A criterion fairly close to the allocation rules of orthodox economic theory is that scarce aid resources should be distributed between countries so that at the margin a unit of aid has the same development impact everywhere. The 'development effect' of aid may then be formally defined as the present value of all future increments of consumption resulting from the aid. This calculation gives rise to two problems. First, there is the problem of estimating the effects of various quantities of current aid on future consumption in each country. In principle, 'objective' estimates of this kind can be made which will be discussed shortly. Second, there is the problem of comparing present and future units of consumption and comparing the value of a unit of consumption in different countries. In both cases some form of weighting is required which will express the donor's preference between present and future generations and between different countries. There is no 'objective' way of doing this and it is a matter of judgment, for example, what different weights should be given to a unit increase in present consumption in countries with different per capita income levels and different cultures. This raises all the familiar problems of interpersonal comparisons of utility and whether the utility of marginal real consumption declines. The marginal effectiveness of aid involves, at least, an assessment of the direct and indirect effects of external capital on the national income of the recipient in all future periods. The direct effect is the permanent net addition to national income-this might be called the yield or productivity of external capital. There are reasons for believing that it may be a declining function of the scale of the capital inflow in a given period of time, because of the scarcity of complementary domestic factors such as skilled manpower in the short period. 9 The indirect effect of foreign capital arises from the direct increment of income. The savings out of this extra income (the marginal saving rate) will be invested and in tum add further to income and so on. There is a multiplier effect over time whose size will depend on the productivity of capital and on the marginal savings rate in the country concerned. 10 Although normally the indirect effects of investment are ignored and income saved is not valued differently from income consumed, the justification for taking account of different national marginal savings rates in aid allocation is that the aid-giving agency may wish to discount the future at a different rate from the government or people of the recipient country. The donor establishes for each recipient its own value of aid by

20

The economics ofaid

computing all future increases of consumption (consequent on aid) while discounting them at its own rate of discount. This value of aid concept is then comparable between recipients. Notes

2 3 4 5 6 7 8 9 10

Some P.L. 480 aid is formally on a grant basis. These loans are interest free and repayment is required in the local currency. Since these repayments cannot effectively be used to buy foreign currencies or to export the recipient's products, the loans really constitute grants. i.e., the annual net benefits to the recipient of the £100 loan invested at 7% are £1 per year for 50 years (i.e. £7 minus interest charge of £6). When this stream of benefits are discounted to the present and summed they equal £13·8. Economic benefits are the increased trade possibilities for the donor countries in the long run following growth of the less developed countries from aid. Empirically it would appear that the distribution of income (per capita) between nations is more unequal than the distribution between income groups within the developed countries. It is unrealistic since such a body would reflect the interests of the major donors in its strategy as the World Bank has tended to reflect U.S. ideology in practice. In any case it would develop a political bias of its own. Unless it is assumed that higher levels of personal consumption increase people's productivity directly and significantly. See I.MD.Little and J.M.Clifford, International Aid, pp. 98-9. For a critical account of this hypothesis see H.Myint, Economics ofDeveloping Countries, Ch. 7. The productivity of external capital is a complex concept. The net addition to national income resulting from foreign capital may exceed the yield on capital since idle natural resources or unemployed manpower may be brought into use. At infinity, one unit direct addition to annual national income (from invested capital) will have grown to _I_ where 8=output/capital ratio and =marginal saving rate. 1-Ss

s

3

Aid and economic growth I

Aid and growth The transformation of a poor country with a low or zero growth rate into one capable of an adequate sustained growth rate is the essence of the development problem. 1 Faster growth requires an improvement in the skills of a country's labour force, a growth in its capital stock, and substantial changes in the composition of output and accompanying changes in attitudes and institutions. Although only the most mechanically-minded economist believes that there is a fixed relationship between foreign resource flows and the growth rate of the receiving country, nevertheless, it is helpful to spell out formally the ways in which foreign assistance can augment a country's capital stock, imports of specific commodities and skills and so remove some of the constraints on growth in a society which would otherwise have to depend entirely on its domestic resources. This section will attempt to show that the impact of foreign assistance on the growth of a poor country will depend on what are the dominant constraints on growth-whether savings, foreign exchange availability or skills. It will go on to consider the conditions under which aid flows will permit growth to become ultimately self-sustaining.

Two roles offoreign assistance If skill problems and other institutional requirements for growth are ignored there are two views of the contribution of foreign capital assistance to economic growth. 2 The first view sees the maximum level of domestic savings as the dominant constraint on growth, and foreign capital as a supplement for domestic savings which permits higher investment and growth. In this view, domestic capacity is assumed to provide the required import of foreign goods, (via exports) provided domestic consumption is restrained enough to allow the desired level of exports. The only constraint on growth is that imposed by the maximum feasible domestic savings. The second view stresses that many goods have strategic importance for growth but cannot be produced domestically at early stages of development and have to be imported. It questions the capacity of certain poor countries to freely export their domestic output either because of international demand conditions (with restricted access to markets in the developed countries and inelastic foreign demand curves) or because of domestic supply conditions at low levels of economic development. Bottlenecks may, therefore, arise because of inadequate supplies of certain foreign goods. The maximum potential domestic savings of a poor economy may not be fully used for capital formation and the dominant constraint on growth will be availability of foreign exchange. Foreign assistance in this context is required primarily not to supplement domestic savings but to relieve the import

DOI: 10.4324/9780203840177-3

22

The economics ofaid

constraint on growth. When this bottleneck constraint holds, foreign assistance will have a greater effect on the growth rate than if the savings constraint is binding. Hence two separable constraints on growth-savings and imports-are identified. The contribution to economic growth of a given flow of foreign assistance will differ depending on which is the binding constraint in an economy. 3 Alternatively, the requirements of foreign assistance to achieve a given growth rate in an economy will vary according to which constraint is considered the dominant one. (This issue is considered in the next chapter.) Let us now consider a simple model of an economy receiving no foreign capital where the dominant constraint on growth, in one case, is set by the maximum feasible savings rate and, in the other case, by the maximum import (export) rate. It is then possible to show the differential effect of foreign capital assistance on the growth rate. It is initially assumed that the only imports are capital goods which the economy, at an early stage of development, does not have the physical capacity or skills to produce domestically. The income (Y) of the economy is assumed to grow at the same rate as its output potential (P) because there is assumed to be full utilization of available capacity. Assume the economy has the following type of aggregate production function where Ka and Km represent imports of domestic and foreign capital goods respectively which are required in fixed proportions. (1)

This notation means that output is given by whichever bracketed expression yields the smallest value. The units for capital goods are such that one unit of output capacity (P) can be used to construct one unit of Ka or exchange for one unit of Km at fixed terms of trade (i.e. the price of all the inputs and outputs is one). To increase the potential output of the economy requires net investment expenditure on domestic and foreign capital goods and the level of investment (I) is determined by the propensity to save (s) and current income. The extra output obtained from unit of investment

(d/)

and therefore the growth rate made possible by a given savings rate (s) will depend on whether there are any constraints on the availability of the two inputs-domestic and foreign capital goods, other than that set by availability of savings.

Savings limited growth We may assume that units of domestic output are freely convertible into foreign capital goods, because the export capabilities of the economy are sufficient to finance the required imported capital goods arising from the investment level. In this case the growth rate will be constrained only by the domestic savings rate (s). This is the usual assumption of the Harrod-Domar type models. The growth rate of the economy is given by 15s where 15 is the output/investment ratio. What is the value of 15 given the above production function? Since there is no constraint on either input (other than that set by available investible resources) foreign and domestic capital goods can be combined in the appropriate proportions. The extra output arising from a unit of investment

Aid and economic growth I

. (savmg)

Brx = dP -/ = dY -I = c5 = B+rx

4

23

(2)

Trade limited growth However, if the export capabilities of the economy are not sufficient to finance the required imported capital goods then the growth rate 15s will not be achieved and a foreign exchange bottleneck will exist. Suppose that maximum possible exports are a constant proportion (x) of domestic output (Y). Then, the maximum possible rate of growth of the economy, whatever the savings rate, will be Bx. This follows from the nature of the production function assumed in the earlier section, since x represents the maximum proportion of output which can be converted into imported capital goods (K,). If BxO, m> 1

(5)

It is assumed that the value of output exceeds the value of imported current inputs (M) so that m is greater than unity. If the maximum export rate of the economy (x) is inadequate I to sustain the necessary current imports, capital transfers (f) will have to exceed ;;; - x for any growth of capacity to occur at all. In other words, a foreign aid transfer will not

Aid and economic growth I

25

release a bottleneck (import) on growth unless it is more than adequate to finance the uncovered current import needs of the economy. If foreign aid exceeds the current import

x)

x

gap ( i.e. f > ~ then the economy will grow at a rate B ( + f - ~). See Figure 2. For aid flows over and above current import needs the impact on the growth rate will be determined by the value of B. It should be noted in this case, that only the proportion of domestic

P( I-~))

output (P) which is net of imported current inputs (i.e. is available for domestic consumption and capital formation. The savings propensity out of net output must be large enough to ensure thats, the savings rate, is the sanie as in the case where no imported current inputs are required. Suppose, now, to take a more extreme case, an economy had developed an advanced domestic capital goods industry so that no foreign capital goods were required for production, but its export capability was not adequate to finance the required current imports to ensure Figure 2 Effect ofcapital inflow on growth rate

Growth rate(r)

I I I I I I I

0S1

I

I I I I I

o,,

I

111111111,

Capital inflow f

f1

rm (rate of growth with import constraint)

=B(x+f-~)

rs (rate of growth with savings constraint) = b(f + s) 1

1

/1 = -m - x where -m > x full utilization of its domestic productive capacity. (This is not unlike the situation which confronted India in the early 1960s. India, of course, required imported capital goods but

26

The economics ofaid

the pressing problem was its inability to finance large enough 'maintenance' import requirements.) Foreign capital which fills some of the gap between export earnings and current import needs would) of course, raise utilization of existing capacity but it would not

¼- x)

it would affect the level of output permit any growth of capacity. Iff was less than ( (income) but not the growth rate. However, if the aid flows are adequate to fill the whole of the current import gap it would have an effect on growth. In terms of Figure 2, if the rate of capital inflow were less than.I; no growth would occur. Beyond the critical level off; aid, growth will occur at a rate rm constrained by foreign exchange availability. The analysis suggests that aid can be 'developmental' or growth permitting, even if it does not finance the import of capital goods. Assistance for financing adequate imports of current materials inputs in this situation permits growth of output. It enables a country to effectively use its existing capacity as well as its domestic savings potential to add to capacity. Of course, even if aid is not adequate to permit capital accumulation it is still justifiable as a 'rescue operation' to relieve the current import constraint and allow fuller utilization of its existing capacity until the export rate can be raised.

The role ofaid in the labour surplus economy So far we have ignored labour inputs in production on the assumption that the supply of unskilled labour does not constrain growth in most poor countries. The main constraints are maximum feasible domestic savings, rates or availability of imported capital and current materials. However, let us suppose that foreign capital goods and materials are not required for production and further that investment in domestic capital goods only requires labour. In this case it would appear that in a surplus-labour economy, capital accumulation could proceed using idle labour without any domestic saving or import constraint; provided labour is transferred from agriculture (say) where its marginal product is zero, to work on capital construction at the same real wage as before, and provided those remaining in agriculture do not consume any extra goods after the departure of their colleagues, there appears to be no limit on capital accumulation while idle labour exists and no need for foreign capital inflows. However, if we ignore the unreality of the production function assumptions, there is an important reason why labour-intensive investment may be limited. Institutionally, there is a problem in most poor countries in preventing a net increase in consumption when idle labour is transferred into investment activity. It is politically difficult (via taxes, etc.) to prevent a rise in consumption per capita in the agricultural sector from which the idle labour is normally drawn. In this case consumer goods cannot be released beyond a certain point for feeding the newly employed workers. This institutional constraint on extra capital formation can be relieved if a foreign capital inflow permits the import of extra consumer goods (mainly food) required by the investment workers. Aid in this case not only permits a fuller utilization of the productive potential of labour in the economy, it also allows a higher rate of capital formation. Food aid which can often be justified as a 'relief' measure would in this context permit greater capital formation. 7

Aid and economic growth I

27

Aid and self-sustained growth If a positive rate of aid flow (aid as proportion of G.N.P.) is necessary for achieving the desired growth rate, the above model has the implication that aid transfer in absolute terms must rise continuously over time. This follows from the assumptions of the model; that savings and export propensities are constant, that possibilities of import substitution are ignored and that there are fixed coefficients between production inputs. When these assumptions are relaxed, the possibility that aid transfers can eventually be phased out emerges, so that aid can be viewed as a 'pump priming' instrument which ultimately permits self-sustained growth. It is possible to spell out the conditions necessary to ensure that the specified growth rate ultimately becomes self-supporting for any particular economy. This is an important calculation because neither donors or recipients find the prospect of continuously rising aid flows attractive. There are clearly many possible configurations of structural changes which could eventually ensure self-sustaining growth. Two simple cases may be considered, assuming first, a savings constraint, and, second, a trade constraint on growth, but assuming that the production function remains unchanged.

Savings constraint only Consider an economy in which there are no export limitations on growth but only a domestic savings constraint. Initially the domestic savings rate is too low to permit the desired growth rate without foreign assistance but the required foreign capital is provided. What must be the nature of the domestic savings function if the desired growth rate is to be achieved but foreign aid injections are to fall to zero over a finite period? Let The target growth rate= f Incremental output/capital ratio = 15 Then the required investment rate=

f

Also let the absolute level of income, investment, domestic savings and foreign assistance be Yf If S, and F, respectively:. r

I_n_ t = 0, investment lo . k. Y0 and savings S0 positive to ensure the reqmred mvestment. Income grows at rate f, so

Ye

=

Y0 (1

+ f)'

=

i. -

Y0-F0 , since foreign aid is initially

(6) (7)

(8)

where

S=marginal domestic savings rate.

28

The economics ofaid Hence foreign assistance in year t=F1=J1-S1

= (

J- s) (

Yt - Yo)

+F

0,

(F0 > 0)

(9)



- ........ f

For F, to decline s ~ In words, for self-sustaining growth to be achieved the marginal domestic savings rate must exceed the required investment rate. Otherwise the maintenance of the growth rate,

J°,

f , would require increasing injections of foreign aid over time. The more S exceeds the more quickly will foreign aid be phased out in the absence of other structural changes in the aided economy.

The prospects for self-sustained growth Does the performance of developing countries in recent years appear to satisfy the savings criterion for self-sustained growth? The criteria in terms of this model are that empirically the average domestic savings rate (s) or the marginal domestic savings rate (s) should be equal to, or exceed, the investment rate required to maintain a satisfactory target rate of growth of G.N.P. (f) ( i.e. s ~ ~ or s ~ ~) . If it is assumed that 0·058 represents a satisfactory minimum rate of growth for all (f) developing countries, then the desired investment rate for each country can be calculated using values for the output-capital ratio (15) which have been recorded in these countries in a recent period. Table 6 sets out data for 31 developing countries which refer to the period 1957---62. In so far as recorded performance in this short period is a true indication of each country's future

Table 6 Performance ofdeveloping countries 1957-62 It is assumed that the desired rate of growth of all these countries (f) is 0·05. The 'savings' conditions for self-sustained growth are that S >

§or S > §·

Actual savings performance Country

Capital Required inflow 1962 investment rate (FIY)

(0

Average rate (s) 1962

Marginal rate Actual growth Cs> 1957-62 rate (G.NP) 1957-62

A Countries meeting the saving criteria Burma

0·00

0·205

0·17

0·21

0·046

Israel

0·20

0·154

0·11

0· 15

0·103

Jordan

0·24

0·068

-0·07

0·09

0· 111

Korea

0·10

0·172

0·03

0·27

0·040

Aid and economic growth I Malaya

-0·04

0·116

0·22

0·26

0·062

Pakistan

0·04

0· 117

0·09

0·25

0·041

Panama Peru Philippines

0·06

0·156

0·12

0·37

0·051

-0·01

0· 155

0·21

0·31

0·073

0·02

0·139

0·12

0·30

0·050

Taiwan

0·07

0·134

0· 15

0·29

0·074

Thailand TrinidadTobago

0·01

0·106

0·16

0·22

0·080

0·10

0·217

0·22

0·11

0·078

Argentina

0·03

0·533

0·21

0·83

0·019

Brazil

0·03

0·132

0· 15

0· 19

0·067

Greece

0·06

0· 151

0· 15

0·26

0·060

Honduras India Nigeria

-0·01

0·203

0·13

0·25

0·033

0·02

0·145

0·12

0·20

0·048

0·05

0· 185

0·09

0· 19

0·033

29

B Countries not meeting the saving criteria

Iran

0·01

0·177

0·14

0·11

0·049

Mexico

0·01

0· 153

0·13

0·11

0·050

-0·06

0·326

0·27

---0·26

0·043

Venezuela Bolivia

0·07

0·216

0·04

---0· l 6

0·029

Chile

0·06

0·138

0·07

0·10

0·038

Colombia

0·04

0·208

0·16

---0· 12

0·050

Costa Rica

0·05

0·236

0·11

---0· 10

0·039

Guatemala

0·02

0·176

0·08

---0·03

0·036

Liberia

0·56

0·390

0·11

0·21

0·046

Mauritius

0·09

0·249

0·10

---0· 39

0·034

Paraguay

0·03

0·318

0·13

0·08

0·026

Tunisia Turkey

0· 18

0·245

0·08

---0·84

0·034

0·03

0·240

0·12

---0·02

0·030

Source: Data derived from H.Chenery and A.Strout, 'Foreign Assistance and Economic Development', American Economic Review, September 1966, Table 6. savings capabilities, it may be concluded that 18 of the 31 countries satisfy the saving condition for self-sustained growth. Not all of these countries were receiving foreign assistance in 1962; of the 26 receiving aid, 14 satisfied this test. However, recent statistical evidence on the magnitude of saving and other parameters may be very misleading for prediction-an issue which is discussed in the next chapter. For this reason an alternative approach to the estimation of long-run domestic savings behaviour in developing countries may yield a more reliable assessment of their capacity for ultimately self-sustained growth.

30

The economics ofaid

Over time as income grows, the incomes not likely to be available for investment are composed of three main elements: (a) the increase in income required to support the increase in population; (b) the increase in income which must be earmarked for consumption to provide incentives for growth or to meet rising expectations; (c) the increase in domestic production lost through deterioration in the terms of trade. Looked at in this way, the prospects for the aided countries do not appear as optimistic as the Table 6 data on marginal savings suggest. Thus, for the aided countries in Section A of Table 6, the average saving rate in 1962 was typically 0· 10 and required investment rate typically 0· 16. If their population growth is 0·03 and the 'incentive' increase in consumption is 0·02 per annum (no allowance for terms of trade deterioration), then with a G.N.P. growth rate of 0·05 the marginal savings rate would be no higher than the 1962 average rate and well below the required investment rate. Of course, population growth is not always as high as this and it is possible for governments to hold down increases in consumption as incomes rise (for short periods at least). Time periodfor self-sustained growth

One of the strategies for allocation of aid among developing countries is that it should be concentrated on those countries likely to achieve independence of external assistance within a given period of time. 9 How long will it take a country to achieve self-sustained growth and how much assistance will be required? Clearly this problem can be answered in terms of the savings constraint model outlined above, provided empirical values can be given to the main parameters. Given a target rate of growth (f) the value of the output-capital ratio and the initial domestic savings rate, the time required for an economy to achieve self-sustained growth at rate f will depend (in the savings constraint model) on the marginal savings rate (s) . This is intuitively obvious and the following figures for l (years to independence of e~.temal assistance) indicate its sensitivity to

s.

s

{ (years)

0·20

a

0·22

40

0·25

27

0·30

18

0-40

14

This calculation makes the assumptions that 1 the growth rate is 0·06, output-capital ratio is 0·3 and the initial domestic savings rate is zero. 2 the magnitude of the marginal savings rate will remain constant for the whole period l. 3 that capital transfers are entirely in grant form so there is no debt service which would otherwise prolong the period of time required to become completely independent of foreign assistance.

Aid and economic growth I

31

It is noticeable that for marginal savings rates within empirically plausible limits (0· 10 to 0·30), the time period is extremely sensitive to the actual savings rate.

Export constraint Growth at the desired rate may be constrained by lack of export capability rather than domestic savings. In this case what condition is required to ensure that aid flows to supplement export earnings initially, will eventually fall to zero? Suppose only capital goods are imported but exports initially are inadequate to finance the required level of imported capital goods to achieve the desired growth rate. By a similar analysis to that in the previous section it can be shown that growth will eventually be sustained by the country's own export earnings if the marginal export propensity x > -f,. 10 It is not easy to test this condition empirically because most poor countries import current inputs as well as capital goods and the value of B is difficult to ascertain. The greater the extent to which X exceeds the ratio, the faster will self-sustained growth be achieved and the smaller the total aid transfer required.

-i

Skill constraints The preceding analysis has assumed that there are no domestic institutional limits on the achievement of the desired growth rate other than savings and export potentiality. In practice, however, growth depends heavily on the availability of skilled workers, managers, technical personnel and civil servants. The lack of these skills can severely limit the amount of productive investment which can be planned, organized and executed and sets what is commonly called the 'absorptive capacity' of an economy. One treatment of skill shortages would be to postulate various skill/output coefficients where skills are seen as inputs in a production function in the same way as domestic and imported capital goods. Alternatively, skill availability can be viewed as a function of the development of an economy, independent of its capacity to save or import. In this case it is possible to postulate a limit on the 'ability to invest' productively. In the early phases of development many economies may not have the managerial, organizational and technical skills and experience to achieve the level of investment which their domestic savings (plus foreign capital) would otherwise permit. This constraint on the ability to invest may limit an economy's absoiptive capacity for foreign capital and its growth rate. The absoiptive capacity can however be raised and the extent to which this can be done is partly a function of savings and external resources and partly a function of time. Skills can be augmented by investing in the education and training of people and this requires savings. Skills can also be increased by importing foreign personnel to use directly or indirectly in training local people. Foreign assistance can finance both and hence can relieve skill bottlenecks. However, the human capabilities of an economy are very much a function of time, because the ability to carry out an investment progranune effectively is largely a matter of 'learning by doing'. It is the passage of time and the speed of the development process which is likely to set the important and overriding limit on the absorptive capacity of an economy for capital.

32

The economics ofaid

It has therefore been suggested that the ability to invest effectively may be most plausibly represented by the maximum rate of growth of investment over time where investment is measured in absolute terms. This view can be set out formally as follows where q represents the maximum growth rate of investment over time set by skills and experience. 11 It is assumed for simplicity to be a constant.

(i) Ability to invest 1,=lp +q)' (ii) Savings constraint S,

(10)

= S0 + s(Yt -

(11)

Y0)

(iii) Incremental output/capital ratio 0 =dY I

Foreign assistance F,=1,-Sf Hence solving for the level of foreign assistance which can be used for investment gives Fi

=(

1- i)(Yt -

Yo)

+F

0,

where F0

=/

0 -

S0

(12)

By comparing this equation with equations (6) and (9) above it is clear that if q is less than o(s+j) or B(x+j) then there is a 'skill determined' limit on growth which overrides the other constraints. This lowers the rate of growth and puts a limit on the inflow of foreign capital (f) which can be effectively used. Some economies may well be in this early phase of skill-limited development. In this situation the increased flow of foreign assistance (F,-FJ finances the difference between the increment of investment permitted by skills and the increment of domestic savings. At the end of this skill dominated phase of development the rate of increase of investment will drop to f and foreign aid flows to achieve f will be determined by the savings or import constraints on growth.

Conclusion The main conclusions of this chapter are that, apart from skills, there are two separable constraints on growth-savings and foreign exchange. Hence an increment of foreign assistance may have a different effect on growth from an increment of domestic savings. These conclusions follow logically from the assumptions of the model used. The crucial assumptions are (a) that exports are a fixed proportion of domestic income and (b) that imported inputs are essential for domestic production and the aggregate production function has fixed coefficients. The plausibility of these conclusions therefore depend on whether these are plausible technological and economic assumptions to make in the context of most developing countries. Whether such extreme lack of substitutability between domestic and foreign resources exists is an empirical question and the next chapter is concerned with attempting an answer to this question.

Aid and economic growth I

33

Notes

2 3 4

5 6

It will not be necessary at this stage to specify an adequate growth rate for a poor country, but this will need to be done when calculating foreign capital requirements for development in the following chapter. An adequate growth rate might be defined as one which would permit the minimum growth in output per capita consistent with political and social stability, and it would depend on the growth of population and how rapidly aspirations are growing. Effects of skill limitations will be considered later. An evaluation of the second view is provided in Ch. 4 below. This is sometimes termed the 'structuralist' view. The rest of the analysis in this chapter accepts that there are two separable constraints on growth. In this case, a unit of investment will be expended on Km and Kd in the proportions 8 : a: and 8 This follows because a KJ=B Km and (Kd +K,) = 1. The extra output given by 8 + "' respectively. . of.mvestment 1s . th en= 8 B+a: a: one umt The assumptions of fixed production coefficients and fixed export rates are crucial in the above analysis and their economic plausibility is questioned in the next chapter. It has been shown that o

=

:. B

rx

rxB

+

B

:. o(rx

+ B) =

rxB :. rxo

= rxB - oB

0

= 1 - o/rx

But o/a > 0 < 1 (a> 0, B > 0) so B > o. 'Relief' food aid is required to prevent an unacceptable decline in food consumption per head when domestic food output is disrupted. 8 This is an arbitrary value but one which would ensure that per capita income would double in 50 years with a population growth rate of0-03. 9 See Ch. 2. 10 Foreign assistance (F) for growth rate (r) = M, - x, where M 1=imported capital goods X,=exports M, = and X, = X0 + x(Y, - Y0) where x=marginal propensity to export. Hence F, = ( .!.._ x) y, + xY, - x,. F, will decline if .i > 11 In economic -theory, the appropriate level of investment is usually assessed in relation to a schedule of the marginal efficiency of inveshnent or the marginal rate of return. To specify a rigid limit to the scale of feasible investment, as is done here, implies that on an investment schedule the rate of return would drop, at some point, sharply to zero.

7

i,

fi.

4 Aid and economic growth II

The model outlined in the last chapter can clearly form the basis of a methodology for empirical estimation of the level of foreign assistance required to achieve growth targets in developing economies, individually or as a whole. Ignoring skill constraints this model implies that two resource gaps-a 'savings' gap and a 'trade' gap-need to be estimated. It further implies that these two gaps need not be identical and that foreign capital must be adequate to fill the larger gap if the planned growth rate is to materialize. Two-gap computable models of this kind have been used in practice to estimate aid requirements of developing countries 1 (see Table 7). It is the economic theory underlying this methodology of foreign resource needs which must now be clarified and examined.

Ex-ante and ex-post concepts The first point to clarify is the distinction between ex-ante and ex-post magnitudes of the resource gaps. In an ex-post or accounting sense the two resource gaps are identical; it is only when the savings and foreign exchange gaps are viewed in an ex-ante or planned sense that they may differ. The ex-post or accounting identity of two resource gaps is an elementary formal proposition and can be demonstrated as follows: Goods and services which are put to use in an economy come from two sources, home production (Y) and imports of goods and services (M). These are used for consumption (C), investment (I) and exports of goods and services (X) where C and I include imported consumer goods and investment goods. Hence, (i)

Y+M=C+l+X or Y=C+l+X-M

Domestic production (Y) gives rise to incomes which are identical to the total value of production2 and this income is spent on consumer goods (C) or saved (S) so that (ii)

Y=C+S From (i) and (ii) it can be seen that C is common to both sides of the equation and the equation can be rearranged as

(iii) J=S+M-X This indicates that M-X is the excess of imports of goods and services over exports, or the deficit (or surplus) on the current account of a country's balance of payments. This must be identical with the net capital inflow (F), defined to include any net change in the external reserves of the economy. Thus DOI: 10.4324/9780203840177-4

Aid and economic growth II

35

(iv) M-X=F When (iii) is rearranged we have 1-S=M-X=F

It is clear that the net inflow of foreign capital plays a dual role but in the accounting sense the savings gap is identical with the trade gap.

Long-run ex-ante resource gaps However, this necessary identity arises after the event (ex post) when it is possible to look back retrospectively at what happened. Ex ante, looking forward, the resource gaps may differ because, in the long run, those who make the decisions about saving, investing, importing and exporting are not always the same people and they are not all affected by the same factors. The situation can be illustrated by Figure 3 which relates the two ex-ante gaps or required levels of net foreign capital inflow (F) to different target growth rates of income . The X-M schedule represents the ex-ante export-import gap in any particular period 't' for various growth rates. The 1-S schedule, similarly, represents the ex-ante savings gap in period 't'. The particular ex-ante schedules of Figure 3 are based on the following five equations which correspond closely to the type of model outlined in the last chapter.

Yt = Y0(1 Ii =

f

b Yt

+ i)'

St = S0 + s( Yt - Y0 ) Mt = M0 + ni(Yt - Y0) Xt = Xo(I + x)' where mis the minimum marginal propensity to import and xis the maximum anticipated rate of growth of exports which is assumed to be exogenously determined. 3 The other variables have already been defined in the last chapter. The ex-ante gap is measured either by l 1_ S1 or~- X, and these are both a function of the growth rate (f), given the initial values of income, saving, exports and imports and the exogenously determined rate of growth of exports (x). It must be stressed at this stage that the assumed functional relationships are somewhat unorthodox. Both investment and saving are uniquely determined by the level of income and no other factors. Imports (M) are assumed to be the minimum consistent with a given income level so that mrepresents the marginal 'necessity' to import. While, for exports, x represents the maximum possible rate of growth over time. It follows that there are two unique and independent schedules relating the 1-S gap and the M-X gap to various growth rates. These functions are unorthodox in not including a price variable explicitly. 4 This implies that either prices are assumed to be inflexible or that, with price flexibility, demand and supply conditions for exports and imports (for example) are such that no price changes would alter export earnings or import payments at a given level of income. These implicit

36

The economics ofaid

and somewhat restrictive assumptions are reconsidered later but if accepted it is quite possible to have two independent schedules as in Figure 3. At Q in Figure 3 both ex-ante gaps are equal (0F J. If the target growth rate happened to be r 0, then this rate of growth would be achieved if net foreign capital inflow of OF0 were actually forthcoming. 5 However, suppose the target growth rate is r 1 then the trade gap is larger than the savings gap, ex ante. If foreign capital is not adequate to fill the larger gap (1-S)t (M-X)t "O

.Q cij

a. .!:

c,

'O

0

r1 ro Target growth rate of GNP

r2

Figure 3 Ex-ante 'saving' and 'trade' gaps and growth rates (OF j this growth rate will not be achieved. Short-run forces will bring about ex-post equality of the two gaps but at the expense of frustrating the growth target. Similarly if the target rate is r 2, capital inflow must meet the larger savings gap (0F.). On the basis of this basic methodology several estimates have been made of ex-ante aid requirements of individual developing countries and developing countries as a whole. Table 7 sets out some of these, distinguishing those based respectively on 'savings' and 'trade' gaps. The Appendix discusses statistical problems in estimation of these gaps. Evaluation ofthe dual gap hypothesis One crucial assumption of the 'dual constraint' or 'dual gap' model must now be examined. The model set out in the last chapter shows that an increment of foreign assistance can have a different effect on growth from an increment of domestic savings. This follows from the assumption that there is inadequate scope for substitution between domestic and foreign resources. More specifically, two main rigidities are built into the models, viz.,

a that there is a limit on the scope for expanding export earnings because of inelasticities of either foreign demand or domestic supply and

Aid and economic growth II

37

b that imported inputs are essential for production and the aggregate production function assumes fixed input coefficients. With these rigidities the ex-ante M-X gap may diverge from the ex-ante 1-S gap. Export earnings are inexpandable and imports incompressible beyond a certain level. The 'trade' Table 7 Estimates ofaggregate savings and trade gaps ofdeveloping countries

A Savings gap (billion $) Source

Period of projection

U.N. 1 (1949)

Rosenstein-Rodan (1961)

UNCTAD (1968)

Required annual capital inflow

5·8

8·5

2·0 2·0

3·0

6-5

1960-9

2·0

3·0

7·0

1962-6

l ·8

2·8

6-4

1967-71

2·2

2·8

6-4

1971-6

2·5

2·8

5·0

Millikan/Rostow (1953) 4

Capitaloutput coefficient

1950-60 2

Hoffmann3 (1959)

Growth rate ofG.N.P per capita(%)

1975 (L)

3·6

1975 (H)

7·6

B Trade gap (billion $)

Source

Year of projection

Imports

Exports

Trade gap

Debt service gap

Foreign exchange gap

1970

41·0

29·0

12·0

8·0

20·0

1970

41·0

31 ·0

10·0

8·0

18·0

B.Belassa

1970

38·0

33·0

5·0

5·5

10·5

1975

49·0

42·0

7·0

6-9

UNCTAD 8

1975 (L)

12·5

1975 (H)

17·6

U.N.s G.Blau

6 7

13-7

U.N. Measures for Economic Development 1949. 2 M.Millikan & W.Rostow, A Proposal: Key to Effective Foreign Policy, 1956. 3 PG.Hoffmann, One hundred Countries, 1960. 4 P.RosensteinRodan, Review of Economics & Statistics, May 1961. 5 U.N. World Economic Survey 1962. 6 G.Blau, Commodity Export Earnings & Economic Growth, 1963. 7 B .Belassa, Problem ofGrowth in less developed Economies (O.E.C.D.), 1963. 8 UNCTAD, Trade Prospects & Capital Needs of Developing Countries, 1968. These estimates at 1960 prices refer to the sum of the gaps in Asian and Latin American countries. A high and a low estimate is given. 1

gap may then exceed the 'savings' gap and unless foreign assistance fills the larger gap the target growth rate will not be achieved. Why should these rigidities be assumed to reflect the situation in developing countries?

38

The economics ofaid

The view of some economists is that there are 'structural' factors which create these rigidities in developing countries. Policy measures such as changes in domestic prices or exchange rates in these countries will not be effective in altering export earnings or import payments at a given level of domestic incomes or outputs. Thus a unity elasticity of foreign demand for exports or the lack of any effective domestic substitute for certain imported goods used in the production process would constitute structural rigidities. The alternative and more traditional view is that there are unlikely to be such rigidities if prices are flexible. If resources are optimally allocated there can only be one constraint (savings) on growth and one ex-ante resource gap (savings). If appropriate price policies or exchange rates policies are followed, resources would be switched to eliminate the difference between the growth effect of imports and domestic saving and hence the difference in the ex-ante gaps. On this view, if a dominating ex-ante trade gap exists it indicates current or past misallocation of resources and must be the result of inappropriate price policies. To illustrate the difference between these two views consider an economy where all capital goods are imported (and only capital goods are imported) and where all exports are consumer goods. With these assumptions, to save means to export and to invest means to import. Consider exports first and postulate a foreign demand curve, a domestic demand curve and a domestic supply curve for a typical export (consumer good) as in Figure 4. First, assume that the foreign demand curve is perfectly elastic (F1F). In this case, an increase in saving will shift the domestic demand curve (DP) leftward to Dp 2 • The quantity of consumer goods consumed at home falls by the same amount (ab) as exports increase and hence permits an equivalent import of capital goods. 6 The trade constraint and the saving constraint on capital formation are identical. However, suppose the foreign demand curve for exports to have unity elasticity or to be kinked such that it is inelastic for a fall in price and elastic for a rise in price. In this case it is clear that an increase in saving cannot increase the value of exports. This type of rigidity is 'structural' in the sense that a domestic policy change in the economy, such as devaluation, cannot break the bottleneck on foreign exchange. In this simple economy where only consumer goods are exported and only capital goods imported, this outcome seems to require that foreign demand curves be of this kind for all goods produced by the economy. Are these plausible conditions to assume for developing countries? Taking developing countries as a whole, and certain primary products, international demand curves may approximate to unity elasticity. For individual countries which have a monopolistic position in the international market (e.g. Pakistan jute) demand conditions may be of this kind. Also, where international agreements and market sharing arrangements exist, a kinked demand curve may exist for the individual supplier. However, it is implausible to assume that these demand conditions face all primary products and all individual country suppliers. In the case of manufactures (vis-a-vis primary products) the aggregate and individual demand curves for developing countries are likely to be highly elastic with respect to price except where these manufactures face quantitative restrictions on their entry into the markets of developed countries. 7 There may, however, be a supply problem with exports from a developing country even if demand is elastic. Domestic market protection of industries may have created a cost

Aid and economic growth II

39

structure which makes them internationally uncompetitive. Thus in Figure 4 the supply curve of exports may be S 1S2 where domestic suppliers are unwilling to export at all whatever the shift in domestic demand (increase in savings). This, however, may reflect an overvaluation of the domestic currency and a devaluation would normally raise the foreign curve (in domestic currency) to say F 2F 2 in Figure 4 and induce exports of de when savings increase. Figure 4 Demand and supply ofexports

£

f:z ------

d

e • ....-

- - •. -·· ---- - ---------- G

~2, .-·

i:;r'---------"F-----4~----,,,,."------

Quantity In this case there would appear to be no structural rigidity if the appropriate price policies were pursued. Moreover, if primary products face a unity elasticity of demand and manufactures face an elastic demand, then differential effective exchange rate changes would permit a net expansion of export earnings. What about imports? A structural problem would mean that it is impossible, by changing relative prices, to reduce imports without lowering the level of domestic output. In the simple economy where all capital goods are imported and only consumer goods produced domestically, a structural problem implies that (a) in production no substitution between imported capital goods and domestic factors is possible and (b) in consumption no substitution between different consumer goods is possible. 8 Although for certain products and processes rigidities of this kind exist, they are unlikely to hold generally. The simplifications of this approach and its static nature must be borne in mind but it does suggest that the existence of a trade constraint which overrides the saving constraint on growth requires rather restrictive assumptions which are not likely to hold generally although they may apply to particular countries at a particular time. 9

40

The economics ofaid

Policy implications The long-run and short-run policy implications of these alternative views are considerable. If these structural rigidities do not prevail generally then an ex-ante dominance of the trade gap over the saving gap would not be expected. If empirical calculations suggest that this pattern is likely to emerge generally (see Table 7) then an important long-run variable, relative prices, has probably been left out of the formulation. Indeed, the existence of a dominant ex-ante trade gap may well reflect past and present misallocation of resources due to faulty pricing policies. Instead of using foreign assistance to fill the difference between the 'saving' gap and the larger 'trade' gap, a change in prices would perform this function. The establishment of an equilibrium exchange rate would alter relative home and foreign prices and help to reallocate resources in the economy from serving domestic to serving foreign demand and from import-using to import-saving activities. Foreign assistance would still be required to meet a savings deficiency at the target growth rate but it would not bolster inappro-priate price policies and perpetuate misallocation of resources in the assisted economy. On the other hand, if an economy genuinely faces 'structural' rigidities, no price or exchange rate policies will help to close the ex-ante gaps and foreign assistance flows should fill any larger 'trade' gap if the target growth is to be achieved. Foreign assistance may then be viewed as helping to weaken rigidities a by financing the domestic development of a more diversified industrial structure capable of producing domestic substitutes for imports, and goods which have an elastic foreign demand; and b by tiding over the period while developed countries alter their restrictive trade policies towards many developing country exports. Let us now assume an economy with a genuine problem of inelastic exports and imports and an ex-ante gap inequality. How does the short-run adjustment from ex-ante inequality to ex-post equality come about? In the first case where the 'hard core' ex-ante M-X gap exceeds the ex-ante 1-S gap at the desired growth rate (r) and net foreign capital actually forthcoming fills the larger gap, there will be a deficiency of effective demand. This can be illustrated as follows: Suppose an economy is projected to have a net national income (Y) = I 000. Also that projected investment (1)=200, savings (S'.)=100, exports (X)=lO0, imports (M)=250. ltisfurtherassumedthatforeign resources equivalent to the larger (M-X) gap are forthcoming (i.e. 150). Then aggregate demand for goods and services =J+C=200+900= 1,100 while supplies available for domestic use =Y+M-X=l000+250-100=1,150. A deficiency of effective demand of 50 exists. If the general price level is inflexible downwards, market forces would cause unplanned stocks accumulation, reduced utilization of capacity and reduced impetus to invest. This would mean a shortfall in the level of income and the target investment and growth rates. Even if prices are flexible there could be some loss of confidence by entrepreneurs with similar effects on investment and growth. The government will therefore have to intervene to stimulate domestic consumption (reduce domestic savings) by (say) reducing taxation to ensure that ex-post equality ofthe two gaps is realized without a change in desired investment and growth

Aid and economic growth II 41 rate. It would be inappropriate to stimulate investment since this would conflict with the predetermined growth target and also cause an unplanned increase in import requirements. In the second case where the foreign resources forthcomings are equal to a dominant ex-ante 1-S gap at the target growth rate, there will be an excess of aggregate demand over aggregate supply at current prices. The appropriate government policy would be to stimulate the demand for imported consumer goods, or divert export goods which can be consumed at home, from foreign markets to the home market This might be effected by changes in the exchange rate or in import duties and export taxes. To permit greater imports of capital goods might cause increased investment and hence widen the 1-S gap and foreign resources forthcoming would then be inadequate. In the absence of these policy measures there will be an accumulation of foreign exchange reserves and the domestic excess demand will cause unplanned stocks reduction or inflation. Even if fixed investment is unaffected, there will be an eventual replenishment of depleted stocks from imports (running down the increased foreign exchange reserves) or rising domestic prices will reduce exports, due to increased competition from domestic buyers and reduced foreign demand. Hence, one way or another, the M-X gap would adjust to the 1-S gap ex post. In the first case, however, it is important for the government to take the appropriate policy measures otherwise the adjustment may frustrate the target growth rate. In the second case, an absence of government intervention seems unlikely to affect target growth. The preceding two chapters have considered the theoretical nature of the 'two gap' analysis of aid but it does not consider the usefulness of this approach to aid policy in practice. A sceptical view of the possibilities of this approach is provided in the last chapter of this book.

Notes The most recent computations are those of UNCTAD, Trade Prospects and Capital Needs of Developing Countries, New York, 1968. A similar methodology has been used by A.Maizels et al., Exports and Economic Growth ofDeveloping Countries, N.I.E.S.R., 1968, to predict growth rates on certain assumptions about export growth of developing countries and plausible foreign capital inflows. 2 Net factor receipts and payments for the economy are assumed to be zero. 3 Note that in this model in Ch. 3, x represented the (constant) proportion of exports to income, whereas here x represents the growth in exports over time. The difference is that exports are exogenously determined and not set by growth of income. 4 Thus investment and saving are not a function of the rate of interest and exports and imports are not a function of the price of domestic relative to foreign traded goods and services. 5 It should be noted that even equality of the two gaps ex ante is not a sufficient condition of equilibrium (i.e. of achieving the specified growth targets) as in the Keynesian system. It also requires that actual foreign resources forthcoming are equal to the ex-ante gap. 6 The price of capital goods is assumed unchanged. 7 The most important manufactured consumer good exported from developing countries is cotton textiles and the global and country-wise quotas imposed on imports into developed economies, at present, give the foreign demand curve an effective kink which must frustrate any attempt to raise export earnings from price adjustments by the suppliers.

42 8 9

The economics ofaid These are the basic Leontieff assumptions of fixed coefficients in production and consumption as opposed to the neoclassical assumptions of smooth continuous isoquants and preference/indifference curves. India has been suggested as one country where a trade constraint or gap may well dominate the savings constraint or gap.

5 Debt and the terms of aid

Introduction So far the analysis has been concerned with assessing the effect of net capital inflows on growth, but no attention has been given to the implications of indebtedness arising from these capital flows and the servicing of the debt by the borrowing country. What effect will the need to service debt have on growth of the borrower in the short run and long run? The answer depends, of course, on the attitudes or policies of the lending countries. Whether debt will present an obstacle to desired growth rates of developing countries will depend on the scale and the terms on which lenders are prepared to transfer resources to these countries. This clearly does not involve fundamental issues in growth theory. Nevertheless from a policy viewpoint it is valuable to analyse possible behaviour patterns of debt on different scales and terms of capital inflow. The debt-service issue will be analysed in the context of long-run growth and also in the context of short-run management. It will be assumed, that there is one lender and one borrower. If the lender is only willing to lend for a limited period on fixed terms (of interest and repayment) then the long-run target growth rate in the borrowing country will have to be tailored to allow a sufficient margin of resources for servicing debt. If on the other hand, the lender is prepared to lend at interest in unlimited amounts for indefinite periods in order to permit the borrower to achieve a specified growth objective, there would appear to be no debt problem. Yet even in this case it can be shown that formally an 'unmanageable' debt problem can arise and cause a breakdown. Thus, if the interest rate on loans exceeds a certain 'critical' level, debt (and hence interest payments) may explode in relation to the gross national product of the borrower. The policy implication is that where growth conditions in the borrower seem unfavourable, interest rates must be kept below this critical level by the lender to ensure that no breakdown in the borrower's ability to meet interest charges occurs. The debt problem is therefore essentially a problem of the terms on which assistance is provided. For the economic analyst the main interest lies in establishing the consistency of growth targets set by the borrower with the terms of capital assistance set by the lender. If lenders commit themselves to certain growth objectives in the borrowing country, the terms of assistance required to achieve these targets can be calculated formally using the type of simple macro-economic model employed in earlier chapters. In addition to the long-run implications of debt servicing there is also the short-term problem. Against a background of fluctuating export earnings the borrowing economy is faced with rigid debt-service payments. This may present acute problems of short-run management especially if interest and amortization constitute a large proportion of export earnings. DOI: 10.4324/9780203840177-5

44

The economics ofaid

These long-run and short-run issues will be discussed more fully in the following sections.

Debt and long-run growth The economic conditions which must be fulfilled if the borrower is to service the debt arising can be stated formally in terms of total resource availabilities and uses and foreign exchange availabilities and uses. These conditions are that: 1 output produced plus gross foreign capital inflow must exceed domestic consumption and investment by the amount of debt-service payments, or domestic savings plus capital inflow must exceed domestic investment by the amount of debt-service payments. And 2 foreign exchange receipts plus gross foreign capital inflow 1 must exceed payments for imports by the amount of the debt-service payments. An economy must adjust claims on total resource, savings, and foreign exchange in any given year and over time so as to release the resources required to meet debt-service obligations. Formally there is nothing in the price or income adjustment mechanism which makes it impossible for this adjustment to be made if the rules of the game are kept. However, these adjustments may conflict with other central objectives of the debtor country; they may involve reduced employment and activity in the short run and lower investment levels and hence lower growth rates (than desired) in the long run. Since the debt-servicing capacity of an economy is clearly related to its growth rate, it is most conveniently analysed in the context of a simple macro-economic model of the savings-investment kind employed earlier. In this model the main parameters are average (s) and marginal (s) savings rates, output capital ratio (15) and the target growth rate of G.D.P. (f). The required investment rate is then For simplicity, though not realism, it is assumed that these parameters remain constant for the foreseeable future. To build in the debt implications it is also necessary to assume a given average interest rate on foreign assistance (i). Now a borrowing economy can hypothetically pass through three stages in its debt profile and this is illustrated in Figure 5. In Stage one domestic savings are inadequate to finance required investment and the economy has to borrow not only to finance part of the investment but also to pay interest on debt accumulated and to cover amortization. During this phase the burden of servicing foreign capital is continuously postponed and the debt compounds rapidly. Stage two begins when domestic savings are adequate to finance required investment. The economy's target growth can now be potentially self-sustained. If the borrower decides not to default on debt-service payments it still has to borrow to pay these. 2 This stage will only begin (in terms of this model) if the marginal savings rate (s) is at least equal to the required investment rate, since in the long run the average domestic savings rate will tend towards the marginal rate. During this stage no net inflow of capital is required but new borrowing continues to finance interest and repayment of borrowed capital. As average domestic savings continue to grow an increasing part of interest on outstanding debt can be met from domestic savings. Total debt grows at a diminishing rate



Debt and the terms ofaid 45 till it reaches a peak when new external borrowing is only required to cover amortization. Savings are just adequate to finance required investment and interest on debt. Figure 5 Hypothetical debt profile £

Time

Stage three begins when domestic savings have grown adequately to cover not only required investment and interest on external debt but also yield a smplus for amortizing debt which now declines rapidly (compound interest in reverse). This stylized debt cycle indicates how an economy can be assisted to achieve and maintain a self-sustained growth rate and still repay capital with interest. It is clear that the rate of growth which can be supported will be lower: a the lower the marginal savings ratio (s) b the lower the output/capital ratio (15) c the higher the interest rate on loans (i) d the shorter the period which the borrower is allowed by the lender to become independent of capital assistance or free of debt. However, there is nothing automatic or inevitable about this growth process and debt cycle. The growth process is not as mechanical and predictable as this simple model assumes. There are some economies which, on present evidence, seem unlikely to leave stage one. These are the economies where low marginal savings rates suggest that domestic saving in the foreseeable future will not cover their investment requirement for sustaining a 'satisfactory' growth rate. These 'long haul' countries will tend to be those with large population relative to natural resources, low per capita income and high population growth

46

The economics ofaid

rates. Such economies could remain borrowers indefinitely and could only sustain their target growth rates if the lender is prepared to lend indefinitely. There are other economies which are likely to become just self-sustaining in growth, (i.e. enter stage two where domestic savings equal required investment). Interest payments will then increase annually at a rate equal to the rate of interest. Once again the country would have to borrow indefinitely whatever the interest rate (if it is other than zero). For economies which become more than just self-sustaining, the behaviour of debt and the borrowing horizon will depend on the level of interest rates charged.

Indefinite borrowing and the critical interest rate The maintenance of desired growth rates in certain economies will depend, therefore, on the willingness of the lender to lend indefinitely. There is no a priori answer to whether these conditions will be met. However, if unlimited borrowing is possible, it is important to note that 'unmanageable' debt can still arise for the borrower if debt (and hence interest payments) rise faster than its national product. Unlimited borrowing and debt rising for ever is manageable provided it does not 'explode' relative to national income. There is a 'critical' rate of interest on borrowed funds above which this explosion will occur; and this can be demonstrated as follows: In the long run the average saving rate (s) will approach more and more closely to the marginal savings rate {s). Thus the surplus of domestic savings over required investment, needed to service debt, will approach the proportion of G.N.P. equal to the difference between marginal savings and required investment. At one extreme, when the surplus just covers interest obligations debt is constant. At the other extreme when there is no surplus, debt grows at a rate equal to the rate of interest (i). If this rate of interest exceeds the growth of G.N.P. then debt ultimately becomes unmanageable. In between these two long-run extremes lies the critical case where debt would grow at the same rate as national product. It is possible to identify a 'critical interest rate' which is a function of initial and marginal savings rates, the output-capital ratio and desired growth rate. 3 The policy implication is clear for the lender. In certain cases it may not only be necessary to lend indefinitely to sustain desired growth rates but it may also be necessary to adjust the interest rate on loans to ensure that debt does not explode relative to the G.N.P. of the borrower. Whether the concept of the critical interest rate is significant is more debatable. If a country is unlikely to become self-sustaining in growth and must borrow indefinitely, it seems pointless to charge interest at all. Charging a positive interest rate only goes to swell an unrepayable debt indefinitely. 4 If, on the other hand, a country can become more than self-sustaining (so that the critical interest is higher than the rate of growth) then it is hardly vital to calculate the critical interest rate since in this situation the borrower can default on his debt-service obligation without jeopardizing his desired growth rate. In any case, these calculations are excessively formal and, in practice, it is not interest rates so much as the willingness of lenders to continue providing resources to the borrowers which is likely to be the obstacle to sustaining necessary growth rates in the long run. There are also short-term management problems arising for the borrower in meeting debtservice obligations.

Debt and the terms ofaid 47 Short-run debt servicing Interest on debt is the most rigid item in a country's balance of payments and must normally be paid in convertible currency. Export earnings, the main source of convertible currency, however, are often subject to considerable fluctuations in borrowing countries. 5 As interest payments rise to a large proportion of a borrower's export earnings, the rigidity becomes of greater significance. In principle, there is no problem if the lenders provide adequate new loans to cover necessary imports as well as the debt-service obligations. In practice, however, liquidity problems can arise if the new loans are inconvertible (e.g. tied to the purchase of goods in the lending country). Capital inflows of this kind are not perfectly substitutable for debtservice payments made in convertible currency. In the later stage of the debt cycle when the debtor country has become a net repayer of capital there is still no formal problem. When exports fall a deflation of domestic income will reduce imports until a sufficient margin of foreign exchange is released to pay the fixed debt-service charges. In practice, however, the structure of the economy may make it so difficult to substitute domestically for many imports (especially fuel, industrial maintenance goods and certain capital goods) that drastic reductions in the utilization of existing capacity and the level or the productivity of the domestic investment programme may be necessary to ensure that foreign exchange uses and availabilities are balanced. An escape from this problem may be found in the use of foreign exchange reserves and stabilization policies by the exporting country but again experience suggests that many difficulties can arise. The management of many developing countries' balance of payments may be considerably complicated therefore by the need to service debt. Where management is inefficient, long-run growth can be reduced.

Notes Including any net call on the borrower's foreign exchange reserves. Clearly this is the point at which a single borrower can default with no cost and much saving to itself. From a purely economic viewpoint the main discouragement to it defaulting lies in the possibility that its own growth parameters will change unfavourably in future or that its behaviour will dry up the supply of capital to other developing countries who have not yet reached the stage of potentially self-sustained growth. 3 The concept of the critical interest rate is due to JP.Hayes. The critical interest rate (i) is given by the following formula when initial period debt is zero and S0 represents the savings rate in the initial period.

2

I = r(So - s) r So -

8

It is noticeable that even if the marginal savings rate (s) is less than the desired investment rate a country can borrow indefinitely without interest payments continuously increasing faster than domestic product, provided interest rates are low enough (obviously if i = 0).

48 4

5

The economics ofaid However lending governments may in practice be under two forms of pressure which require them to lend indefinitely but at positive interest rates. On the one hand, there is the moral and political pressure to assist very poor countries indefinitely. On the other hand there is the influence of an ethical system which insists on the obligation of the borrower to repay with interest. This is further supported by the contention that unless interest is charged the borrower will use the funds inefficiently. (There is no a priori or (yet) empirical justification for this view however.) Unstable export earnings are said to arise in primary producing countries because they are often dependent on anarrowrange ofexport commodities whose markets are affected by minorrecessions in industrial countries and whose supply is often vulnerable to bad harvests, domestic political instability, etc. The rigidity of debt payments has also increased in recent years as developing countries have come to rely more heavily on borrowing from other governments. Previously most of the capital inflow was in the form of private equity invested in export industries. This ensured that profit and dividend remittances fluctuated with fluctuations in export earnings.

6 Aid tying: trade and resource allocation effects

What is tying? External financial assistance to developing countries is provided in a wide variety of forms. This is largely a matter of the degree of restriction placed by the provider of the finance on the use to which these resources should be put and the sources from which they should be obtained. At one extreme is completely untied finance which the recipient can use for any purpose and which can be used to procure goods and services from any international source. At the other extremes would be finance tied to procurement of goods in the assisting country and tied to use on a specific project (or for specific goods) in the recipient country. There are a variety of intermediate forms. In practice the forms of aid will vary greatly for different countries but there has been a general trend towards increased 'sourcetying' of aid since 1960. There are two main economic reasons why aid is tied by source and use: First, aiding countries are concerned with the effect which aid has on their exports. This concern with the trade effects of aid arises because an important 'lobby' supporting development assistance in developed countries are certain domestic industries which hope to profit from export contracts under this assistance. Further, the governments of aiding countries are often concerned about the cost of aid to their own balance of payments. They wish to minimize the charge which their aid programmes make on their foreign exchange reserves especially if their reserves are inadequate. 1 Second, the 'aiding country wishes to' exert influence over the allocation of resources in the recipient. Whether tying makes this feasible has been the subject of some debate among economists. Suppose, for example, that aid is tied to project A which is examined and considered 'high priority' by the aiding country. If this project would, in any case, have been part of the recipient country's investment programme and would otherwise have been undertaken with the country's own resources, then the project actually financed would be quite another one from that to which the aid is ostensibly tied. Tying can be effective in this respect if the aiding and the aided countries differ in their assessment of investment priorities, but this raises the important question ofwhich country is the best judge of development priorities. 2

The rest of this chapter is devoted to a simple analysis of the effects of aid tying on (a) Trade Patterns and (b) Resource Allocation.

Aid tying and trade patterns To examine the effect of aid tying on trade patterns a simple arithmetical example is used with one recipient country importing from three other countries, two of which provide aid. DOI: 10.4324/9780203840177-6

50

The economics ofaid

It is assumed that any aid provided will in aggregate generate an equivalent quantity of extra imports, i.e. that a typical developing country is assumed not to add to its reseIVes when it receives aid in any period. The analysis will consider the trade effects of a marginal addition to aid. a Untied aid

When aid is completely untied it can be used for any item from any source. Since Recipient can purchase aided imports from its most preferred sources, the import pattern on untied aid will be the 'free trade' pattern. In the following example, therefore, recipient's average propensity to import from different sources out of its own earned foreign exchange is taken to represent the 'free trade' pattern of its imports. These average propensities are assumed to operate when untied aid is spent. 3 In this situation it is highly unlikely that aid provided by each individual country will generate an equivalent amount of its own export to Recipient. This will only occur if recipient's free trade (f.t.) propensity to import from each aiding country is the same as each donor's share of a marginal increment of aid to recipient. Such an occurrence would be purely coincidental since the f.t. propensities to import from different countries will depend on its commodity pattern of imports and the comparative advantage of the various countries supplying these commodities, while the share of each country in aid to Recipient will depend on various factors which are not likely to be systematically related trade shares. 4 Table 8 provides a simple illustration of the untied aid/trade situation. Table 8 Recipient's imports with untied aid Imports without aid Country A

400

Untied aid 60

Aided imports 40

(40%)

B

500

40

50

550 (50%)

100

10

(10%) 1000

440 (40%)

(50%) C

Total imports

110 (10%)

100

100

1100

The provision of untied aid leaves the overall import pattern (by source) unchanged. However, aiding country A gains only 40 extra imports for 60 untied aid, while countries B and C gain more exports than their aid contributions.

b Unilateral tying If A has a balance of payments problem it will be under pressure to tie its aid to raise its export/aid ratio. Let us consider now the trade effects of unilateral tying by A, i.e., on the assumption that other donors do not respond by tying themselves. The trade effects of tying

Aid tying: trade and resource allocation effects

51

by A will depend on its form of tying. We can distinguish conceptually two cases of tying and their effectiveness in altering trade patterns.

Case 1 Tying to source only If country A ties its aid to procurement of commodities only from A but leaves the recipient free choice of commodities, such tying may be completely ineffective In generating extra exports. This will certainly be the case if the tied aid is used to purchase goods which would have been bought from A anyway, and the total value of the goods which would be bought from A without aid exceeds the amount of A's aid. A's aid is then said to be 'switched' and the trade/aid relationship between A and the recipient will be exactly the same as in the untied aid situation. In the Table 8 example A's source tied aid of 60 is 'switched' and releases Recipient's own free exchange resources for spending according to its free trade propensity to import from the three countries, A, B and C. A gains no more exports from 'source-tying' its aid; such tying is purely formal and has no trade diversionary effects. Case 2 Tying to source and use Suppose A not only confines expenditure from its aid to A's goods but also specifies the use of the aid. Two sub-cases can be distinguished where such double-tying will be effective in diverting trade to A. The first case is where aid is tied to a marginal project-a project which Recipient would not otherwise have undertaken. 5 Aid is then used for the purchase of imports from A which would not otherwise have been made from any source. In this way A alters the pattern of Recipient's trade in its favour and it will be seen from the illustrative example in Table 9 that A not only ensures that its own aid is 100% effective in generating extra exports but also it gains some exports at the expense of other donors (still untied) aid. The second case is where A's aid is innon-projectform, but tied to goods which the recipient would otherwise import from the other countries B and C. This type of tying is deliberately aggressive and aimed at diverting Recipient's imports from its most preferred sources. This type of tying is harmful to the recipient since it forces it to import from sources which do not necessarily have a comparative advantage in the commodities concerned. This reduces the value of aid provided; a problem that will be taken up in the next section. Table 9 Recipient's imports-A's aid tied to marginal project Country Imports without Aid Imports from tied aid aid A

400

60

Imports* from untied aid

60

16

(40%)

B

500

40

20

520 (47·3%)

100

4

(10%) 1000

476 (43·6%)

(50%) C

Total imports

104 (9·1 %)

100

* i.e. Total untied aid (40)xpropensity to import from each country.

1100

52

The economics ofaid

Our concern here is to show that trade is diverted from the most preferred sources by this type of tying. The situation is illustrated in Table 10. It will be seen that A's aid is used to buy goods which would otherwise have been bought from B and C, and hence releases Recipient's own foreign exchange resources to use freely according to its free trade propensity to import from different sources. The net effect is to shift the trade pattern even more in A's favour than aid which is untied or tied to a marginal project. 6

Table 10 Recipients imports-with tied non-project aid Country

Imports without aid

Aid

Import diversion through tied aid

Imports from freely usable resources (c)

A

400

60 (a)

+60

+40

Total imports 500 (45·5%)

B

500

40 (b)

-50

+50

500 (45·5%)

C

100

-10

+10

100 (9·0%)

1000

1100

(a) tied, (b) untied, (c) i.e. untied aid+free resources released by A's tied aid.

c Reciprocal tying

This deliberate diversion of trade through A's tying practice is likely to spark off a 'tying war', with other donors reciprocating these tying practices to regain some of their lost trade. In this simple model the only other aiding country B would now tie its aid to goods which would otherwise have been supplied by A or C. In this situation it is impossible to say a priori how the final trade pattern will be distorted from that where all aid is untied. This will depend on the country-wise incidence of the trade diversion, effects of each donors tied aid and the shares of aid provided. It should be noted that aid tying of this kind can displace the trade of other poor countries with Recipient. Many poor countries are faced with difficulty in expanding their exports to the markets of the developed countries. Greater intra-trade between developing countries would help to circumvent this problem. In so far as aid tying limits trade between poor countries it is working against one of the main purposes for which aid is provided-i.e. to relieve a foreign exchange constraint in the developing country. So far we have been concerned with the effects of aid tying on trade patterns. This is the main concern of the donors. However, the greater the degree oftying imposed by countries in protecting their uncompetitive export industries or their balance of payments position, the more is the value of a given amount of aid reduced to the recipient.7 This issue will now be examined.

Aid tying-effects on recipient countries Aid tying generally reduces the value of a given amount of aid to the recipient country. Because aid is tied by source, and often to specific commodities, the recipient is not able

Aid tying: trade and resource allocation effects

53

to purchase goods from those sources which have a comparative advantage in producing them. Aid tying, therefore, may force recipients into non-optimal patterns of resource allocation (which are not the result of their own inefficient domestic policies). This problem can be analysed and illustrated formally by some fairly simple models.

Model I Source-tied aid Let us assume, initially, that Recipient produces a single product and no exports. No domestic factors are required for production but two inputs, A and B, are imported on aid which is tied to purchases in the two aid-giving countries I and II. Source-tied aid is therefore the only way of financing the required inputs. It is assumed that the supply of input A is cheapest from source I and input B is cheapest from source II; prices being fixed at the source of supply. Finally the production function in Recipient is assumed to have isoquants convex to the origin. In Figure 6, A 1B 1 represents the available combinations of A and B Figure 6 Aid tying

Commodity A

H,

G J CommodityB under aid from source I where A is cheaper. A 2B 2 is the equivalent aid availability line from source II where B is cheaper. The total tied aid vailability line is EFG (Where FG II to A 1B 1 andEF 11 toA 2B 2 and OE=OA 1+0A 2 and OG=OB 1+ OB 2 • If the same total aid were untied to source, the aid availability line would be HF.18 which represents all possible combinations ofA and B bought at their cheapest source. It is only at

54

The economics ofaid

combination F that the two inputs can be bought at their cheapest source when aid is tied, and it is only by coincidence that the highest isoquant would touch EFG (tied aid availability curve) at F It is likely to touch at some point such as Q on isoquant ~. At Q, output is sub-optimal in relation to the level possible with untied aid, (point Pon isoquantXu). The domestic price ratio for the two inputs (=marginal rate of substitution between them) at equilibrium Q, would diverge from the 'best' international price ratio (the slope of HJ). Aid tying in this case not only distorts the allocation of resources in the recipient country, it can have the same effect on the donor country. In the above case not all requirements of input A can be met from the cheapest source (I). Hence there is a demand in source II on the industry producing A which has a comparative cost disadvantage relative to B.

Model II Source and commodity-tied aid If aid is tied to commodities as well as source in a very rigid way, then resource allocation is again made non-optimal in the recipient country. For example, suppose source I will provide a fixed quantity OA 2 of good A only and source II will provide OB of good B only on aid. The amount of output produced with these inputs would lie on the isoquant passing through F. If HJ represents the international price ratio for A and B, then F is a sub-optimal position, where the domestic price ratio diverges from the international price ratio. If aid were tied rigidly to source and commodity in this way F would be on an even lower isoquant (or level of output) than if aid were merely source tied (i.e. Q). 9 This may be a somewhat extreme case which arises where aid is provided in kind (e.g. wheat or cotton surpluses under U.S. P.L. 480 aid). However if aid is tied to the purchase of a limited range of goods (quantities not rigidly specified) as well as source, the analysis under Model I does not reveal explicitly a further reason why tying reduces the value of aid to the recipient. This is the opportunity which 'source and item' tying provides for 'monopolistic' pricing by industries in the donor countries. It will be remembered that Model I assumed that the prices of A and B were set at the source of supply. These prices, however, need not be 'competitive' prices. If, as happens frequently in practice, donor governments specify both source and items permitted under their aid, then the specific donor industry producing these goods becomes, in effect, the only source of supply. When the suppliers of these Items are few in number 'collusion' in fixing prices above the competitive level can occur on aid contracts. 10 Even if collusion does not take place the recipient of aid is likely to suffer from monopolistic 'price discrimination'. It is well known in formal economic theory that a monopolistic supplier will maximize his profits by charging a higher price in his domestic market and a lower price in the international market where the demand curve is normally more elastic. If the supplying firms in the donor country treat aid-financed sales as equivalent to domestic sales they will charge the higher monopolistic price rather than their competitive export price for the same commodity. 11 This means that double-tied aid is worth less in real terms to the recipient than equivalent untied aid which can be used to buy any items from any international source. Moreover, even if the source to which the aid is tied has a comparative cost advantage in a particular commodity (e.g. A from source I) the recipient may not fully benefit from this. Formally, in terms of Figure 1, double-tying resulting in 'collusion' and/or 'price

Aid tying: trade and resource allocation effects

55

discrimination' practices at source, will shift the aid availability curve EFG to the left (compared with a situation where competitive prices for A and B are charged at the supply source). In this model the recipients will now operate on a lower isoquant than~. The above model is umealistic in assuming that the recipient country produces only one product which is not exported and which is produced with only imported resources. If these assumptions are relaxed it can be shown that the recipient of tied aid can mitigate the adverse effects of tying on the available level of real resources and their optimum allocation.

Substitution possibilities Recipient can effectively untie aid, which is formally tied to source, if, from its own export earnings, it would have purchased anyway some goods from the aid sources. The conditions are that (a) aid is tied to source only, (b) the value of the tied aid does not exceed the value of the goods it would have purchased in any case from a given source, (c) the aid-giving source does not insist that aid purchases should be 'additional' to purchases which would have been made anyway. 12 If a recipient can fully exploit these substitution possibilities, even though aid is tied, it can avoid non-optimal allocation of resources. Thus, suppose in the case above, Recipient would have liked to buy OB3 of goods B from source II from non-aid funds. 13 It would also buy SP of B from source I on tied aid anyway because Pis an optimal point If OB3+SP exceeds the amount of tied aid available from source II (OB.) this aid is effectively untied. Total aid is also effectively untied since all the aid from source I would be used to purchase A from this source under optimal conditions. Hence Recipient by switching its tied aid from source I to finance purchases which would otherwise have been made from this source, releases free exchange to buy goods A from its cheapest source. The aid availability line then effectively becomes HJ.

Model III A more realistic model A more realistic model can be constructed by assuming that Recipient, uses domestic resources for production, produces more than one good and exports goods which finance the import of goods in addition to those available on aid. It is also assumed that the donor country insists on and achieves 'additionality' so that Recipient cannot substitute tied aid for financing imports it would otherwise have made from its export earnings. This situation is illustrated in Figure 7. Recipient can produce two commodities X and Y with domestic resources and the production possibility curve is convex. If JK is the international price ratio the country would produce a combination P. With the social welfare function represented by the U curves, the country would consume combination C0 (exporting QP of X and importing QC0 of Y). A single donor provides aid equivalent to KH of X or FJ of Y goods if aid were untied. However, the donor demands 'additionality' by specifying that the recipient should import GH of Y (on aid) in addition to its import QC of Y from its own resources.

56

The economics ofaid

Figure 7 Aid tying

y

F J

X The optimal consumption would be C* with untied aid. With the aid tied to Y and 'additional', one possible equilibrium would be C1 . This is clearly suboptimal and would require a consumption tax/subsidy policy. A 'second-best' optimization policy permitting a higher level of welfare than C1 could be achieved by an appropriate tax-cum-subsidy policy permitting the use of all the aid but changing the domestic output pattern in favour of X from P to P 2 • This might yield a position C2 which is preferable to C1 but still a lower level of welfare than C* under untied aid.

Empirical aspects The preceding analysis assumes that in the aid-receiving country there is full knowledge of production functions and unambiguous social welfare functions so that optimum positions would be taken if the tying practices of aiding countries did not prevent this. This, of course, is not always the view of the government providing the aid. One of their reasons for tying is implicitly that they consider themselves a better judge of priorities in the allocation of resources than the local planners. This seems to imply either that their technical (production) knowledge is superior and/or that their relative valuation of different goods more nearly reflects the valuation of the aid-receiving society than that of the society's

Aid tying: trade and resource allocation effects

57

politicians and planners. This important issue cannot be discussed any further here but to some extent it requires an empirical answer. Is there any empirical evidence on the magnitude of aid-tying inefficiencies? Clearly the extent to which tying interferes with the optimum choice of investment projects and techniques in the recipient country is largely unquantifiable. However, some qualification of the direct cost of aid tying (by source) is possible. The direct cost of aid tying can be defined for empirical purposes as the excess of the delivered price actually paid for goods by the recipient over the lowest delivered price which the country would have to pay if aid were not tied. The practical difficulties of measurement stem from lack of comparability of products imported from different sources and the problem of obtaining information on comparative prices from international tenders. Such evidence as exists 1 on Pakistan, Chile, Iran and Tunisia suggests that tying raises the procurement price of goods at least 10-20% above the price from the cheapest alternative source of supply. Notes

2 3

4 5

6

This would suggest that those donors with the greatest 'liquidity' problems would be the ones which tie their aid most strictly. Observation, however, does not confirm this correlation. See H. Singer, 'External Aid: For Plans or Projects', Economic J oumal, September 1965. In other words, the marginal propensity to import from different sources is assumed to be the same as the average. This may be unrealistic as the composition of imports will change as more resources become available for import. However, the assumption is used for simplicity. There may be some relationship because historical ties between countries affect their trade and aid relationships (e.g., U.K. and certain Commonwealth countries; U.S. and Latin America). See Ch. 1 above. It is impossible to ascertain by direct observation whether a project is marginal since it involves a comparison of an actual situation (with aid) and the hypothetical situation (without aid). Some inference may be made in practice, however. The precise diversionary effects ofA's tied aid on other individual countries cannot be specified a priori. It is assumed in Table 10 that they occur in proportion to the free trade shares ofB and C in Recipient's imports. It is clear, however, that A gains at the expense of other exporters as a whole by this type of tying. The size of the trade gain will depend on the free trade propensities to import and Recipient's non-aided imports relative to aid. It also does long-run damage to the export trade of the aiding country.

7 8 HFJ II A!B2" 9 If the aid sources tied their aid to that commodity in which they had a comparative cost disadvantage (A from source II and B from source I) the recipient would find that its loss of output from aid tying would be greater still. 10 It may be possible for the aid-giving authority to examine prices and prevent this practice. Administratively, this often presents great difficulty because of the range of goods, the comparability problem and the detailed knowledge of prices required. 11 This may not occur systematically since not all firms follow such pricing practices. However, there is empirical evidence that generally domestic prices and f.o. b. (competitive) prices differ in industrial countries. 12 When no imports from a source would otherwise have been made, there is no scope for substitution. However, condition (c) is difficult to enforce in practice, since it is not possible to infer the amount of goods which would hypothetically have been purchased in the absence of aid. In practice, a donor may be very effective in achieving 'additionality' if it specifies items purchasable

58

The economics ofaid

under tied aid which the recipient has been observed not to purchase in the past from it out of its own foreign exchanges. A policy ofthis kind, however, would seem to defeat the whole objective of aid by tying it to goods which have a low or zero development value. 13 For some other use than the production of X in the above simple model. 14 See bibliography.

7 Some questions and further issues

The preceding chapters have been concerned with setting up an analytical framework for the economics of aid. If these chapters have succeeded in giving the reader a simple, orderly conceptual apparatus for approaching aid issues they will have fulfilled their intention. But the simplicity and neatness of the analysis is misleading and dangerous because the development process and the role of aid is infinitely more complex and subtle than these simple models and principles suggest. It is salutary to end on a somewhat sceptical note in case the impression is given that economists know more than they really do about the key factors in economic growth or that the preceding analysis provides any more than a first step in providing a basis for aid policy decisions. This chapter will briefly raise a number of questions about (a) The role of capital in economic growth, (b) the use of performance criteria for allocation of aid, (c) the use of aggregate models for aid policy, (d) the relationship of trade and aid policy in development.

Role ofcapital and other factors in growth With little qualification the underlying rationale of aid has been that capital formation is an essential and major factor in growth and traditionally much economic theory has provided support for this view. Foreign capital is typically seen as a supplement to domestic savings which enables domestic investment to be higher than it would be on the basis of domestic efforts alone. A predictable, and often fixed, relationship is postulated between investment during a period and the extra flow of output, assumed to be generated by it. This view which relates output growth to a single factor-capital invested in physical goods-strains credulity. It takes little perception to realize that many other factors than financially measured investment must determine any extra measured output which is produced in a period. The importance of the appropriate skills of the labour force, the enterprise and organizational capacity of those making decisions, the natural environment and the timing of discoveries of new natural resources, the level of technical knowledge and the stability of the political environment are only some of the other general influences on economical productive activity. Casual observation reveals that some countries have had rapid growth without significant capital formation and without substantial capital import; others have had high rates of investment and external assistance yet low growth rates. Neither extra domestic capital formation nor more external capital is a necessary or sufficient condition for faster growth. Recent systematic analysis, which cannot directly answer questions of causality, has nevertheless suggested that investment rates in different economies 'explain' only a small part of the observed differences in the growth rates of these economies. Other factors appear much more important and these may not be universal factors but influences which are specific to each economy. DOI: 10.4324/9780203840177-7

60

The economics ofaid

In any case, is it the savings and investment which generate the growth or is it the growth which generates the opportunities for productive investment and the conditions for high savings propensities? There is clearly a two-way relationship which raises even more doubts about the interpretation of the observed statistical association between investment and domestic savings rates, on one hand, and growth rates on the other. The answer to these fundamental growth questions is highly relevant to any theory of aid or aid policy. Can we say only that in environments where conditions are favourable to growth, capital inflows will be highly effective and that where aid is likely to fall on stony ground it should not be made available? Or can aid be used directly or indirectly to create more favourable growth conditions? Can externally financed and chosen projects break bottlenecks in an economy and so create growth conditions or stimulate new economic activities by their demonstration effects? Or are foreign financed projects more likely to do harm than good? It has been contended, for example, that aid-financed projects are frequently too grandiose, too capital or import intensive and too umelated to the requirements of the domestic economy. Aiding governments or their officials may, for various reasons, have a bias towards financing large industrial or overhead projects. This may cause a diversion of local resources and interest from smaller projects in the agricultural sector of developing economies which might well yield higher returns and relieve a major constraint on growth. More indirectly, should aid be used as a 'carrot and stick' device to induce governments of certain developing countries to follow policies and activities likely to be more favourable to growth and development? This raises the question of 'performance criteria' in aid allocation, which is discussed at greater length below.

Performance criteria The suggestion has frequently been made that aid should be made dependent or conditional on the performance of the recipient. The idea has considerable appeal to those who think help should be given to those who help themselves. In Chapter 2 one such criterion discussed was the proposal to make aid available only to those countries likely to become independent of aid in a given period of time. Several major problems arise in this type of approach. First, what are the development objectives against which judgment is to be made? Growth of conventionally measured national product (or national product per head)? This largely ignores other elements which might be termed 'quality oflife' factors as well as the distribution of measurable income. It is possible of course to devise 'composite' indicators of development which bring together quantitative indices of its various strands such as indices of improvement in the educational system, health and nutrition and measures of political and social stability where they can be devised. However, apart from the problem of choosing the relevant components, what weight is to be given to them in the sum? Difficult value judgments are involved here but does the composite index present any greater problem than that involved in using national product indices which weight the components by their market valuation? An alternative approach is to measure a country's performance by 'intermediate' variables considered to be instrumental in development; such as savings rates, export growth, tax revenue. Whatever variables are used a further problem arises-the need to distinguish

Some questions and further issues

61

between performance and success; between the controllable and the uncontrollable elements in an economy's observed growth, savings, exports or any other target variables. Ideally, aid should be related to that part of the observed outcome which is under the control of the government of a country. Because too little is known about the precise impact of various policy measures it is not possible in practice to identify the controllable elements in the situation. An improvement in the savings rate of a country may be due to greater tax efforts or to a fortuitous change in the terms of trade. A third problem arises, even if performance can be measured in a meaningful way. This is the weight to be attached to 'poverty' as distinct from 'performance' in the allocation of assistance. This presents very difficult value judgments for the aiding countries since the poorest countries are often likely to be the worst performers.

Aggregate models Whether 'performance' tests are used or not there is a need to estimate the volume as well as the nature of external capital flows required by developing countries. There are, in brief, two approaches to the problem-macro and the micro methods. Aggregate growth models which provide a method for estimating the aggregate resource gap have been discussed at some length in Chapters 3 and 4. Attention must now be drawn to the weakness of this approach in practice and the alternative (or at least supplementary) view of total aid inflows as the sum of a series of micro-economic analyses and decisions. The aggregate two-gap models examined in earlier chapters have many attractions. To the tidy-minded theorist they are neat, clear-cut and rigorous. The growth process is reduced to a simple mechanical model whose parameters can be altered to see what the implications are for aid flows. To the policy maker these models are also attractive. They provide an objective (a target rate of self-sustained growth), some performance standards in the aided country (e.g. domestic savings rates or export growth rates) and a method of calculating the volume and duration of aid flows required to achieve the objective. Theoretical questions about these macro-models have already been raised in earlier chapters. Here it is necessary to consider the weaknesses of these models as a practical guide to policy. In the first place the models are too mechanical. They can only be relied on to predict future aid requirements if it can be assumed that the values of the parameters will remain stable in future or that past experience is an accurate guide to future behaviour. This is often highly implausible. In the early stages of development, for example, heavy investment in infrastructure may be necessary to establish conditions for future growth and the capitaloutput ratio which prevailed at this stage would be valueless as a guide to that likely at later stages of development. Similarly, marginal savings rates can shift radically over timeboth upwards and downwards depending on domestic conditions and policies. In practice, therefore, no long-run predictions could be safely made from these models and continuous assessment would be necessary; based on informed judgment about the behaviour of the main parameters and not merely on statistical projection. In the second place, the aggregate nature of these models makes no allowance for the absorptive capacity of the economy for capital. The ability of a country to formulate and execute productive projects is not taken directly into account in aggregate exercises of this

62

The economics ofaid

nature. Local skills, administrative capacities and experience may effectively constrain the amount of capital inflow that can be productively used in a given period below the amount that aggregate models may suggest. The result would be a diversion of some or all of the capital inflow into consumption. Thirdly, aggregate models do not indicate the appropriate allocation of aid. Where are the bottlenecks? Where is the rate of return highest? These are micro-economic questions which involve a detailed examination of the different sectors, industries and projects and require techniques of cost-benefit appraisal and linear programming. They require much data collection, enquiry and local judgment over a wide area. This approach to aid requirements is less spectacular and simple but is likely to come much closer to assessing the appropriate aid flows for a particular country as well as the most strategic points to which it should flow. Since one of aid's vital functions is to relieve strategic shortages and bottlenecks in various sectors or areas of an economy, it cannot be viewed simply as an aggregate supplement to domestic resources. Its importance lies largely in its disaggregate nature and impact at many different points and places. The appraisal of projects considered for foreign capital assistance, involves the study of investment economics and cost-benefit analysis. An introduction to this field is provided in a companion book in this series (Investment Economics by J.L.Carr). 1 It is hardly necessary to add that to make an effective appraisal of projects at the micro-level, it is necessary to have knowledge of the prospective total or macro-situation in an economy. To this extent both micro and macro analysis are complementary to one another in the difficult task of assessing the scale, pattern and impact of aid flows.

Aid and trade Discussion in this book has been confined to the economics of aid and nothing has been said about the trade policies of the wealthier countries and their effect on developing countries. What is the relationship between trade and aid? Are they substitutes or complements? Some trade measures designed to help the export earnings of developing countries involve an implicit transfer of resources and therefore have a certain aid content. International Commodity Agreements to raise primary product prices or preferential trading arrangements that permit less developed countries to charge higher prices for their industrial products in developed country markets, involve an explicit resource transfer. This is a direct transfer from the consumers in developed countries to the governments or individuals of less developed countries. Most trade measures (such as reduced tariffs or quantitative restrictions) taken by developed countries would increase export opportunities to developing countries. They involve no direct resource transfer. However, is there a sense in which increased trading opportunities are substitutable for aid? It has been argued, for example, that aid is a 'soft option'. For the rich countries it is easier politically to place the burden of assisting development on the taxpayer (via aid) than to overcome the resistance of domestic producers to commercial policies which would expose them to greater competition from producers in the developing countries (e.g. textiles). For the poor countries, moreover, the artificial barriers to their exports in developed

Some questions and further issues

63

countries may force them to establish domestic industries which replace imports. These industries are often inefficient because they are sheltered by protection from the need to compete internationally. Inefficiency can therefore be encouraged in both rich and poor countries because it seems easier to provide aid than to provide freer access to developed country markets. Aid, in fact, may be financing some of the unsuitable, uncompetitive import-saving industries which developing countries are partly forced to establish because the industries where they have a comparative advantage over the rich countries (textiles, leather goods, etc.) are restricted in access to the rich country markets. It is possible to show (following Professor H.Johnson) that greater trade opportunities may be compared to greater aid flows. Clearly additional trade opportunities do not provide additional real resources for investment like aid. Instead, they provide the opportunity to convert extra domestic resources into foreign resources without the costs that would result if the trade opportunities did not exist. If the alternative were to establish import-saving industries, the cost of this is the extent to which domestic production of import-substitutes exceeds their world price. If the alternative were to sell more exports by lowering prices then the cost is the difference between the price at which they would be sold given the trade opportunity and the marginal revenue from selling more exports without the opportunity. In both cases the opportunity cost of the extra trade opportunity represents a real resource gain (like aid) to the country receiving it. These are the gains of maximum specialization. Although greater trade possibilities for developing countries have some resource transfer element in them, they clearly are more complementary to aid flows than substitutable for them. Development requires both; the greatest scope for specialization and trade as well as external resources for investment to exploit these opportunities to the fullest possible extent. Notes An excellent practical guide to industrial investment appraisal in the context of developing countries is Manual of Industrial Project Analysis, Vol. II, by I.MD.Little and I.A.Mirrlees, O.E.C.D., 1969.

Appendix Resource gap models: statistical estimation ofthe parameters

The ex-ante and ex-post relationship between growth and foreign capital inflows were examined in the context of an aggregate model in Chapter 4. The three major structural parameters in the model were: The marginal output-investment relation (and) or its universe the incremental capital-output ratio (I.C.O.R.) (ii) The savings-income relation: the marginal savings ratio (s) (iii) The relationship of imports to income: the marginal propensity to import (m). (i)

The behaviour of exports can also be related to income in the same way as imports but exports are more plausibly treated as exogenously determined. The prediction of foreign capital requirements by the use of ex-ante 'saving' and 'trade' gap models requires empirical estimation of these parameters. This raises a number of problems about the meaning of these concepts, the degree of aggregation involved in their use and the difficulty of measuring their value from historical data for prediction. Each structural parameter will be examined in turn.

(i) Marginal output-capital ratio The marginal or incremental output-capital ratio is used in most studies to predict the level of investment required to achieve a given growth in output for an economy. Some disaggregation is usually attempted since the economy's structure normally changes as it grows and different sectors have different coefficients. Instead of a single value of national product in any period, 'n' sectoral value-added magnitudes may be predetermined and each sector's required investment will be a function of value added in the sector. The aggregate of the sectoral investment requirements yields total investment requirements. It is usually easier to make a forecast of sectoral growth than aggregate growth and this procedure has the advantage that alternative estimates of investment needs can be made based on alternative development programmes or projected structures of an economy. The problem still remains of obtaining a value for the marginal output-capital ratio for each sector which is basically defined as increase in value of output during year t investment in year t. This may be in net or gross terms and it may involve a lag relating output changes to investment in any earlier period. This coefficient may be estimated from time series data. However,

Appendix: Resource gap models: statistical estimation ofthe parameters

65

historical data usually show a very unstable relationship between investment and the increment of output. The reasons are obvious. Many factors other than investment in physical capital affect the addition to output or value added, particularly in the short period. The influence of technical change, availability of scarce manpower, degree of utilization of existing and additional capacity and (in the case of the agricultural sector especially) weather conditions, are all picked up and inputed to investment by the historical approach to measurement of this relationship. It is not surprising that observed incremental output-capital ratios (aggregate and sectoral) are not statistically independent of growth rates but vary directly with them. (See H.Leibenstein, 'Incremental Capital-Output Ratios & Growth Rates in the Short Run', Review ofEconomics & Statistics, Vol. 48, February 1966.) History does not always repeat itself; the coefficient derived from some past boom period and its particular technological conditions may not be a suitable guide to a future period when conditions may be completely different Unfortunately, the estimation of foreign resource gaps over a short period of time (say 5 years) is particularly sensitive to the value of this parameter so that errors in its estimation can have very significant implications for foreign aid requirements.

(ii) Marginal savings ratio The second major parameter of the growth models is the ratio of extra aggregate saving to extra aggregate income; the marginal savings ratio. When used for prediction there are problems arising from the high degree of aggregation involved in this concept. There are also difficulties in using historical data for estimating the value of the parameters, because shifts in the saving function can occur in developing economies if there is a growth in financial institutions and new opportunities for the use of savings as well as changes in government policies and tax/expenditure behaviour. It is useful to disaggregate into at least three major sectors-government, household and corporate and estimate savings functions for each sector. In most developing countries government savings (i.e. excess of government current revenues over expenditures) play a very important role. The government savings function is particularly difficult to specify as it will depend on the existing tax structure, changes made in the tax structure and in tax rates, which depend on the political situation and attitude and strength of the country's rulers. Since the majority of developing countries have a predominantly 'commodity' tax structure it may not be umeasonable to assume that government revenues would rise only proportionately with G.N.P. If government current expenditures rise in the same way, government savings would be proportional to G.N.P. However this limited elasticity of indirect tax systems in developing countries is not inevitable if changes in the tax structure are made. Household savings behaviour in developing countries presents difficulties because of paucity of data. It is generally held that wage-earners have a very low or zero propensity to save in most developing countries. However, the household sector also includes peasants and owners of unincorporated enterprises who are considered by many economists to have high savings propensities out of disposable incomes. The savings behaviour of these groups may be strongly affected by institutional factors such as the extent of banking and credit

66 Appendix: Resource gap models: Statistical estimation ofthe parameters

insurance and pension institutions, savings propaganda and incentives, and the outcomes may well depend heavily on the institutional character of the economy and government policies with respect to savings. Savings of the corporate sector should be isolated in developing countries because not only does it tend to have the highest sectoral propensity to save but it is usually the fastest growing sector in the economy. In those semi-industrialized economies where the corporate sector largely serves the domestic market, some estimate of cmporate savings may be made from the predetermined growth of value added in sectors like manufacturing, transport and public utilities where cmporate organization is predominant. However, in many less developed countries, cmporate organization can be identified with the export sector of the economy including the often foreign-owned mining, plantation and trading companies. In this case cmporate profits and savings will be a function of the country's export earnings. Government saving may also be mainly affected by export earnings since much of their revenue often comes from indirect taxes on foreign trade and direct taxes on export companies. Savings will then tend to be a function of the largely exogenous factors which determine the future cycle and trend of exports of the particular commodities in which the economy specializes. Moreover, the marginal effect of profits and company savings will be larger, if the given change in export earnings is a result of a change in the price of exports rather than a change in their volume. More generally this influence of the terms of trade in an export economy can have a crucial effect on both savings behaviour and, of course, the value of exports when expressed in the numeraire of import prices.

(iii) Marginal propensity to import Predicting the behaviour of imports is again likely to be more accurate if some disaggregation is attempted. To estimate the total import payments requires a distinction between debt service payments, invisible imports and visible imports. The main problem is to estimate visible imports. Two disaggregated approaches are possible; first by classifying imports by economic sector of destination and second by classifying imports by type of goods. The first procedure involves the use of an input-output matrix for the economy and is usually less practicable in most developing countries than one which classifies imports into (1) consumer goods (2) finished capital goods (3) raw materials and semifinished (intermediate goods). Imports of consumer goods are likely to be a fairly stable function of total national income. Imports of intermediate are usually a fairly stable function of value added in the manufacturing sector of the economy. The primary producing sector and the services sector of an economy are unlikely to be significant users of imported intermediate goods (with the exception of fertilizers and transport fuels). Capital goods import will be closely related to the sectoral investment levels since exports of capital goods are negligible in most developing countries. The major conceptual problem in estimation of import requirements is to allow for the effect of import-substitution which occurs in a growing economy. It might be expected that investment in industries producing import substitutes would tend to reduce imports over time. In practice, although imports may fall as a proportion of G.N.P. they tend to increase

Appendix: Resource gap models: statistical estimation ofthe parameters

67

absolutely with growth. The main explanation of this somewhat paradoxical result lies in some or all of the following factors. a The establishment of new (import-saving) industries requires investment with a high import content. So long as the investment level and the (imported) capital-output ratio remains constant, the impact of this year's investment on the demand for imports should be more than offset by last year's investment in import substitutes now coming to fruition. However, if the investment level increases overtime and the (imported) capital-output ratio rises with the move towards a more capital-intensive type of industry, then import requirements will not necessarily fall over time. b Further, the additional output of these new industries increases the requirements of non-indigenous materials, fuels and components, etc. Hence the net import-substitution, or value added domestically, is often a small proportion of the value of the finished output. c The additional income (wages and distributed profits) generated by the new activity creates extra expenditure on imported consumer goods and services. The effect on imported consumer goods will depend, among other things, on the government's domestic tax policies, the effective exchange rate and whether there are quantitative import controls. This combination of forces can therefore result in rising import requirements over time, despite investment in import-saving industries. The precise outcome depends on so many parameters that no simple model can be constructed.

Further reading

Three major works on economic aspects of aid are: I.MD.Little and I.Clifford, International Aid, 1965 R.Mikesell, The Economics ofForeign Aid, 1968. LB.Pearson (Chairman), Partners in Development, 1969.

Ch. 1. Politics and Statistics ofAid The politics of aid has now developed a large literature. Some useful references are: I.MD.Little and I.Clifford, International Aid, chs. 1-3. C.Kaplan, The Challenge ofForeign Aid, 1966. Joan MNelson, Aid, Influence & Foreign Policy, 1968. Tom Soper, 'Western Attitudes to Aid', Lloyds Bank Review, October 1969. The main sources of statistics on assistance are the Development Assistance Committee of O.E.C.D., Annual Review and The Flow ofFinancial Resources to Developing Countries (Annual). United Nations; International Flow ofLong-Term Capital and Official Donations, 1951-9 and Annual World Economic Survey. For the U.K., the Ministry of Overseas Development publishes annually British Aid Statistics. For official U.K. policy, reference should be made to the White Papers, Cnmd. 2147 and 2736Aid to Developing Countries, 1963 and 1965. Cnmd. 3180 Overseas Development: The Work in Hand, 1967. A further critical insight into the administration of the U.K. aid programme can be gained from a study of the Reports of the House of Commons Select Committee on Overseas Aid.

Ch. 2. Aid Sharing and Allocation I.MD.Little, International Aid, ch. 3. I.Pincus, Economic Aid & International Cost Sharing (Rand), 1965, chs. 3 and 5. 'Cost ofForeignAid', Review ofEconomics & Statistics, November 1963. WE.Schmidt, 'Economics of Charity', Journal ofPolitical Economy, August 1964. W.Kravis and I.Davenport 'The Political Arithmetic of International Burden Sharing', Journal of Political Economy, Vol. 71, No. 4, August 1963. J.Fei and D.Paauw, 'Foreign Assistance & Self Help', Review ofEconomics & Statistics, Vol. 47, August 1965.

Ch. 3. Aid and Growth R.McKinnon, 'Foreign Exchange Constraints in Economic Development', Economic Journal, June 1964. Reprinted in International Trade, Ed. J.Bhagwati (Penguin Modem Economics), 1968. H.Chenery and A.Strout, 'Foreign Assistance & Economic Development', American Economic Review, Vol. 56, No. 4, September 1966.

Further reading 69 Ch. 4. Estimating Aid Requirements P.Rosenstein-Rodan, 'International Aid for Undeveloped Countries', Review ofEconomics & Statistics, Vol. 43, No. 2, May 1961. G.Ohlin, Foreign Aid Policies Reconsidered (O.E.C.D.), 1966, ch. 4. Chenery & Strout, op. cit. I.Vanek, Estimating Foreign Resource Needs for Economic Development, New York, 1967, chs. 1 and 6 especially.

Ch. 5. Debt and Terms ofAid D.Avramovic (Ed), External Debt & Economic Growth (I.B.R.D., Johns Hopkins) 1964, chs. 3 and 5. G.M.Alter, 'The Servicing of Foreign Capital Inflows by Underdeveloped Countries' in Economic Developmentfor Latin America (Eds. H.S.Ellis, H.C.Wallich) 1961. Pearson, op. cit. ch. 8.

Ch. 6. Aid Tying Little and Clifford, op. cit, ch. 7 and 12. H.W.Singer, 'External Aid: For Plans or Projects', Economic Journal, September 1965. A.Lovell, 'How Should Aid Be Given?', Lloyds Bank Review, April 1966. J.Bhagwati, 'Tying ofAid', UNCTAD. Doc. TD/7/Supp. 4, 1 November 1967. Empirical Estimates are given in The Cost of Tying to Recipient Countries', UNCTAD Doc. TD/7/ Supp. 8, 21 November 1967. Pearson, op. cit. ch. 9.

7 General Texts on Development Economics H.Myint, Economics ofDeveloping Countries (Oxford). B.Higgins, Economic Development (Norton). H.J.Bruton, Development Economics.

Ch. 8. Further Issues L.Pearson, Partners in Development. This covers the issues of performance criteria for aid and trade policies for development. This document, the product of an International Commission, is the latest and most comprehensive review of the development issue and it provides not only a good summary of the major policy issues in this field but also points the way forward to greater and more efficient international co-operation for aid and development.

ROUTLEDGE LIBRARY EDITIONS: DEVELOPMENT

AID AND INEQUALITY IN KENYA

AID AND INEQUALITY IN KENYA British Development Assistance to Kenya

GERALD HOLTHAM and ARTHUR HAZLEWOOD

Volume 9

LONDON AND NEW YORK

First published in 1976 This edition first published in 2011 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 This edition published in the Taylor & Francis e-Library, 2011. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk. Routledge is an imprint of the Taylor & Francis Group, an informa business © 1976 Overseas Development Institute All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 0-203-84016-X Master e-book ISBN

ISBN 13: 978-0-415-58414-2 (Set) eISBN 13: 978-0-203-84035-1 (Set) ISBN 13: 978-0-415-59273-4 (Volume 9) eISBN 13: 978-0-203-84016-0 (Volume 9) Publisher’s Note The publisher has gone to great lengths to ensure the quality of this reprint but points out that some imperfections in the original copies may be apparent. Disclaimer The publisher has made every effort to trace copyright holders and welcomes correspondence from those they have been unable to contact.

AID AND INEQUALITY IN KENYA BRITISH DEVELOPMENT ASSISTANCE TO KENYA

GERALD HOLTHAM and ARTHUR HAZLEWOOD

CROOM HELM LONDON in association with THE OVERSEAS DEVELOPMENT INSTITUTE

First published 1976 This edition published in the Taylor & Francis e-Library, 2011. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk. © 1976 Overseas Development Institute Croom Helm Ltd, 2–10 St John’s Road, London SW11 ISBN 0-203-84016-X Master e-book ISBN

ISBN 0-85664-344-0 (Print Edition)

CONTENTS

1 2

3

4

5

Acknowledgements Introduction The Background Aims and Achievements Since Independence

ix x 1

The Agricultural Sector

7

7

Manufacturing

12

African Participation

17

The Role of Government

21

The Current Situation

24

The Facts of Aid British Capital Aid

31

British Technical Assistance

42

The Commonwealth Development Corporation

43

Aid From Other Sources

45

The Costs of Aid

47

British Aid Policy and Administration in Kenya The Early Years, 1963–69

51

Aid Administration Today

55

Policy Issues

60

Other Donors’ Policies

68

The Commonwealth Development Corporation

72

Aid in Action Part One: The Land Transfer Programme

76

34

51

76

1960–65

77

1966–74

82

The Effects of British Aid on Land Settlement

87

The Evaluation of Aid for Land Transfer

91

Appendix: A Model of Expropriation

101

Contents Part Two: The Mumias Sugar Company

6

7 8

vii 109

History of the Project

110

The Project in Prospect and Practice

112

The Outgrowers

115

Mumias and Rural Development

116

Mumias and ‘Appropriate Technology’

119

Other Considerations on Mumias as an Aid Project

122

Part Three: The Special Rural Development Programme

125

SRDP in Kwale

129

The Lessons of SRDP for Aid Donors

135

SRDP and District Development

137

The Influence of Aid Fungibility

141

Public Finance

146

Other Government Policies

155

Administration

162

Political Economy

168

Conclusion: Incomes and Income Equality

172

Policy Implications for British Aid Controversies Over Aid

180 191

Index

204

141

LIST OF TABLES

Table 1

Sector shares of monetary GDP (constant 1964 prices)

39

Table 2

Use of resources (current prices)

40

Table 3

Source of aid disbursements to Kenya

50

Table 4

Technical assistance to Kenya and EAC by country of origin, September 1973

51

Table 5

British official capital aid commitments to Kenya since independence

53

Table 6

Total British aid disbursements to Kenya and the East African Community

56

Table 7

British aid disbursements at 1964 prices

57

Table 8

British technical assistance disbursements to Kenya and the East African Community

61

Table 9

British technical assistance to Kenya: by recipient ministry at 31 March 1974

62

Table 10

World Bank commitments to Kenya 1970–74

68

Table 11

HMG finance for Million-Acre Scheme

109

Table 12

Britain’s commitments to land transfer since 1970

114

Table 13

Estimated value of production by large mixed farms in Kenya, 1963–71

129

Table 14

Costs and potential benefits of land transfer

131

Table 15

Private capital flows in the independence period

132

Table 16

Imports of sugar

146

Table 17

Kenya government expenditure and finance

192

Table 18

Distribution of commercial bank lending

193

Table 19

Use of resources available in the monetary economy

256

ACKNOWLEDGEMENTS

Our thanks must first go to the Social Science Research Council which has financed the research for this book, as it did the companion studies on British aid to Malawi and to Botswana, Lesotho and Swaziland. We are indebted to the Director of the Oxford University Institute of Economics and Statistics for permission for one of us to devote his time to the study, and to the Director of the Institute for Development Studies, University of Nairobi, for providing accommodation and a congenial working atmosphere during our stay in Nairobi. Our other acknowledgements for help we have received in carrying out the study must be brief, not because we have few debts of gratitude but because we have so many. A large number of civil servants of various nationalities in Kenya and others at the Ministry of Overseas Development, UK, gave of their time and experience and we have had the benefit of conversations and correspondence with academic colleagues in both countries. We should like to thank them all. The research also had the benefit of an Advisory Committee of eminent academics, administrators and businessmen chaired by Professor Ian Little, and colleagues at the Overseas Development Institute, headed by the Director, Robert Wood, took a lively interest throughout. We should like to thank them. Gratitude is also due to the secretarial staff at ODI who typed, reproduced, collated and despatched numerous drafts. For errors of fact and judgement that remain the authors are solely responsible, having only the refuge of co-authorship, that they can unhesitatingly blame each other.

INTRODUCTION

Our task has been to describe and assess the implications of aid for the economic development of independent Kenya and to suggest conclusions for policy, particularly British policy. Economic development cannot be assessed without making value judgements. Even after these have been made it is necessary to remember that there is no such thing as an economy. Movements in variables commonly regarded as economic have effects that are commonly classified as political or social; these in turn will certainly affect economic developments. To take a narrowly economistic view is often to misunderstand patterns of causation. So in saying something about the influence of aid on the economic face of Kenyan society we have tried to keep the administrative and political faces in focus. Even so, we despair of taking everything properly into account. We can think of four ways in which we might have proceeded: firstly we could have analysed economic, political, social and aid statistics and attempted to infer causation; secondly, we could have made case studies and attempted to generalise; thirdly, we could have talked to lots of people, studied statements of policy and administrative procedures, collected theories, impressions, anecdotes, opinions with supporting facts and attempted to form a coherent view—a sort of judicial approach; fourthly we could have prejudged the issue and collected selected facts to support our predetermined conclusion. In writings on aid and development, the last is quite a popular technique. Indeed, it has been claimed to be the only respectable technique on the grounds that no one can really be disinterested and objective and an approach that claims to be so can only be sinister ideological apologetics. We are reluctant to descend into this abyss of relativism. Through the mists of conditioning and (after all—distant) self-interest it is sometimes possible to identify a spade as a spade for all practical purposes and to secure agreement in calling it such. One can then say whether one likes it or not. In any case, we had (and in some cases still have) insufficient self-confidence in the correctness of our judgements, to make this fourth a sustainable approach. The difficulties with the first approach in studying aid to Kenya are more than the usual ones of sparse and unreliable data and the compatibility of tests of statistical significance with more than one theory of causation. In Kenya the size of aid flows in money terms has not been great relative to many other economic aggregates which are subject to other influences. It is, therefore, impossible to study the latter and infer anything, even probabilistically, about the effects of the former without a properly specified model of the whole political economy. Furthermore, statistical aggregates disguise components of considerable heterogeneity and this is particularly true of aid statistics. One senior adviser’s salary may be insignificant macroeconomically but his influence conceivably could be out of all proportion to it. Parameters will therefore be unstable and hypothesis-testing likely to yield consistently negative conclusions. That probably just reflects the impossibility of making valid high-level generalisations about aid. Data, however, are certainly inadequate for more refined statistical analysis of less general propositions. Statistical analysis can in our view be no more than a check on conclusions arrived at by other means.

Introduction xi We have relied therefore largely on the second and third approaches. These methods are, of course, far from ideal. Case studies can be of special cases that do not generalise and anyway it may not be possible to get an impression of the characteristics of a wood from a study of single trees. Neither is the truth always anything like a correctly weighted summary of people’s perceptions, even if one could weight accurately and people related their perceptions with unvarying truthfulness. Other methodological problems confront any approach. Many of the propositions in this study must be based on ‘counterfactual hypotheses’ that, by their nature, cannot be tested. A lot of argument about aid is possible even if people agree in some minimal sense about what it did, because they can fall out over what would have happened without it. Often those most vociferous in accusing others of disguised value judgements are guilty here; they do not distinguish between the most probable outcome in the absence of aid and their preferred outcome. Hence aid to the same country can be said by the radical left to have reinforced dependency and postponed socialism, by a radical liberal to have pre-empted indigenous capitalism causing centralisation and pauperism, and by an aid-employed bureaucrat to have prevented economic stagnation and political chaos. We have heard all these views expressed about Kenya, the first and third more than the second, We shall now admit a bias in our perception: we are sceptical of any cataclysmic view. Societies are organisms not usually given to violent changes of course as a result of relatively minor stimuli. Looking at aid in a ‘partial equilibrium’ framework, our money would be on the proposition that not one of these views correctly describes the workings of aid to Kenya although aid has had effects described, and exaggerated, in each one of them. We can only hope that this bias has not blinded us to the truth on occasions when aid was critical. The perception is, of course, partly the result of our strictly limited approach. We really have been concerned with the aid relationship and the influence of aid rather than with the whole complex of relations between Britain and Kenya. Other elements of the relationship have been considered only intermittently where they seemed to impinge directly on aid. This, we believe, is justifiable. Whatever the genesis of aid it is now administered in Britain and elsewhere by a distinct government department that in important day-to-day decisions enjoys considerable autonomy. It is possible therefore to study the effects of these decisions separately within the larger framework. Even those people, for example, who are satisfied that Britain’s relationship with Kenya is one of steady exploitation may like to know whether aid is merely part of the medicine, or the sugar around the pill. At this point we shall not define our terms—on the grounds that there is a considerable consensus about the denotation of ‘aid’ even if people cannot agree about a statement of its connotation. We have dealt mainly with the ‘aid’ programme to Kenya administered by the Ministry of Overseas Development, ignoring both non-governmental aid by voluntary agencies and some general ODM activities like financing research, and students visiting Britain, which benefit Kenya along with other less developed countries (Idcs). ‘Development’ is, admittedly, more difficult than ‘aid’ and we have sometimes used this piece of shorthand. Where serious ambiguity was possible we have tried to circumvent it by discussing positivistically the effects of aid and then saying whether we like them, rather than disputing whether aid has been conducive to ‘development’. In pursuit of the approach outlined the book has the following plan: the first two chapters provide a background to the remainder of the study. Chapter One gives a description of the

xii Introduction main features of the Kenyan economy at the time of independence; Chapter Two relates and discusses the progress of the economy since independence and describes and interprets Kenya government policy. The next four chapters contain the bulk of the study’s findings. Chapter Three gives the main facts of British capital aid and technical assistance, and brief summaries of the aid of some other donors and attempts to refine official statistics. Chapter Four is an historical discussion of British aid policy and a description of the current British policy stance and administrative arrangements, with some comparison with other donors. Chapter Five contains a detailed account and analysis of the Land Transfer Programme, quantitatively and historically the most important part of British aid, and case studies of two other aided projects. Chapter Six analyses the concept of fungibility and its implications for aid evaluation and then discusses in turn aid’s effects on Kenyan public finance, development policy, the Kenyan administration and the Kenyan political scene. It concludes with judgements of aid’s effects on income levels and income equality in Kenya. The last two chapters are concerned with the implications of the study’s findings. Chapter Seven suggests the policy implications for donors, particularly Britain. Chapter Eight reviews some academic controversies about aid in the light of the study.

1 THE BACKGROUND1

Kenya is no longer White Man’s Country2 but its economic structure, its economic and social policies, and the form and function of the international aid it has received and is receiving are deeply affected by its history as a land of European and Asian settlement. Before Kenya achieved national independence in December 1963, a fundamental determinant of the nature of its economy and of the policies of its government was the existence of wealthy and—by tropical African standards—relatively large non-African communities. By the time of independence, changes had been in progress for several years, and 1959 and 1960 were the last years in which the situation had not been deeply influenced by the approach of majority rule.3 In 1960 the total population was estimated to be 8.1m. of which 7.8m. were Africans and 169,000 Asians. The European population numbered 61,000. There are no adequate data of the racial distribution of money income, but from all the evidence it is clear that, despite the overwhelming numerical preponderance of Africans, non-Africans received a high proportion of the total. Tax returns show 92 per cent of Europeans receiving incomes of over £400 p.a. while only 0.5 per cent of Africans were in this income group.4 Eighty per cent of the value of the marketed produce of agriculture came from the European-owned farms and estates; 55 per cent of the total wage-bill accrued to non-Africans, though they amounted to only 10 per cent of the labour force; profits from manufacturing and trade were received almost entirely by non-African individuals or companies. Pre-independence agriculture was characterised above all by the division of the land between Europeans and Africans. Asians were largely excluded from the ownership of agricultural land, and Africans were prohibited from acquiring land in the ‘White Highlands’, which by the Agricultural Ordinance of 1955 became officially invested with the more neutral title of the Scheduled Areas. These European-owned Scheduled Areas occupied some 7½m. acres, about half of which was suitable for arable farming, the remainder being limited to pastoral use by lack of rainfall. The African lands—the Reserves, or Native Trust Lands—totalled about 130m. acres, but only about 18m. were suitable for agriculture. Low rainfall restricted the use of the remainder of the non-scheduled areas to grazing, nine-tenths of it being classed as ‘suited only to poor quality ranching or wild life exploitation’.5 Farming in the Scheduled Areas included estate production of permanent crops for export—coffee, tea, sisal—and livestock ranching, as well as mixed farming. European mixed farming had at one time concentrated on cereals, but by 1960 efforts made after the Second World War to develop and diversify mixed farming had achieved considerable success. Farming in the African areas was overwhelmingly for the household consumption of the farmers, not for sale, and although cash-cropping was increasing in importance, perhaps only about 15 per cent of total output was marketed. In consequence, some 80 per cent of the total marketed output of agriculture came from the European areas. Livestock

2 Aid and Inequality in Kenya: British Development Assistance to Kenya and dairy produce accounted, roughly equally, for about one-quarter of total sales from the scheduled areas; coffee, tea and sisal contributed 45 per cent; cereals, mainly wheat, and to a lesser extent maize, provided another 15 per cent of the total, and the remainder of total sales was made up of a number of crops, including cotton, tobacco, pyrethrum, sugar and oilseeds. Livestock, together with a very small value of dairy produce, also accounted for about one-quarter of sales from the non-scheduled areas, and coffee, a crop which Africans had only recently been permitted to cultivate on any scale,6 accounted for another one-quarter; 14 per cent was produced by cereals, mainly maize, and a number of minor crops accounted for the remainder of sales, of which the most important was cotton. The limited extent of the shift that had taken place in African agriculture towards cash-cropping by 1960 was to an important degree caused by the administrative and legal restraints on such development, notably those on the cultivation of coffee. It was also partly the result of the limited development of transport in the African areas, and of the fact that Kenya’s highly developed and controlled structure of marketing and credit focused on the Scheduled Areas, as did the research and other agricultural services. Although the government agricultural service dealt with both African and European farmers, the statutory boards, committees and organisations concerned with marketing, mainly administered by the farmers themselves, were primarily concerned with the Scheduled Areas. It was natural for European settlers to establish European-type institutions to serve European settler interests. Until late in the colonial period, the possibility of radical developments in African agriculture and of African participation in the modern economy, except as employees of Europeans, cannot have seemed to most Europeans to be matters of any practical importance. A great divide between the European and African economies was an inevitable consequence of settlement by Europeans concerned to create and maintain their own particular standards and way of life. Substantial changes in land tenure and in the occupation of land began towards the end of the colonial period. They were a response to a belief in the inability of the African reserves under existing tenure systems and agricultural practices to accommodate the expanding African population. There were two kinds of change. One was the transfer of land from European ownership and the settlement of African farmers on it. The other was the ‘commercialisation’ of African-occupied lands by means of consolidation and adjudication and registration of title. Both kinds of change were advocated in 1955 by the East Africa Royal Commission, the analysis and recommendations of which were described as ‘Adam Smith in East Africa’ and had already become government policy. The Royal Commission proposed the abandonment of the ‘tribal approach’ to land, including the ‘racial approach to the Highlands question’, and prescribed ‘individualisation of land ownership and mobility in the transfer of land’. The commercialisation of land had been a feature of the Swynnerton Plan7 of 1954, which argued that the reform of African land tenure was a prerequisite of agricultural improvement. Consolidation, enclosure and registration of title, it was argued, would make credit obtainable for improvements and enable progressive farmers to acquire more land. The African lands would be enabled to move away from being overwhelmingly devoted to production for subsistence towards a commercial agriculture:

The Background

3

able, energetic or rich Africans will be able to acquire more land and bad or poor farmers less, creating a landed and a landless class. This is a normal step in the evolution of a country.

There had already been some individualisation of land-holding in different parts of the country, even in the absence of machinery to adjudicate and register titles, and landlessness was not unknown in traditional society. The already existing pressures on the African lands in Central Province had been raised by the forced return of Kikuyu to their home areas from other parts of the country (and from elsewhere in East Africa) under the Emergency Regulations introduced in 1952 to combat the Mau Mau rebellion. In the White Highlands a class of landless Africans had become established in the form of squatters on European farms, who provided wage-labour and were allowed to cultivate some land for their own subsistence. The political circumstances were favourable for government action. Under the Emergency Regulations many African politicians who might have used the changes in tenure as a stick to beat the government were in detention; in parts of the country people who had formerly lived scattered on their holdings had been gathered together in villages, making consolidation and redistribution of the land easier; money was available to support policies which might help to defeat and to remove the causes of the rebellion; and the authorities were able, if necessary, to exert force to make people conform. Consolidation of land and registration of title began in 1955 and by the year before independence about half the land of high potential had been consolidated and enclosed, and about half of that had been registered. Registration had, in fact, been completed in Central Province, but had not proceeded significantly elsewhere, where the pressures from the rebellion were less severe. The land tenure changes were supported by credit and extension services and by the final removal of restrictions on the growth of cash crops. It is probable that the removal of these restrictions, particularly those on the cultivation of coffee, was by far the most important cause of the increase in marketed production by small farmers. The value of produce sold from small holdings increased from £5.1m. in 1955 to £9.5m. in 1960 and to £11.6m. in 1963. The contribution of coffee to total sales increased from 6 per cent to 27 per cent,8 as the large-scale investment in coffee in the late 1950s came into full production. However, the large farms retained their dominant position in production for the market. In 1963, the large farms still accounted for as much as 78 per cent of total sales, compared with 86 per cent in 1955. The large farms also remained the source of the bulk of agricultural exports. Nevertheless, by the end of the colonial period Kenya had made decisive progress towards the establishment of a peasant cash-crop agriculture in what had been the African Trust Lands. In 1959 it was decided by the colonial Kenya government that the racial allocation of land should be abandoned, and this decision paved the way for African ownership of land in the White Highlands. Schemes for transferring European farms to Africans began to be devised in 1960. The new policy marked a sharp change of direction, and seems all the more radical in the light of the fact that the settlement of new European immigrants in the Highlands was among the most important projects of the first post-war development plan. The fundamental principle of the land transfer schemes was that farms were to be purchased from Europeans and sold to Africans. They were not schemes for expropriation—there is a view that farms were over-priced9—and were designed as much to reassure the Europeans who remained as to transfer the land of those who departed to African small holders. At

4 Aid and Inequality in Kenya: British Development Assistance to Kenya the time, the vital importance of largescale farming to the economy of Kenya was accepted doctrine,10 and the preservation of large-scale farming as a major economic sector was a fundamental consideration in the design of the Land Transfer Programme. External aid was sought for land purchase and settlement. Funds were raised from the World Bank (IBRD) and the Commonwealth Development Corporation (CDC—Colonial Development Corporation, as it then was) for the purchase of farms and their settlement at a ‘low density’ to provide annual incomes, in addition to subsistence and loan charges, of £100 and more. The following year the UK government agreed to provide funds on loan terms for land purchase, and this programme was expanded at the end of 1962 into what became known as the Million Acre Scheme. ‘High density’ settlement was planned under this scheme with the intention that settlers should be able to obtain a net annual income of between £25 and £70. By the time of independence, 236,000 hectares had been purchased and ten thousand families settled under the Million Acre Scheme, and a further thousand or so families had been settled on low-density schemes. Small-scale African farming had been firmly established in the Highlands. The influx of Africans into large farming had begun, but remained unimportant until after independence. The existence of the non-African population, both as producers and consumers, provided the initial stimulus for the development of manufacturing and processing in Kenya. The constraints on African monetary agriculture limited the possibility of a manufacturing industry to serve a largely African market. The earliest developments were in the processing of the products of European agriculture, but as the size of the market expanded, particularly during and after the Second World War, and with the increased use of tariff protection, the manufacturing sector of the economy became more diversified. It was not only the European and Asian communities of Kenya which provided a market for Kenya manufactures. Kenya had for long been associated in a common market with Tanganyika and Uganda, and the early impetus to the development of industry in Kenya resulting from its relatively large non-African population put it in a strong position to capture the markets of its common market partners. By the late 1950s Kenya had become the manufacturing centre for the whole of East Africa, and something like 20 per cent of the output of its manufacturing industry was sold to Tanganyika and Uganda. Whatever the origin of the market for Kenya manufactures, by the time of independence the African market was of great importance. A number of products had a widespread market among Africans—shoes, cigarettes, beer and grain-milling products, for instance. In the middle 1950s manufacturing and construction accounted for about 20 per cent of GDP in the monetary economy (as large, in fact, as that of agriculture and livestock production), and manufacturing alone for 13 per cent. Industry was still heavily oriented towards agricultural processing. Food, beverages and tobacco industries together produced one-half of the gross production of manufacturing industry.11 Repairing transport equipment was an important activity (11 per cent of gross production) fostered by the growth in the number of motor vehicles and by the location in Kenya of the railways and harbours and airways head-quarters for the whole of East Africa. The ownership of industry was divided between Kenyan Europeans (grain milling, dairy produce, pig products, sisal products, canning), East African Asians (sugar milling, bakery products), and what are now known as ‘multinationals’ (tobacco, footwear, pharmaceuticals, cement, paints, soap), but with expansion and a scarcity of capital in the locally-owned firms the participation of the multinationals was increasing.

The Background

5

The construction industry had received a great stimulus during the war. The stimulus continued into the post-war years with the demands of development plans, which were heavily oriented towards public works, continued military expenditure, construction arising from the Emergency, and the expansion of facilities for the East African common services. Much of the industry was in the hands of Kenyan Asian firms, but international construction companies were becoming of increasing importance by the time of independence. Wholesale and retail trade accounted for 17 per cent of gross domestic product recorded in the monetary economy in 1960. Large-scale domestic commerce in exporting, importing and in the distribution of domestic production was divided between European firms, some of them branches of international companies, and Asians. Asians were widely engaged also in small-scale retailing, and they were to be found in the smallest and remotest centres. Africans had hardly begun to enter large-scale commerce and their trading activities were carried on in a very small way and largely in the rural areas. The existence of the non-African communities, particularly the Europeans, was responsible for the development of technical, financial and government services to a much greater extent than might have been expected from the low average level of income in the population as a whole. European residents demanded public and private services appropriate to their own level of income. European agriculture was associated with a highly developed structure of marketing and advisory services. The position of Kenya at the centre of the common arrangements in East Africa substantially increased the activities of government. These factors account for the remarkable development of the city of Nairobi. Its attractions were cumulative. Originally a railway construction encampment, it became the seat of government for Kenya and for those activities (notably in transport and communications) administered in common for the whole of East Africa. It developed as the centre for the provision of services to agriculture and manufacturing, and as the location of most manufacturing activity, and became the natural location for the headquarters of international firms entering the East African market. By the time of independence it had long been established as a centre relevant to a level of income and way of life totally different from that of the vast majority of the population of Kenya. The part played by Africans—except as wage-earners—in manufacturing, construction and trade was of little importance. Of course, the smallest enterprises escape the statistical net (surveys of manufacturing, for instance, were confined to firms with five or more employees) and so understate the role of African business activity. But African participation and initiative in any but the smallest manufacturing, construction and trading activities remained negligible until the end of the colonial period. The inheritance of independent Kenya was, therefore, an economy the modern sector of which had been fashioned largely in response to the existence of a non-African population. Change had already gone some way in agriculture, although not far enough to destroy the central position of the former White Highlands. In other sectors of the economy change had hardly begun. In the few years preceding independence the economic growth of the post-war period was halted by the uncertainties of the future. There was a sharp downturn in construction; employment declined; capital flowed out; the government’s finances deteriorated and Britain had to provide grants-in-aid. In the event, many of the fears proved groundless.12

6 Aid and Inequality in Kenya: British Development Assistance to Kenya Jomo Kenyatta had declared that: The Government of an independent Kenya will not be a gangster government. Those who have been panicky…can now rest assured that the future African government…will not deprive them of their property or rights of ownership. We will encourage investors…to come to Kenya…to bring prosperity to this country.13

When economic growth began again it was firmly within the structure established in the colonial period. The years since independence have seen many changes, but equally noteworthy has been the continuity with the past.

NOTES 1. In writing this chapter we had the advantage of seeing in advance the contribution of Michael McWilliam, ‘The Kenya Economy, 1945 to 1963’, to Volume 3 of the Oxford History of East Africa, which has subsequently been published. London, Chatto and Windus, 1968. 3. The European population reached its peak in 1960. It had increased from less than 30,000 in 1945, and by 1961 it had already fallen to 59,000. In valuing European farms under the Land Transfer Programme (see Chapter Five) land prices in 1959 were taken as a basis. 4. See Development Plan 1966–1970, Government Printer, Nairobi, 1966, p. 29. 5. Kenya African Agricultural Sample Census, 1960/61, Part I, pp. 1–2. In 1960 there were only 3,609 holdings in the Scheduled Areas. 278,000 Africans were employed on these holdings as wage-earners, as were 1,905 Europeans and 866 Asians. The vast majority of Kenya’s nearly 8m. Africans lived in the non-scheduled areas. 6. In 1954 coffee accounted for less than 4 per cent of the value of sales of produce from small farms. 7. A Plan to Intensify the Development of African Agriculture in Kenya, Nairobi, Government Printer, 1955. 8. The increase in coffee production was the major reason for the growth in the value of small holders’ total sales: Coffee production in non-scheduled areas (‘000 tons) 1955/56 0.8 1960/61 7.9 1963/64 15.3 9. It is said that some farmers sold out, bought another farm and sold out again, which suggests that they got a good bargain. 10. The view is still held today, though it may receive a jolt from the need to provide funds for the rehabilitation of certain large farms. A loan has been provided for this purpose by the IBRD. (See The Standard, 19 August 1975.) 11. Survey of Industrial Production, 1957. 7. 12. Though not those of the Asian population, particularly of Asians who did not acquire Kenya citizenship. 13. Reprinted in Jomo Kenyatta, Suffering without Bitterness, Nairobi, East African Publishing House, 1968, p. 157.

2 AIMS AND ACHIEVEMENTS SINCE INDEPENDENCE

The economic policies of independent Kenya have been essentially pragmatic, unencumbered by much in the way of overt ideology. To foster the entrepreneurial society and to increase the role of Africans within it is, perhaps, the most summary form in which the general aims of policy can be expressed. Formally, the policy is labelled African Socialism. Sessional Paper No. 10 of 1965, On African Socialism and its Application to Planning in Kenya defined the social and economic objectives of the Kenya government and outlined the policies through which these objectives were to be pursued. The policies set out there have been repeated and elaborated in the various Development Plans and in other documents. African Socialism is not a programme for public ownership on ideological grounds. Nationalisation, the Sessional Paper declared, will be used only where the national security is threatened, higher social benefits can be obtained, or productive resources are seriously and clearly being misused, when other means of control are ineffective and financial resources permit, or where a service is vital to the people and must be provided by the Government as a part of its responsibility to the nation.1

It has been suggested, in fact, that Kenya is committed not to Socialism but to a capitalist mode of production, and that African Socialism is a ‘verbal pretence’. The idea of African Socialism is best seen as that of softening the impact of the market economy by bringing into play the ‘mutual social responsibility’ which operated in traditional African society. However, there is little evidence that there has been any serious attempt to elaborate this line of thought or to give it practical application. Certainly, the policy of African Socialism has involved no revolutionary break with the past; great changes there have been, but the thread of continuity has been strong.

The Agricultural Sector It is in agriculture that change has been the most dramatic, while continuity has, nevertheless, been marked. The changes in the ownership of land and in the nature of land tenure during the decade after independence resulted from the application and intensification of the policies adopted before independence had been achieved. Agricultural marketing arrangements and price-fixing policies betray the strong influence of the past. The policies initiated in 1960 under which European farms were to be transferred to African farmers continued to operate. There was to be no expropriation and no free land.2 The programme for the purchase of European mixed farms suitable for subdivision into small holdings was the subject of a series of aid agreements between the Kenya and

8 Aid and Inequality in Kenya: British Development Assistance to Kenya British governments. The transfer of land under the Million Acre Scheme and the lowdensity scheme, begun before independence, was completed by 1970. A minor additional programme (the so-called Harambee scheme) of relatively low-density settlement was carried out on land transferred under the 1965 and subsequent aid agreements. A new form of settlement was important for a time when the Kenya government recognised the need to make some provision for squatters. In the Development Plan for 1970–74 squatter settlement, renamed Haraka, had a substantial place. The intention to establish further Haraka settlements was abandoned in 1971, though it is proposed in the Development Plan for 1974–78 to put more money into the existing settlements. They had been designed to yield little more than subsistence for the settlers, who had been provided with few facilities and resources. No additional land was purchased for Hamka settlements as the squatters were mainly on farms which had been abandoned or from which the owners had been evicted for neglecting them.3 The Haraka schemes were seen by the government as a political necessity rather than as a desirable form of economic development, though there is a view that they soon became indistinguishable from high-density settlements under the more elaborate and costly schemes. There have been problems with the settlement schemes, particularly in obtaining loan repayments from settlers. Little more than half the amount due in loan repayments has been collected, and most settlers are in arrears. To alleviate the problem it is proposed to offer settlers the option of converting their freehold title into a leasehold title from the government.4 During the process of settlement under the Million Acre and Harambee schemes there was some disruption of production, but production recovered and considerably greater outputs per acre are being obtained from the settlement areas than before settlement. There remains a need to improve credit facilities, extension services, and roads and water supplies, despite the fact that settlers are much better supplied with these services than small farmers in other areas, In 1971, in fact, it was decided to abandon the old form of high-density settlement, undertaken by the subdivision of large farms. Its place was taken by co-operative or Shirika settlement in which it is hoped that the avoidance of subdivision into individual holdings will bring greater efficiency. However, it cannot be assumed that there will no further settlement on subdivided farms because the Development Plan 1974–78 declares that ‘most farm products can be produced very successfully on small-scale farms. In the long run, therefore, a considerable amount of land used for large-scale farms will be subdivided.’5 Not all the transfer of land from Europeans to Africans has taken place through official settlement schemes. There has been substantial direct purchase by African individuals, partnerships and companies. The funds for these purchases have come from private institutions, from the Agricultural Finance Corporation, and to a significant extent from private accumulations of wealth, often acquired outside agriculture. The purchasers have often had their primary interest elsewhere, in business, politics, and the civil service.6 Some of these new large-scale farmers lacked both experience and capital, and a programme of rehabilitation is being embarked upon, with the aid of funds from the World Bank. The Development Plan says that ‘where appropriate the Ministry will encourage subdivision of these farms into smaller more manageable units’.7 The statistics confirm the success of the programmes for the transfer of land from Europeans to Africans. Some 3.5m. acres of the Scheduled Areas were under mixed farming before land

Aims and Achievements Since Independence 9 transfer got under way; rather more than one-third of this area has been transferred under settlement schemes. Something over two million acres of mixed farming land remain in 1,540 large farms (1972 figures). Of these large farms, 1,234 embracing about half of the land area of the remaining large farms, are in African ownership, either individuals, partnerships or companies; a few more, where special circumstances prevailed, have been taken into state ownership.8 More than two-thirds of the European mixed farming area had, therefore, passed into African hands by the end of the first decade of independence. There had in addition been some African purchases of estates and ranches. There remained outside African ownership nearly 300 mixed farms, covering an area of about one million acres, large commercial ranches, and some 1,500 tea, coffee and sugar estates, mostly owned by companies. Commercial agriculture by non-African companies has not proved to be incompatible with the development of African agriculture, however, and the two have been linked by associating small holders with company estates, in particular where access to a factory is essential to producers, as in the cultivation of sugar-cane.9 Small-holder and estate production have also been associated, though less successfully and perhaps with less concern for success, in the production of pineapples for canning. In addition to the settlement schemes, the other leg of the land reform programme, the adjudication, consolidation, and registration of land so as to replace communal by individual tenure in the former African areas, proceeded rapidly after independence. In 1963 only 5 per cent of the registrable land had been adjudicated; at the end of 1973 adjudication and registration had been completed on 20 per cent of the land and the process was under way on a further 20 per cent. Indeed, there is a view that the procedures are being applied indiscriminately, so as to include areas where they are inappropriate, as in areas devoted to migratory pastoralism. No simple comparison is possible between the quantitative importance of settlement on the one hand, and development of the former African Trust Lands on the other, by which the significance of settlement can be judged. In terms of population, of course, the contribution of settlement seems tiny. It has been estimated that there are about 1.7m. rural households and 1.2m. settled agricultural holdings in Kenya. About 60,000 families have been settled on former European farms. However, despite the difficulty of statistical comparisons, it is evident that development of the traditional African lands is fundamental to any progress. Indeed, by 1965 there was official disillusionment with settlement, and Sessional Paper No. 10 declared, stating a view that had already received some consideration, that: We have to consider what emphasis should be given in future to settlement as against development in African areas. The same money spent on land consolidation, survey, registration and development in the African areas would increase productivity and output on four to six times as many acres and benefit four to six times as many Africans. It therefore follows that if our resources must be used to achieve maximum growth we must give priority in the future to development in the former African areas.10

Settlement can be a safety-valve, relieving the pressures exerted by the landless, but it is doubtful if it can take place on a sufficient scale to be more than a safety-valve, and to make a serious contribution towards solving the problem of landlessness. Nevertheless, the political necessity of further settlement as an obvious aid to the landless remains, and

10 Aid and Inequality in Kenya: British Development Assistance to Kenya indeed is strengthened by the ‘individualisation’ of land in place of customary tenure in the former African lands. Although the land reform in the African areas has been a major change, it would be wrong to see it as the fundamental cause of progress in smallscale agriculture. Progress has not been confined to areas where the land reform has been carried furthest. In any case, the fragmentation of holdings was not a serious problem in many areas, so that substantial benefits could not be expected from consolidation, and in areas where consolidation was carried out at an early date an informal fragmentation has subsequently taken place. Much of the benefit of the tenure changes was expected to arise from their effect on the availability of credit to the farmers. ‘The need to develop and invest requires credit,’ Sessional Paper No. 10 argued, ‘and a credit economy rests heavily on a system of land titles and registration. The ownership of land must, therefore, be made more definite and explicit.’11 The political difficulty of dispossessing small farmers12 must cast doubt on the value of title to land as a security for credit and, in fact, there has not been a large influx of capital into small-scale farming. The Agricultural Finance Corporation provides credit to virtually all large-scale farmers but to little more than one per cent of small farmers, and only about one-fifth of its total lending is to the small-farm sector.13 The progress in smallscale agriculture has occurred through the expansion of cash crops, particularly coffee, which followed the relaxation and eventual abolition of the restrictions that had existed on their cultivation by African farmers. That there has been greatly increased participation of Africans in monetary agriculture cannot be doubted. The output of the large-farm sector has increased and the African share has become substantial.14 The marketed output from small farms has increased fivefold in value in the decade since independence, and the share of small farms in marketed agricultural output has risen from a quarter to a half of the total. It would not be universally accepted that the increase in African participation and African money income in the aggregate is sufficient evidence on which to pronounce the policy towards agriculture and land tenure a success. It can be argued, as it was argued in the report of the ILO mission in 1972,15 that the changes have perpetuated and, indeed, increased the inequalities in the distribution of income and wealth which were characteristic of colonial Kenya, merely eliminating—or at least reducing—the racial dimension in the inequality. The broad distribution of population has not been fundamentally affected by the removal of the racial barriers to land ownership and by settlement on European farms. The African population is still densely settled in what were the African reserves, and the ratio of population to land is still much higher in those areas than in the former Scheduled Areas. When the agriculture of the White Highlands was the overwhelmingly important component of monetary agriculture, it is not surprising that the arrangements for marketing, credit, price-fixing, extension services and transport were focused on its needs. It has been argued that these services are still focused on the large farms, and will continue to be agents of inequality until there is a major reorientation of their activities. The commercialisation of land might have been expected—as it clearly was expected in the Swynnerton Plan—to lead to the accumulation of large holdings by some farmers and the loss of land by others through the operation of the market, once the original consolidation and adjudication had been made. There is no statistical evidence by which it can be determined whether or not such a process is taking place. It is reasonable to

Aims and Achievements Since Independence 11 expect that the growth of population would lead to an increase in the number of landless even without the changes in tenure in the African areas; indeed, it had already done so before land adjudication and registration were far advanced. Although the extent to which the operation of the market has displaced people from the land is not known, it is clear that a trend towards differentiation within the African population accelerated in the years following independence. This is clear, if for no other reason, from the fact that some large African farmers have become established by succeeding, in effect, in stepping into the shoes of departing Europeans, and, as was remarked above, these farmers often have outside interests in business, politics, or the civil service. Among the small holders, the impressionistic conclusion of the ILO Mission was that about 225,000, or one-fifth of all small holders, have been able to benefit sufficiently from settlement and land registration to achieve rapidly rising incomes; they have become significant employers of labour so as to give themselves more leisure and more time to engage in business and salaried employment. The report of the ILO Mission suggests that a further quarter million small holders have been able to commercialise their activities enough to achieve incomes of between £60 and £110 a year, but have been unable to do more because they cultivate too little land or too poor land or lack knowledge or capital. The majority of farming families—some 620,000—have, according to the ILO Mission, benefited from the developments since independence only marginally, if at all. These families have an income from farming of less than £60 p.a. Some have income from other sources—remittances from relatives working in the towns, for example—but some can have no additional income; some are squatters on land they do not own, and some are landless. But this is only one view, and other observers would say that the great majority of Kenya’s rural population has become significantly better-off since independence. On this view, the growth of economic differentiation and inequality within the African population has not proved incompatible with an improvement in the absolute standards of the mass of rural families. Whatever the truth about the absolute living standards of the population, the growth of economic differentiation should be no surprise. The policies enunciated in the Swynnerton Plan and by the East Africa Royal Commission were designed to give the ‘progressive farmer’ scope to exercise his talents, and to help those most able to help themselves, rather than those most in need. The application of these policies is a major example of the continuity which has been a feature of the post-independence years. It is easy to believe that they have fostered the growth of inequality of income and wealth within the African population. They do not necessarily stand condemned by that fact; greater inequality may be accompanied by, and indeed be a condition of, a general rise in living standards. It would be a more serious reflection on the success of the policies if the improvement had been confined to a few, and if the position of the masses had not improved or had worsened. But even if that had been the case, it would not follow that the policies which offered opportunities to be grasped by the successful were themselves responsible for the failure of others to advance. Given the pressures on the land that had already built up,16 and the frightening rate of population growth, a failure to change land tenure in the African areas and to regulate and plan the influx into the Highlands may simply have made more difficult the path to improvement for all. Certainly it should be the aim to extend the benefits of economic growth more widely,

12 Aid and Inequality in Kenya: British Development Assistance to Kenya and this is official policy. The Development Plan for 1974–78 proclaims a fundamental goal to be ‘an improvement in the distribution of the national income’, but this is not a change of heart—it was also the declared aim in the Plan for 1966–70.17 If this aim is to be achieved, one requirement is a more equitable distribution of agricultural services. According to the 1974–78 Development Plan, an important instrument for greater equality is to be the subdivision of a considerable amount of land still farmed on a large scale. Where large-scale farming is necessary for efficiency an important role is proposed for the Agricultural Development Corporation, and ‘Where private farmland is retained as large-scale units, the Government will encourage the occupation of these farms by groups of people or co-operative societies rather than by a few privileged individuals’.18 That would involve a very considerable shift in the trend of development, in the face of vested interests, and one wonders about its compatibility with the possibilities for private accumulation resulting from the commercialisation of land. The process of differentiation within the African population has not only a personal, but also a tribal and—to a large degree the same thing—a regional dimension. The opportunities opened by the changes in land policy and by the removal of restrictions on cash-crop production were more eagerly grasped in some areas than in others, and by some peoples than by others. The distribution of high-potential land favoured some and not others, and the widely recognised relative expansion in the wealth and strength of the Kikuyu is in part the result of the natural advantage of the land in Central Province. Nyanza Province and Central Province had roughly the same value of marketed agricultural output until 1957, after which date Central Province drew rapidly ahead, so that at the time of independence its marketed output from small farms was roughly twice that of Nyanza Province. Central Province has retained its lead in independent Kenya.19 The advance of Central Province was based to a large degree on coffee production, which expanded rapidly once the restrictions on its cultivation had disappeared. The exploitation of the opportunities offered by the removal of these restrictions—in coffee and other products—was facilitated by the high level of infrastructure and of extension services that had been established in the Province as a consequence of the Emergency. In contrast to the progress of Central Province, money income may even have declined in some areas during the 1960s, as in South Nyanza and Busia, where disease and poor quality led to the uprooting of robusta coffee, and in some marginal and drought-stricken areas where production has not kept pace with the growth of population. That the growth of regional and tribal inequality results in part from such natural and historical circumstances does not, however, diminish the political tensions it creates. And it is therefore all the more important that policy-makers—including aid donors—should be alive to the dangers of fostering the cumulative growth of inequality.

Manufacturing The process of import substitution through the establishment of domestic manufacturing gained speed after independence. It was fostered by policies to attract foreign capital and at the same time to promote a greater participation by Africans in ownership and management. During the 1950s there had begun to be a sizeable inflow of Kenyan-Asian (and other East African)-owned capital from trade into manufacturing. In the years after independence capital from abroad became relatively more important.

Aims and Achievements Since Independence 13 The first step to stimulate the inflow of capital was to create an appropriate ‘investment climate’. In 1961 and 1962 foreigners had been withdrawing capital, and a reversal of this flow required a reversal of the expectations that had stimulated the outflow. Even before independence it had become clear that—in the words of Jomo Kenyatta already quoted—the British were not to be succeeded by ‘a gangster Government’ and that the new government would ‘encourage investors in various projects to come to Kenya’; in 1963 capital was already flowing in again. With independence, new legislation reinforced and codified the favourable attitude towards foreign capital. Under the Foreign Investment Protection Act (No. 35 of 1964) an undertaking can be granted an Approved Enterprise Certificate which entitles the holder to transfer profits and capital out of Kenya. A constraint on foreign enterprises as a channel for the export of local capital is provided by restrictions on local borrowing, Policies concerning taxation, protection, and the provision of finance have been designed to encourage private investment, including investment from abroad. Company profits are taxable at a rate of 45 per cent, though the rate is 52½ per cent for branches of foreign companies. There is an initial investment allowance of 20 per cent of the cost of industrial building, plant and machinery. Wear-and-tear deductions are allowed at the rate of 2½ per cent on buildings, 12½ per cent on plant and machinery, and 25 per cent on certain vehicles. The government advertises that the effect of all allowances is that ‘an industrialist can write off 120 per cent of his investment against taxable income over a period of a few years’.20 However, the tax rates are relatively high, and the allowances not conspicuously generous. It is to be noted that there is no provision for tax holidays. Foreign capital has been encouraged, therefore, rather by the attempt to establish a reputation for stability and order than by the offer of excessively generous financial terms. Protection of industry is given by the common external tariff of the East African Community.21 It is also given by import licensing and quantitative restrictions. In addition, a system of import-duty refunds and remissions operates to assist producers to compete with imports and in foreign markets, though it is intended to dispense with these concessions in the course of a general revision of the tariff. A ‘Manufactures Export Compensation Scheme’ provides a subsidy of 10 per cent of the f.o.b. value or of the foreign currency proceeds of exported local manufactures. After independence tariff and other instruments became increasingly protective, though ‘not as part of a coherent strategy, but piece by piece in response to pressures and temptations of various kinds’.22 Despite the measures designed to attract capital into manufacturing, the growth of output did not accelerate notably in the decade after independence. The official quantity index of manufacturing output increased from 60 to 120 between 1954 and 1964 (1960=100); between 1964 and 1974 it rose from 70 to 149 (1969=100). In value terms, at current prices, output rose from K£34m. in 1964 to K£105m. in 1973, and increased its contribution to GDP in the monetary economy from 14 per cent to 18 per cent, though measured at constant 1964 prices its contribution increased only to 15 per cent.23 Nevertheless, the growth was substantial. Value added in manufacturing increased at an annual rate of 7.9 per cent between 1964 and 1972, while employment in manufacturing rose over the same period at a rate of 6.6 per cent.24 The number of manufacturing establishments employing five or more persons increased from 676 in 1963 to 1,298 in 1971, and with that increase industry became more diversified. The market remained primarily within Kenya, but its export component became less tied to East African markets.

14 Aid and Inequality in Kenya: British Development Assistance to Kenya The Kenya market changed less than might have been expected in the years following independence. The non-African population remained large, though its composition altered. The European population had numbered 31,000 in 1948, and had increased to a peak of 61,000 in 1960. Thereafter it declined, but in 1969, six years after independence, it was still as large as 41,000.25 Employment of Europeans (another indicator of the size of the market provided by the European community) had been 23,000 at its maximum in 1960, and had fallen to 18,200 by 1963. In 1972 it was still more than 13,000, though it declined to 9,500 iii the following year. The Asian population continued to increase to a maximum of 192,000 in 1967, and had fallen to 139,000 in 1969. But the demand for manufactures from the high-income non-African population had come to be supplemented by a growing African demand. As the Development Plan predicted: In the Kenya domestic market the main change will be the shift that is taking place from production for a market largely composed of non-citizens to a market where these are being replaced or added to by a rising African middle-class.

The import substitution of non-durable consumer goods had been carried a long way by the end of the first decade of independence. Production of intermediate products and capital goods had already developed and the greatest scope for future growth was in these sectors. One difficulty was the greater limitation imposed by the size of the market in these industries than in many consumer goods industries. Despite the preferential access to the markets of Tanzania and Uganda under the terms of the 1967 Treaty for East African Cooperation,26 and despite some successes in other export markets,27 the smallness of the Kenya market seemed likely to be a greater constraint on industrial development in the future than in the past. It has been argued that the reliance of industrialisation in Kenya on foreign capital, and particularly on international firms and imported technology, has resulted in the establishment of an industrial structure in which inappropriate goods are produced by inappropriate methods.28 Multinational firms, it is said, have adopted the capital-intensive methods they use in the developed countries to manufacture the products they manufacture in those countries. Industrialisation has, in consequence, become not just a matter of import substitution but of ‘import reproduction’ in which the identical goods, previously imported, are produced domestically. These products are designed for the markets of the developed countries and are often not relevant to the needs of a poor society, but a demand for them among the better-off is created by advertising. These products, it is claimed, are socially divisive. They are the medium for a ‘transfer of tastes’ from the developed countries to the better-off in the poor countries, whose pattern of consumption is made to conform to that of the developed countries, and to diverge from that of the mass of the population in their own country. The activities of the international firms increase the inequality of income because the firms pay their employees (labour costs are not an important component of total costs when capital-intensive methods are used), particularly in management, at rates greatly above the level prevailing in the economy generally. Methods of production are not designed to use locally available materials, so that the demand for imports grows despite import substitution. Products which are not themselves inappropriate to the needs of the mass of the population are often produced—because of the need to maintain the quality of the internationally-marketed brand—to an inappropriately high standard of quality, which

Aims and Achievements Since Independence 15 can only be achieved at a price which confines demand to the rich. Finally, the argument runs, there are strict limits on the growth possible in an industrialisation pattern directed at the needs of the better-off, because in Kenya the market is too small for the efficient production of many goods demanded by high-income consumers. There is a good deal of truth in this analysis of the nature of industrialisation in Kenya. After all, its initial proposition is simply that production is established to serve a market that exists, and when the great majority of the population has very small money incomes, that market will be provided by the better-off minority. That is, of course, why import substitution is a basis for industrialisation, to satisfy by domestic production an existing demand previously satisfied by imports. However, for all its truth, this criticism of the nature of Kenyan industrialisation can easily be carried too far. It is not a fact that Kenyan manufacturing is concerned solely with the demands of the rich. In reality, a wide range of goods is produced for consumption by people who cannot in any absolute sense be considered wealthy. The statistics of industrial production do not provide a sufficiently fine classification of products for their nature to be clearly identified, but they do provide some indication. In the first place, the Census of Industrial Production showed that, in 1963, 16 per cent of total production by private industry was of food, and another 15 per cent of drink and tobacco. The largest output among food industries was from grain-milling, which accounted for nearly one-third of their total output. Clothing, textiles and footwear contributed 8 per cent of industrial production.29 These industries did not produce only for the rich: maize meal, cheap cigarettes, plastic sandals and cooking fat are among their products. It is true that even these products are not available to the very poor. But the very poor, in the rural areas particularly, provide virtually no market for manufactures. Even in the urban areas the purchases of the poor are very restricted. In the budget used for computing the low-income urban price index,30 three-quarters of expenditure on goods is accounted for by food, and another 6 per cent by drink and tobacco. With 8 per cent on clothing and footwear, and 4 per cent on fuel, there is very little left to constitute a demand for a wide range of manufactures. It must be accepted as a fact that when the money incomes of most people are extremely small, manufacturing will inevitably be largely directed at a minority market. Policies might reasonably aim at a widening of the market, but hardly at production only for majority markets. A more equal distribution of money income would doubtless result in a different pattern of demand—an equal distribution of the present aggregate of consumer’s expenditure might eliminate altogether the demand for all but a few manufactures. Growth must depend on a further expansion of aggregate money income. Changes in industrial structure and in the pattern of demand will come about only in response to an expansion of rural incomes, and this may be achieved by the further development of small-scale agriculture. The policies of multinational companies can have no more than a marginal effect in this respect. The ‘transfer of tastes’ is the familiar ‘demonstration effect’, the disadvantages of which have long been recognised. It is not clear that the disadvantages of the ‘rising expectations’ created by the demonstration of higher standards of consumption inevitably outweigh the beneficial incentive effects, and it is even less clear that policies to deal with the disadvantages are available in which the cure is not worse than the disease. Kenya’s development policy, after all, aims at least in some respects to emulate the rich countries,

16 Aid and Inequality in Kenya: British Development Assistance to Kenya and the move to the consumption of superior goods is part of the process of development. Other patterns of change are conceivable; in Kenya they are ‘non-agenda’. And just as it is easy to exaggerate the extent to which Kenyan manufacturing is concerned to serve the rich, so is it easy to exaggerate the disadvantages of the transfer of tastes. It might readily be agreed that a transfer of tastes from maize meal to corn flakes would not be an important ingredient of development. But could the same be said for a shift from thatch to iron for roofing, or the stimulation of a demand for plastic sandals? It has, in fact, been argued that local materials are potentially available that would provide improved roofing without the high import content and high capital intensity of the manufacture of corrugated iron, and that plastic sandals are an ‘inappropriate’ product because the alternative exists of sandals made from old motor tyres. If consumers believed themselves to be as well-off or better-off using some other roofing material, and if they preferred motor-tyre sandals, then it would be reasonable to declare corrugated iron and plastic sandals to be inappropriate products. There is a case here for research into improvements of the alternative products and for the advertising of their benefits. But unless such activity persuades consumers—and with the consumers of these two products, and of many others, we are not dealing with a small, wealthy minority—that the alternative products are to be preferred, it cannot reasonably be maintained that the transfer of tastes is detrimental—not to consumers’ welfare, at any rate. It is not as if the availability of corrugated iron and of plastic sandals precludes the availability of alternative products. By definition, ‘appropriate’ products are not capitalintensive, so that there are no significant economies of scale in their manufacture. The demand for plastic sandals does not, therefore, raise the cost of motor-tyre sandals by limiting the size of their market. They are available for those who wish to buy them, and their availability is not affected by the availability of plastic sandals. It is simply that, for many people, for many uses, plastic sandals are a superior product for which they are prepared to pay. The buyer’s standard of living is higher with plastic sandals than without them, Of course, distortion of factor prices may make plastic sandals too cheap relative to motor-tyre sandals, but that is another question. It does not alter the fact that higher standards of living are often associated with the consumption of so-called ‘inappropriate’ products, the taste for which has been created by their being made available by modemsector manufacturers and by imports. Doubtless there can be a transfer of tastes to products the social utility of which (one’s judgement may suggest) is small.31 This is hardly a problem confined to poor countries, and it provides an argument for intervention (if one’s judgement is shared by those who make policy and those for whom it is made!). It is not an easy matter to know how to intervene. It is not a question of regulating the behaviour of domestic producers. The demonstration effect is exercised by imports,32 by the cinema, by tourists and returning travellers. To insulate the population from the disease would not be easy, and the draconian measures that would be required do not seem to be practical politics in present-day Kenya, even where it is judged that the welfare loss would be outweighed by other benefits. There is a role for fiscal policy and for research and promotion of alternative products. It can also be argued that there is a case for a tougher and more sceptical attitude on the part of the Kenya government to overseas manufacturers than has, perhaps, been the case in the past, and that ‘stricter scrutiny of investment proposals and a more critical appraisal of applications for protection are necessary to achieve appropriate import substitution’.33 In fact, the intention

Aims and Achievements Since Independence 17 to modify policies towards industrialisation had already been announced in 1972. The government had become more aware of the disadvantages of a rather indiscriminate and excessive encouragement of industrialisation through the attraction of foreign capital, and of the costs of highly protected import substitution, which are familiar from the experience of other countries. The 1972 budget speech gave: due warning to our manufacturers that the Government proposes to lower the protective barriers around domestic industry and require our manufacturers to compete on more equal terms with producers from overseas.34

Tariff changes require the agreement of Tanzania and Uganda, the other members of the East African Community, but subject to that agreement the Kenya government’s policy, as set out in the Development Plan for 1974–78, is to establish a more uniform tariff by raising the lowest rates, reducing the highest rates, introducing duties on items, mainly capital and intermediate goods, that now enter free, and eliminating refunds and remissions of duty.35 These changes in the tariff would reduce the opportunity for very high rates of effective protection of domestic processing to emerge. It was said that the changes were intended ‘to render Kenyan industry more competitive, more employment creating, and more development oriented’, and that it was ‘envisaged that these measures will have a substantial impact upon the composition of industries which will be established over the next Plan period’.36 However, in response to the economic situation which arose out of the increase in oil prices and other adverse developments, it was decided to impose tighter quantitative controls on imports. It was announced in 1974 that ‘imports of luxury consumer goods will be banned, and other items now subject to import quota will be cut back in order to reduce the overall import bill’.37 These quantitative restrictions could result in a very high degree of protection for domestic manufactures and encourage the domestic production of luxury goods the import of which is prohibited. Price controls exist which might limit the effect of the increase in protection accorded by quantitative restrictions, and the government is aware of the encouragement to the domestic production of inessentials given by import restrictions.38 It is also intended to improve project appraisal and to impose a ‘stricter scrutiny of investment proposals and a more critical appraisal of applications for protection’.39 ‘The Government will specify where and under what conditions foreign investment will be welcomed. Investment in industries with an export potential will be given priority. Technological and organisational expertise, with long-run benefits to Kenya, and improvement in ability to compete internationally will also be considerations.’40

African Participation The desirability of and, indeed, the likelihood of the adoption of policies adversely affecting the interests of the foreign-owned sector of the economy is in part determined by the extent to which Kenyan participation in and association with the foreign-owned sector has already become substantial. There is no doubt that industrial development since independence has been predominantly financed from abroad. The government has taken an equity interest in a number of enterprises (including oil-refining and banking) and has provided loan finance through different institutions. There has been some private participation in the equity of

18 Aid and Inequality in Kenya: British Development Assistance to Kenya overseas firms, as well as a growth of holdings by Africans in non-African-owned local companies, But so far as ownership is concerned, the intention of African Socialism in Kenya that ‘a large share of the planned new expansion is African owned and managed’ has not been fulfilled, although a few Africans have very large stakes in economic enterprise of all kinds. It is a different matter with respect to management. Restrictions on the issue of work permits for non-citizens, and a realisation by international firms of the advantages of localisation have resulted in a considerable degree of Africanisation of management, particularly in large companies. There has also been an inflow of Africans into local directorships, though the degree of ultimate control exercised by Kenyans may not be great, because of the control exercised by overseas headquarters. But the process of Kenyanisation has gone far enough to ensure that the interests of the foreign companies and of many Kenyans are not in conflict. Kenyanisation—in practice Africanisation41—has, in fact, gone far in all spheres of activity. Within a year or two of independence most of the heads of government departments were Africans. But there was no desire to Africanise regardless of cost and efficiency. Many senior executive positions continued and continue to be occupied by expatriates supported by the British government through OSAS (Overseas Service Aid Scheme) and in the University and other institutions there are many expatriates supported by BESS (British Expatriate Supplementation Scheme). There has been an inflow of technical assistance personnel into advisory positions (though in practice some carry out executive functions), and there is a large contingent of expatriate teachers. Expatriates, therefore, continue to be important, particularly in some activities. Nevertheless, citizens of Kenya occupied threequarters of all the posts in high-and middle-level occupations in 1972, compared with 59 per cent five years earlier.42 The Kenyanisation of trade was pursued through a Trade Licensing Act, under which progressively more areas of trade have been confined to Kenya citizens. Manufacturers have been encouraged or required to distribute their products through citizen traders, and the Kenya National Trading Corporation (KNTC), which has a monopoly in the distribution of certain products, has fostered Kenyanisation by awarding agencies to citizens. The Kenyanisation of existing businesses owned by non-citizens has been pursued, almost to completion for small businesses, through compulsory sales to Kenya citizens. Credit has been made available to finance Kenyans in the purchase and operation of the businesses. Trade had been largely in the hands of Asians, some of whom became Kenyan citizens, but many did not, either because they were not qualified for citizenship, did not apply for it, or were not granted it. The Kenyanisation of trade has therefore been essentially a process of Africanisation.43 By the middle of 1975 it had gone very far, particularly in the rural areas and small towns and was being pushed ahead in Nairobi and Mombasa. Advantage has not been taken in the past of the possibilities for increasing African participation in the money economy through the encouragement of what has become known as the ‘informal sector’. This term refers to a wide range of economic activities which are outside the scope both of normal statistical assessment and of government recognition. Those who operate in the informal sector do so without licences—in an economy where so many activities require licensing—without what would officially be recognised as proper premises, and doubtless infringing certain standards of safety and health and, perhaps on

Aims and Achievements Since Independence 19 occasions, of business probity and reliability. But the informal sector is not an unproductive and parasitic growth, though some of its activities may be so. The ILO Mission took the view that the informal sector employs: a variety of carpenters, masons, tailors and other tradesmen, as well as cooks and taxidrivers, offering virtually the full range of basic skills needed to provide goods and services for a large though often poor section of the population…[It] provides income-earning opportunities for a large number of people. Though it is often regarded as unproductive and stagnant, we see it as providing a wide range of low-cost, labour-intensive, competitive goods and services.44

Far from receiving encouragement and assistance from the authorities, the informal sector is more the subject of neglect and even of positive harassment. Although the ILO Mission perhaps took a somewhat romantic view of the character of the informal sector, it can without doubt be a source of goods and services of a kind and quality in demand by the poorer sections of the population. The official attitude to informal sector activities has reflected the implicit assumption, which was common in the colonial administration, that only ‘modern’, large-scale activities deserved attention. The assumption is nicely illustrated by the story45 of the official in Uganda who, when asked about Kampala’s industrial area, replied, ‘There are no industries there—only a lot of furniture works, bakeries, maize mills and soda water factories.’ The attitude also reflects an attachment to ‘standards’ and ‘quality’ appropriate to a wealthy society. The relevance of this consideration to the structure of manufacturing industry has already been discussed. It is found in the regulation of public transport, where safety requirements are the ostensible reason for restricting the provision to expensive, relatively high-quality services. The harassment by the police of unlicensed matatu46 taxis in Nairobi limits the availability of a cheap service. It has been a common feature of the control of markets and of housing, where an appeal to public health requirements is used to prevent the provision of services which the poor could afford. The destruction from time to time of ‘squatters’ ‘housing and of unlicensed market stalls by askaris of the Nairobi City Council has been a distasteful consequence of this insistence on the standards of a wealthy society. Some actions by the authorities—the periodic clearing of beggars from the streets, for instance—have perhaps also been stimulated by a desire to present a modern and respectable image to the tourist. There are indications, however, that official attitudes are changing. The government now recognises the potential for development that exists within the informal sector. It is proposed to stop the demolition of slum housing when no specific benefit results, to liberalise transport licensing and traders’ licensing, and to encourage the informal sector in various ways.47 The government paper issued in response to the recommendations of the ILO Mission says: The Government acknowledges that there is much counter-productive harassment of the socalled informal sector. This harassment will cease and more realistic standards and controls will be applied. The Government has already taken initial steps to ensure that the informal sector is provided with sufficient credit and management and technical services.48

A change in policy towards the informal sector may take some time to penetrate down from those who make policy to those who execute it.49 The authorities continue to be

20 Aid and Inequality in Kenya: British Development Assistance to Kenya reluctant to allow ‘site and services’ housing development, and the harassment of matatus and the demolition of unauthorised dwellings may not quickly be halted.50 If a change in policy towards the informal sector is fully implemented, it may assist the declared aim of ‘a more equitable distribution of resources and income’.51 The same cannot be said for the increase in the salaries of the higher civil service which followed the recommendations of the Ndegwa Commission on public service structure and remuneration in 1971.52 Although the general level of salary increase recommended was extraordinarily modest—a mere 4 per cent—very substantial percentage increases were proposed for the higher grades (many of whose salaries had not been increased for ten years), that for the Head of the Civil Service being 24 per cent. It was further recommended—doubtless to formalise what had become common practice—that there should be no objection to public servants owning property or businesses. The position of the higher ranks of the civil service was greatly improved as a result of the Ndegwa Commission’s recommendations, but a sense of proportion is necessary in judging them. The ratio of the highest to the lowest public service salaries is certainly much greater than is to be found in developed countries. Even wider is the gap between the salary of the senior civil servant and that of the wage-earner in agriculture.53 The average wage paid by the Kenya government in 1969 was £293 p.a. Wages (including housing allowance) were less than Shs 300/-per month for 45 per cent of all central government employees. In agriculture the average wage was Shs 121/-per month and 97 per cent of all wage-earners in modern agriculture received less than Shs 300/-per month. The salary of the Head of the Civil Service was K£3,200 p.a.54 and the Commission recommended that it should be increased to K£3,960 p.a. Compared with the salaries of corresponding officials in the developed countries55 that is little enough in a country where it is not notably cheaper to pursue a similar way of life. Of course, the senior civil servant of a poor country should not follow the same mode of life as his counter-part in a rich country, and some levels of the civil service may be overpaid, but there is a certain way of life which is necessary for the efficient pursuit of the activities required of such an official, and it would be entirely unrealistic to suppose that these activities could be efficiently carried out by an official whose income was, say, equal to that of the average paid in the civil service.56 It is also relevant that the civil service has to compete with the private sector for searce personnel. Despite the increase in civil service salaries there is a high turnover in senior posts. Efficiency might, indeed, be increased by the payment of even higher salaries, if it were coupled with effective restrictions on private business activity. It is often said that some officials devote more time and effort to their businesses and farms than is compatible with the most efficient discharge of their official duties. The ILO Mission recommended a five-year freeze on salaries above £700 p.a., and the Kenya government has expressed its general support for the Mission’s recommendations on wages policy.57 A freeze on higher salaries seemed to be foreshadowed in the 1972 budget speech, when the Minister said: We must define a minimum poverty line. We must then go all out with total commitment to ensure that everyone below such a minimum line is brought up to at least that minimum standard. If you agree with me in this objective, has not the time come to freeze all incomes above, say, K£700 per annum to enable us to achieve it?

Aims and Achievements Since Independence 21 Although no freeze has, in fact, been instituted, new guidelines on wages issued to the Industrial Court favour the lower-paid, and the minimum wage has been increased.58 There may, therefore, be a narrowing of the gap between rich and poor in wage employment. But this will do nothing to improve, absolutely or relatively, the lot of the poorest, who do not receive wages.

The Role of Government The Constitution of the Republic of Kenya was of fundamental importance in defining the direction of subsequent policy, and particularly the attitude to nationalisation. It provided for prompt payment of full compensation when property was compulsorily acquired and that ‘every person having an interest or right in property which is compulsorily taken possession of shall have a right of direct access to the Court’. Policy statements from the beginning indicate the absence of any ideological commitment to public ownership. The President declared in 1964 that ‘nationalization will not serve to advance the cause of African Socialism’, and Sessional Paper No. 10 of 1965 stated that ‘African Socialism in Kenya does not imply a commitment to indiscriminate nationalization’. The approach in practice is illustrated by the nationalisation of the East African Power and Lighting Company, which was acquired by the open market purchase of shares. Government participation in some other industries—banking, insurance, oil refining, for example—has also taken the form of the acquisition of shares by government ministries, particularly the Treasury. In addition to the direct acquisition of shares by ministries, government financial participation in industrial development has taken place through a number of institutions. There is, however, no development corporation, such as exists in some other countries, through which foreign capital is channelled, and which takes responsibility for the management of enterprises. Projects which qualify for government financial participation through one of the development finance institutions have been classed59 as follows: (a) (b) (c) (d) (e)

profitable but needing encouragement; profitable but risky; only marginally profitable, but involving external economies; too big for individual private domestic investors and entrepreneurs; projects in which there is a strong public interest and in which financial participation is desirable to establish public control, promote Africanisation, prevent abuse of monopoly power, and generally to ensure operation for the public interest.

The first of the financial institutions to be established was the ICDC (Industrial and Commercial Development Corporation), which was established in 1954. The ICDC is concerned with the promotion of African enterprise, although it was not until 1960 that it initiated a scheme of small industrial loans to assist African businessmen. The concern of ICDC with the promotion of African enterprise does not prevent its participating with foreign investors. On the contrary, that has been a major activity and it has a financial interest in, for example, the Firestone tyre plant, textile factories, a metal-forging plant, and a pencil factory. But its bias is towards smaller projects, which it deals with partly

22 Aid and Inequality in Kenya: British Development Assistance to Kenya through a subsidiary, Kenya Industrial Estates. Some of the Corporation’s holdings in foreign enterprises have been transferred to Kenyans, as in the transfer of shares to the ICDC Investment Company, the equity of which is held by Kenya citizens. One analysis of ICDC activities in the field of small business concluded that, although ICDC had had its successes, it had primarily helped successful businessmen to consolidate their success; it had not played an initiating and creative role, and would have been more effective if it had taken equity participation instead of offering fixed-interest loans. A second institution, the Development Finance Company of Kenya (DFCK), is particularly concerned with large-scale projects. It is a privately incorporated company and operates with finance obtained from the Kenya government, through ICDC, and from agencies of the British, Dutch and German Governments. It can participate in a project through loans or equity holdings, and it is the intention that it should raise some of its funds by selling from its equity portfolio to the general public. A new institution appeared on the scene in 1973 with the establishment of an Industrial Development Bank, under separate management from the ICDC though owned jointly by ICDC (51 per cent) and the Kenya Treasury. The ICDC, with its Kenyanisation aims, and its use in the implementation of trade licensing and the transfer of businesses to Kenyans, was not considered by the World Bank as an appropriate vehicle for the investment of its funds. The Industrial Development Bank will raise external loans, including loans from the World Bank, to finance major industrial development projects. Despite the absence of any ideological inclination towards the expansion of government economic activity, there is a widespread impression that the role of government in the economy has greatly expanded. In absolute terms the contribution of government to monetary GDP (at constant prices) increased by 140 per cent between 1964 and 1974.60 However, as a proportion of monetary GDP the contribution of government increased only modestly, from 18 per cent to 21 per cent. Nor is the position significantly different for the public sector as a whole, the contribution of which to monetary GDP rose only from 33 per cent to 35 per cent between 1964 and 1973.61 Within the public sector, and within the narrower government sector, there have been large changes. The share of East African Community institutions has declined and that of Kenya parastatal bodies more than doubled. Administration has declined and social services increased in importance. However, between sectors, it is the constancy of the shares which is striking. Table 1: Sector shares of monetary GDP (constant 1964 prices) (percentages of monetary GDP) 1964 Agriculture Manufacturing Construction Trade Transport Banking etc. Government Other Source: Economic Survey 1975.

22 14 3 14 10 4 18 15

1974 19 16 3 10 10 5 21 16

Aims and Achievements Since Independence 23 It is a different matter with respect to the distribution of resources (monetary GDP plus the import surplus) between consumption and investment. Gross investment took 32 per cent of available resources in the monetary economy in 1974 compared with 16 per cent in 1964. The relative decline in consumption which was the counterpart of this increase was concentrated on private consumption. The share of public consumption was constant, taking approximately one-fifth of total resources; the share of private consumption declined from 64 per cent in 1964 to 48 per cent in 1974.62 The share of the private sector in monetary capital formation also declined from nearly three-quarters of the total in 1964 to barely more than one-half in 1974. The increased role of the public sector is therefore reflected not so much in the sectoral contribution to GDP as in the increasing proportion of resources devoted to investment and the increasing part of investment being carried out by the public sector. Table 2: Use of resources (current prices) 1964 K£m. 267.7 −17.4 —— Total resources available 250.3 Gross investment 41.4 Consumption: Total 208.9 Public 49.4 Private 159.5 Source: Economic Survey 1975, Tables 2.5 and 2.8A.

%

1974 K£m.

%

100 16 84 20 64

791.2 −81.1 —— 872.3 283.1 589.2 169.0 420.2

100 32 68 19 48

Monetary GDP at market prices Import surplus of goods and services

A review of the policies pursued by the Kenya government would be misleading if it contained no reference to the ‘style’ of government in Kenya. To an important extent it is government by the civil service. The initiative of the civil service in policies, particularly economic policies, is far-reaching, but it reaches only so far, and cannot be exercised without thought for the possibility of presidential intervention. Major policy changes have been made from time to time by presidential announcement, without apparently much foreknowledge by civil servants, or even ministers. The desirability of careful preparation and planning of major innovations is not allowed to stand in the way of ‘the felt necessities of the times’—the President’s feel for the political realities, In this manner the price of milk was increased, creating financial problems for Kenya Cooperative Creameries, which had to buy at the higher price, and making it necessary to raise other produce prices to prevent an excessive switch to grazing; so the prices paid to maize, wheat and pyrethrum farmers were raised, although a policy of reducing producer-prices was embodied in the Development Plan.63 Similarly, free primary education in certain grades was introduced, and in a different field of policy Kiswahili was declared to be the national language.

The attitude to aid This chapter has been concerned to set the scene in which the provision of external aid and technical assistance to Kenya takes place. Kenya’s policy towards aid has been consistent

24 Aid and Inequality in Kenya: British Development Assistance to Kenya with its other economic policies. Aid and technical assistance are welcome because they mean that ‘development here need not be limited quite so severely by the shortage of domestic savings and education’; but aid will not be welcomed ‘if it is designed to promote the economic or political dominance of the aiding country’.64 An elaboration of policy towards aid is contained in Sessional Paper No. 10 in connection with the terms of bilateral aid, particularly with reference to aid in inconvertible currencies, the financing of local costs, and the repayment of loans. Aid offers, the paper states, must be examined in the light of foreign exchange policies: We must avoid, for example, aid which requires us merely to advance credit to a developed country or to mortgage our future production to any given country or which makes us a mere dumping ground for products of a donor country. Aid terms must be related to the productivity of the project and its positive contribution to our economy. These are matters which must not be decided upon just on political or emotional considerations.65

The need to develop expertise in the assessment of aid offers was, therefore, recognised early. By 1969 the Minister of Finance in his budget speech was able to remark: We are now experienced in the negotiation of overseas aid, and have been much encouraged by indications that the World Bank, which is already our largest overseas source, will be prepared to increase the rate of its lendings to Kenya. The strings attached to World Bank loans are, in fact, no more than the sort of guide lines which we would, in any case, wish to use…

A less totally optimistic assessment of Kenya’s aid policy was given by a former economic adviser to the Treasury. The tendency to expand aid for its own sake has, he suggested, been resisted: The pressures from abroad (and within for that matter) to build clearly distinguishable plaque-hanging projects to high specifications have in general been resisted. Moreover, many of the opportunities to finance turnkey projects have been rejected following costbenefit analysis. Nevertheless, the tendency exists to tailor projects to donor requirements, to accept aid for imported goods which could be produced domestically, to finance imports through aid even though the true aid element may be dissipated through higher prices, and to reduce or eliminate in the Plan projects and programmes essential for development which do not, however, appear to be attractive to donors. It requires vigilance and perseverance to avoid these pitfalls and Kenya’s record, while not perfect, is certainly good.66

The temptations have not, however, always been resisted. In fact, a Treasury circular to the spending departments informs them that projects have a chance of acceptance only if foreign aid for them has already been arranged, though this does not mean, of course, that projects inconsistent with the Development Plan would be accepted simply because they had the support of a potential aid donor. (These and other issues of policy towards aid are taken up in detail in later chapters.)

The Current Situation The background to an analysis of economic aid—to bring together the various arguments of this chapter—is a decade of substantial progress in Kenya as measured by the conventional indicators, The Minister of Finance was able to say in 1974:

Aims and Achievements Since Independence 25 Since independence, our economy has grown at an average rate of 6.8 per cent per year. Moreover, such a statistic does not reveal the great extent to which the economy has been modernised by employing more efficient methods of production. Manufacturing output (price adjusted) has grown by 8.1 per cent per year since 1963, while agriculture has grown at 6.5 per cent. Before independence, most high- and middle-level occupations were manned by expatriates; by 1972, 74 per cent of them were in the hands of Kenyans. The number of pupils enrolled in primary schools has almost doubled since independence, while those in secondary schools and the university have increased by more than fivefold. The major targets of the second Five-Year Plan (1970–74) were all realised.67

In addition, among other achievements, including a rise in exports by 140 per cent,68 there has been the achievement in agriculture of a largescale and orderly transfer of land into African ownership, and a great expansion in the contribution of small holders. All this was success, indeed. Nevertheless, in the middle of the 1970s strong doubts remain about the effects and the future of the pattern of development pursued since independence. Despite the growth of productive capacity, the progress of import substitution, and the expansion of exports, the potential balance of payments position remained serious, even before the effects of the oil crisis began to be experienced. The current account deficit had fluctuated from year to year, and had been tending to increase, but except for two years (1967 and 1971) had been adequately covered by the inflow of long-term capital and aid. However, the projections of the economy made at the beginning of the 1970s assumed an inflow of capital at a much higher level than had been achieved in the past, and even at the time seemed unrealistic.69 In 1974, the current account deficit jumped to K£1 16.1m. from the 1973 figure of K£47.5m. and as a proportion of GDP increased from less than 6 to more than 12 per cent. Despite the progress that has been made, the absolute number of the poor may not have diminished. Unemployment and underemployment have become worse. Migration to the towns, often into a poverty-stricken and unproductive way of life, has increased, indicating perhaps that the opportunities offered in the rural areas are even worse. This state of affairs in the rural areas persists despite the commitment to ‘direct an increasing share of the total resources available to the nation towards the rural areas’,70 and a policy of redistributing income through increased expenditure on education and health combined with widespread remissions of charges for these services.71 Population growth continues at an accelerating rate,72 adding to the intensity of these problems. This darker side of success has stimulated the recent emphasis on inequality within Kenyan society, though the concern about inequality might not be so great if the absolute level of the poorest sections of the population were seen clearly to be improving. The government is committed to a policy of greater equality, but it has been committed to such a policy from the beginning.73 It is acknowledged that: In spite of the rapid growth of the economy, in the first ten years of independence, the problems associated with a rapidly growing population—unemployment and income disparities—have become more apparent than they were in 1963.74

And the President has declared that economic growth can no longer be regarded as the only objective of our nation. Rather, it is a means of attaining other goals…these include full participation of our people in the economy, greater employment opportunities, and a more equitable distribution of resources and income.75

26 Aid and Inequality in Kenya: British Development Assistance to Kenya The commitment is clear. Unfortunately, the middle 1970s is the least auspicious time since independence for achieving a substantially more egalitarian distribution of income and wealth. The days of relatively easy growth and redistribution are over. The major transfers of land have been completed. The high-potential lands are approaching full utilisation with existing techniques, and with the informal constraints that operate on the use of land by people from other areas. Obvious import substitution has been carried out. Commerce has been Africanised. In general, the easy options have been taken up and the way to further progress is a more difficult one. On top of that change in the opportunities for growth and redistribution came the increase in oil prices, which was a major blow to the economy, seriously affecting the balance of payments, in which structural difficulties were already evident, and boosting the inflationary forces that were already active. A Sessional Paper of early 1975 said that ‘Kenya’s impressive record of economic growth—6.8 per cent per annum during the first decade of independence—is currently in danger of erosion’.76 The Paper forecast that over the Development Plan period the annual increase in GDP in the monetary economy would amount to 6.7 per cent, compared with the Development Plan target of 8.4 per cent, and that real income per head would rise by only 1.3 per cent, compared with the target of 3.7 per cent. The strategy of the ILO Report was ‘redistribution from growth’. There were those who thought that, despite the commitment of policy, it was optimistic to expect substantial redistribution in the face of powerful vested interests in the economic structure that had been created. If there was truth in this view it must indeed be Utopian to expect redistribution in the absence of growth. With a low rate of growth, no major improvements in the situation of the poor, and with the spectre of the population explosion standing over all, the prospect would be uncertain, but menacing. The remainder of this book analyses the part played by external aid, particularly British aid, in the developments in Kenya that have been outlined, and suggests its future role. One point concerning the role of aid has, we believe, already emerged from this summary examination of developments in Kenya. Whatever may have been the alternatives at an earlier stage, decisions made during the transition to independence and changes initiated at that time have had an enormous influence on later events. The structure of the economy and the policies pursued in independent Kenya have been deeply rooted in the past. Equally they are now firmly rooted in the interests of influential sections of Kenyan society. It would be a serious misreading of the situation to see Kenya as a reluctant practitioner of policies forced on it by the leverage’ of foreign aid donors.

Notes 1. Sessional Paper No. 10 of 1965, p. 51. 2. The fact that most settlers are in arrears with loan repayments qualifies in practice the principle that they had to pay for the land. (See below, pp. 110, 122–3.) 3. This was as near as Kenya came to expropriation for settlement. There was dispute between the UK and Kenya governments over the treatment of farmers excluded from their farms for mismanagement. The legislation provided for the compensation of farmers, and this was not done. Eventually the courts ruled that compensation should be paid. 4. Development Plan 1974–1978, Part I, Government Printer, Nairobi, p. 199.

Aims and Achievements Since Independence 27 5. Ibid. 6. ‘The phrase telephone farmers’ has been coined to describe the activities of some of these buyers of farms. 7. Development Plan 1974–1978, Part I, p. 200. 8. ILO, Employment, Incomes and Equality: A Strategy for Increasing Productive Employment in Kenya, Geneva, 1972, pp. 34–5. 9. The Mumias development of sugar production is examined in Chapter Five. 10. Sessional Paper No. 10 of 1965, para. 82. 11. Ibid., pp. 10–11. 12. A senior official of the Ministry of Agriculture ‘dismissed as lies the talk by some people that the Government would auction their farm if they failed to repay the loans’. (The Standard, 19 August 1975.) 13. Development Plan 1974–19 78, Part I, p. 212. 14. Gross marketed production from large farms was £37m. in 1963, £72m. in 1974 (Economic Survey 1967, p. 33, and Economic Survey 1975, p. 89). The changes in producer prices over the period indicate that the increase in value is not simply the result of rising prices: Wheat 125

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

Maize 140

Producer Prices, 1974 (1963=100) Coffee Pyrethrum Sisal Tea 153 115 92 172

Seed Cotton 149

Coffee accounted for only 17 per cent of total sales from large farms in 1967 (the last year for which data are available of the value of output of particular crops from large and small farms separately). Output from large farms has been under a number of influences, including the reduction in area as a result of settlement and the use of improved seeds. ILO, op. cit., Chapter 6, especially pp. 83, 95–8. See below, p. 106. Development Plan 1966–1970, Government Printer, Nairobi, 1966, p. 56. Development Plan 1974–1978, Part I, p. 199. Judith Heyer, IDS Working Paper No. 194, p. 26. Ministry of Commerce and Industry, A Guide to Industrial Investment in Kenya, 2nd ed., l912, p. 3. See A, Hazlewood, Economic Integration: The East African Experience, London, Heinemann, 1975. ILO, op. cit., p. 286. Statistical Abstract and Economic Survey 1975, p. 129. Value of output= manufacturing and repairs component of GDP. Development Plan 1974–1978, Part I, p. 278. No later figures have been published. Hazlewood, op. cit. The allocation of markets to their different plants by multinationals could either benefit or limit Kenyan exports, as compared with a more competitive situation. A product-by-product analysis would be necessary to determine the total effect. See Frances Stewart, ‘Kenya: Strategies for Development’ in Ukandi Danaachi et al. (eds), Development Paths in Africa and China, London, Macmillan, 1976. Census of Industrial Production, 1967. See ‘Lower Income Index of Consumer Prices—Nairobi: New Weights’, Kenya Statistical Digest (quarterly). Roughly 70 per cent of total expenditure is on goods and 30 per cent on services, mainly rent (20 per cent) and school fees (nearly 5 per cent). Products can also be positively deleterious. The British government believes this to be so for milk-based baby foods and feeding bottles. A letter from the Parliamentary Under-Secretary of State for Overseas Development, of July 1975, says:

28 Aid and Inequality in Kenya: British Development Assistance to Kenya There are strong reasons for arguing that these foods are not appropriate for use by mothers in developing countries, who often do not understand how they should be administered or lack the facilities for preparing and sterilising feeding bottles in hygienic conditions. In the MOD we share this concern, and have no doubt that promotion of the sales of milk-based baby foods in developing countries is contributing to infant mortality and malnutrition. 32. Imports were the vehicle in the original exposition of the effect. 33. Development Plan 1974–1978, Part I, p. 280. See also I.M.D.Little, T.Scitovsky and M.F.Scott, Industry and Trade in Some Developing Countries: A Comparative Study, London, OUP for the OECD Development Centre, 1970. 34. Paradoxically, later in the speech the Minister announced that import duties were to be raised in order ‘to restrict the growth of demand for a number of import items in the luxury or semi-luxury class’, some of which are produced domestically. 35. Development Plan 1974–1978, Part I, p. 280. 36. Sessional Paper on Employment (No. 10 of 1973), para. 279. 37. Sessional Paper No. 1 of 1974: On the Current Economic Situation in Kenya, p.6. 38. It is proposed increasingly to replace tariffs and import controls by excise and sales taxes so as to discourage the consumption of inessentials, whether imported or home-produced. See Development Plan 1974–1978, Part I, p. 281. 39. Ibid., p. 280. See also Sessional Paper on Employment (No. 10 of 1973), p. 31, para. 112, and p. 38, para. 138. 40. Development Plan 1974–1978, Part I, p. 284. 41. Employment: percentages of total Citizens Non-citizens Africans 1967 92 8 91 1972 96 4 95 (Calculated from Table 2.2, Development Plan 1974–1978, Part I, p. 39.) 42. 83 per cent of the total in public employment and 68 per cent of the total in private employment. See Development Plan 1974–1978, Part I, p. 38. 43. One cannot but be impressed by the different attitude adopted to British farmers of United Kingdom origin, and British traders of Asian origin. The latter had to dispose of their businesses, when required to do so by the Kenya government, in the market context of such forced disposal. There was no finance from Britain to buy out the traders at a free market price, though it is believed that the Kenya government at one time suggested that the UK should make funds available for this purpose on the lines of the Land Transfer Programme. 44. ILO, op. cit, pp. 5 and 21. 45. See Walter Elkan, ‘Criteria for Industrial Development in Uganda’, East African Economic Review, Vol. 5, No. 2, January 1959. 46. From 30 cents, the fare at one time charged. Tatu is ‘three’ in Kiswahili. 47. See Sessional Paper on Employment (No. 10 of 1973), paras. 154 and 170, and Development Plan 1974–1978, Part I, p. 11. 48. Sessional Paper on Employment, para. 99. 49. At lower levels of the administration the degree of understanding may remain abysmal, as perhaps is illustrated by the newspaper report that the authorities in Kisumu were taking action against shoe-shine boys who, because there were so many of them, were exploiting the public. 50. Despite the new policy it could be reported in August 1975 that ‘Nairobi City Council’s “wrecking squad” yesterday moved into swift action and pulled down several shanties at Mathare Valley leaving 100 families homeless.’ (Daily Nation, 15 August 1975.) 51. Introduction by the President to Development Plan 1974–1978.

Aims and Achievements Since Independence 29 52. See Report of the Commission of Inquiry (Public Service Structure and Remuneration Commission) 1970–71, May 1971 (Ndegwa Report). 53. See Ndegwa Report, op. cit., Chapter VI for the figures. 54. There must be added a housing allowance which for the higher salaries varied according to circumstances between 13 and 30 per cent of salary. Ibid., pp. 246–8. 55. The salary of the Head of the UK civil service at the time was £ 16,750. 56. It might be argued that it is possible if political and ideological commitment is sufficiently strong, and some might point to China and Tanzania as examples. There is no such commitment in Kenya. 57. Sessional Paper on Employment (No. 10 of 1973), para. 212. 58. Sessional Paper No. 4 of 1975, p. 14. 59. Ministry of Commerce and Industry, A Guide to Industrial Investment, 2nd edition, 1972. 60. A greater increase in the relative importance of the government sector is indicated when the government sector is compared with total GDP, monetary and non-monetary, because the official figures of non-monetary production show its share declining from 27 per cent to 21 per cent. Measured in this way the government share increased from 13 per cent to 17 per cent. Given the difficulties associated with the measurement and the meaning of the statistics of non-monetary production, the reality of the sector shares is best expressed in terms of monetary production alone. GDP monetary economy (1964 prices) 1964 1974 K£m. % K£m. % General government 42.5 18 102.5 21 Other sectors 198.6 82 377.4 79 Total 241.1 100 479.9 100 Source: Economic Survey 1975, Table 2.1. 61. Economic Survey 1974. Figures for 1974 are not available. 62. Figures for the contribution of education to GDP are not available for 1974. But education has been a growth sector, partly as the result of deliberate policy—primary school fees have been abolished in Standards I to IV—and partly, one suspects, because the government is unable to control the growth. 63. See C.Leys, Underdevelopment in Kenya: the Political Economy of Neo-Colonialism 1964– 1971, London, Heinemann, 1975, pp. 110 and 112. 64. Sessional Paper No. 10 of 1965, para. 25. 65. Ibid., p. 24. 66. E.O.Edwards, ‘Development Planning in Kenya since Independence’, East African Economic Review, Vol. 4, No. 2, December 1968. 67. Preface to Development Plan for 1974–1978. The growth rate stated is in real terms, i.e. at constant prices. 68. Exports by Kenya 1963 1973 To outside E. Africa 43.8 122.6 Re-exports 7.1 6.3 To Tanzania and Uganda 19.8 38.8 Total 70.7 167.7

30 Aid and Inequality in Kenya: British Development Assistance to Kenya 69. See ILO, op. cit., p. 280. 70. Development Plan 19 70–19 74. 71. Development Plan 1974–1978, Part I, pp. 42–3. Expenditure on social services accounted for 17 per cent of total government expenditure in 1965/6 and for 32 per cent in 1974/5. 72. ‘The 1962 census suggested that the rate of growth was 3 per cent per annum; that of 1969 indicated that it had risen to 3.3 per cent.’ (Development Plan 1974–1978, Part I, p. 43.) It is now said that the figure is 3.5 per cent 73. Compare Sessional Paper No. 10 of 1965 and Sessional Paper No. 10 of 1973. 74. Development Plan 1974–1978, Part I, p. 1. 75. Introduction to Development Plan 1974–1978. 76. Sessional Paper No. 4 of 1975.

3 THE FACTS OF AID

Many different states and international bodies give aid to Kenya. The UK’s relative importance as a donor has quite naturally declined since independence. In 1964 well over 80 per cent of all Kenya’s official aid receipts came from the UK, compared with under a half of aid gross and 22.6 per cent of aid net of amortisation in 1972. West Germany and the United States have remained significant donors for much of the independence period and a development has been the emergence of the Scandinavians and the Dutch as important donors. The World Bank group has been a very important donor of financial aid with a tendency in recent years for credits by the International Development Association (IDA) to decline in importance compared to IBRD lending, so hardening the terms of multilateral aid to Kenya. Table 3 shows these trends. As it relates to aid net of amortisation, however, it overestimates the decline in Britain’s contribution because, as a longer-standing donor, Britain receives more repayments than the others. Technical assistance remains a considerable part of total aid. There does not seem to be any marked trend one way or the other in the number of technical assistance personnel in Kenya since independence. The rundown in operational personnel whose emoluments are partly financed by Britain has been balanced by increases in the numbers of advisers from other countries, Compared with the total number of people employed by the Kenya government, the number of technical assistance personnel is small. Wage employment by the central government was 135,000 in 19731 and there were fewer than 3,000 expatriates. There are, however, some remarkable concentrations. J.R.Nellis2 calculated that in 1971 the Ministry of Finance and Planning had 111 Nairobi-based upper-level posts and a number were vacant (around 19). At that time there were 34 expatriate technical assistance personnel working there altogether, 22 being technical advisers. By the not-foolproof technique of studying names and looking at the number of established posts in different ministries in 1969 and 1972, Nellis compiled percentages of expatriates in senior positions in all ministries. In 1972 they varied from nil in the Ministries of Foreign Affairs, Co-operatives, and Defence to 31.76 in Agriculture (38.23 in 1969). The average figure for all ministries was 15.68 per cent (23.15 per cent in 1969). Those ministries with over 20 per cent senior expatriate workers in 1972 were (1969 figures in brackets): Agriculture—31.76 per cent (38.23); Finance and Planning—30.43 per cent (28.37); Works—29.62 per cent (37.93); Education—20.56 per cent (18.03).

0.5

−1.3 1.0 …

2.9

100

0.5

−0.7 −33.0 …

3.4

100

8.1 – – 0.2 87.3 3.6

1.0 0.3 2.1 0.2 0.2 0.5 74.8 17.7

0.6

1965 B

100

1.7

9.9 4.3 …

1.1

1.9 0.2 6.3 – 0.7 2.2 47.4 24.3

1966 B

100

2.5

28.0 7.2 …

1.9

2.7 0.5 3.9 2.8 1.7 5.2 40.1 3.9

1967 B

100

7.2

8.8 14.3 …

1.1

2.4 1.3 6.2 0.1 2.9 6.2 40.9 8.7

100

100

3.4

9.4 15.3 4.1

– – 10.8 3.1

2.3

2.5 1.8 3.3 0.3 5.4 5.9 36.9 9.5

4.9

6.2 3.6 8.2 – 5.3 10.6 26.8 20.4

(percentages) 1968 1969 B A B A

100

3.6

– – 12.3

3.9

B

100

2.8

13.3 9.8 4.9

2.5

3.0 2.0 7.3 2.7 3.9 4.1 34.9 8.9

1970

6.3 3.0 8.0 6.6 4.7 8.1 28.8 14.7

Source: OECD, Development Co-operation: various issues. Key: A: percentage of total technical assistance received by Kenya. B: percentage of total aid (financial+TA) net of amortisation but gross of interest payments.

Canada Denmark W.Germany Netherlands Norway Sweden UK US Other bilateral IBRD IDA UNDP Other multilateral

1964 B

Table 3: Source of aid disbursements to Kenya

100

2.8

– – 12.7

3.5

7.5 6.8 1.4 8.1 5.0 8.3 21.6 12.3

A

B

100

4.4

19.5 3.6 4.7

8.7

3.0 3.8 6.6 4.1 3.5 4.1 22.3 11.7

1971

100

3.3

– – 9.0

2.9

5.5 9.0 11.0 11.6 4.9 7.9 27.4 7.3

A

B

100

4.3

29.6 5.5 3.5

1.5

2.2 5.2 4.7 5.3 3.5 7.3 22.6 4.7

1972

The Facts of Aid

33

Table 4: Technical assistance to Kenya and EAC by country of origin: September 1973 Operational posts A B Australia UK Canada Denmark Finland France Ford Foundation West Germany India Italy Japan Netherlands Norway Rockefeller Foundation Sweden Switzerland UNDP USA USSR WHO Yugoslavia Totals:

Advisers A B

Volunteers A B

5 5 – – – 987 1377 62 78 107 – – 54 56 _ 30 32 63 65 106 _ _ 10 10 – 19 19 10 10 – – – 4 4 – – – 74 75 34 – – 1 1 – – – 8 8 24 – – 28 28 44 2 2 63 64 73 30 32 39 42 33 – – 6 6 – 43 43 5 5 25 2 3 12 12 – 5 5 91 97 – 11 11 12 20 180 8 8 – – – – – 33 33 – 3 3 – – – Overseas Service Aid Scheme (OSAS)* British Expatriates Supplementation Scheme (BESS)* Operational personnel, not from UK Advisers Volunteers Grand total: A: numbers working for the Kenya government only. B: numbers working for the Kenya government plus East African Community. Source: Kenya government. * See below, p. 63.

– 112 _ 106 – – – 34 – 24 44 73 33 – 25 – – 181 – – –

Total 5 1565 56 208 10 29 4 109 1 32 72 139 107 6 73 15 102 212 8 33 3 1134 160 246 614 637 2,791

Two ministries, Finance and Planning and Education, were the only ones to show absolute and percentage increases in the number of expatriates employed between 1969 and 1972. Since 1974, however, there has been a reduction—from 20 to 11—of British expatriates in the Ministry of Finance and most of the survivors are in technical rather than policy positions, in the Tax Department and Bureau of Statistics. It is interesting that the ministries on Nellis’s list are by-and-large those whose efficiency is most criticised by donor agencies except for the Ministry of Finance, which is, however, the major link with donors. The conjunction suggests that technical assistance may be a response to the felt needs of donor agencies as much as of the Kenyan government.3

34 Aid and Inequality in Kenya: British Development Assistance to Kenya Table 4 gives a breakdown of technical assistance in Kenya in 1973 by type of personnel and country of origin.

British Capital Aid British aid to Kenya has not been typical either of the programmes of other donors there, or of British aid in general. About 45 per cent of British capital aid since independence has been for purposes other than general development. Most of this money has gone into the Land Transfer Programme (LTP) or into land adjudication and registration. Excluding the independence settlement (see Table 5) and considering only commitments made since 1964, over half of British aid has been for land transfer, settlement or adjudication. Despite the bias towards LTP, British aid was very important in financing the Kenya government’s development budget in the early years of independence. In 1963/64, gross financial aid accounted for 87 per cent of development revenue, and 77 per cent of that revenue came from Britain.4 Aid, net of loan repayments and interest, was equivalent to 5.9 per cent of Kenya’s GDP in 1964, and the proportion has changed little since. In 1970 it was 43 per cent and in 1972 5.4 per cent. Owing to the expansion of public sector development spending, however, financial aid has declined in importance as a means of financing the development budget. In recent years it has averaged around 50 per cent of development revenue, and Britain’s contribution has fallen to around 10–15 per cent. Britain also contributed to Kenya’s recurrent budget expenditure with technical assistance grants but aid is now an insignificant proportion of the recurrent budget. For some years aid has been around 12–15 per cent of total government expenditure, but despite this relatively modest proportion the Kenyans have difficulty in spending the aid on offer. This certainly applies to British aid. In 1974, three-quarters of the loan agreed in 1973 had not been definitely allocated to any project; nearly £1m., one-fifth, of the 1970 loan was unallocated, and there were even small sums outstanding from the 1966 loan and the pre-independence Colonial Development and Welfare grant. Other donors have had similar experiences. One reason Table 5: British official capital aid commitments to Kenya since Independence £m. 1964 Independence settlement Land Transfer Programmea: Unissued balance of commitment of £21.3m. as grant and loan

11.0

Further loan for Land Bank

1.0

Development grants and loans: unissued balances

4.3

Development grant 1964/5

1.0

Development loan 1964/5

2.0

Pension loans

13.6

Budgetary grant

1.25

Special land purchase scheme loan

1.275 34.475

The Facts of Aid

35

1965/6 General development loan (interest-free, 42% import content)

3.0

Land Bank loan

1.0 4.0

1966/70 Land Transfer Programme (interest-free loan)

6.3

General development loan (interest-free, 60% import content)

8.7

Land adjudication loan (previously classified under general development)

3.0 18.0

1970 Land Transfer Programme: Settlement (grant)

2.5

Agricultural Development Corporation and Agricultural Finance Corporation loan (2% interest)

1.0

Special scheme grant General development loan (2% interest, 75% import content) Land adjudication/Rural development loan (2% interest)

0.25 5.0 2.75 11.5

1973 Land Transfer Programme: Settlement grant Agricultural Finance Corporation loan (2% interest) General development loan (2% interest, 50% import content)

6.0 2.0 10.0 17.0

TOTAL

85.975

In addition between 1963 and 1969 about £1m. was provided specially for the purchase of farms of compassionate cases outside the settlement areas. a

is the procurement-tying of funds, whereby Kenya is obliged to spend part or all of an aid commitment in the donor country. That is not the main reason, however. Of the £1m. of British aid outstanding from 1970, only some £103,000 had to be spent on British goods, showing that most of the tied aid had been spent and much of the untied portion of aid loans had not.5 Up to 1970, British financial aid for general development was spent in a large number of mainly small packets. Having a Development Plan including many small projects, the Kenyans approached HMG with a ‘shopping list’ of aid requirements, mainly for capital equipment, The only projects carried out before 1970 where the British aid contribution was over a quarter of a million Sterling were: construction of the Nairobi-Mombasa road, which was financed by money from more than one loan (£2.2m.); the Chemelil sugar factory (£405,000), which also received much German aid; afforestation schemes (£715,000); and provision of road-building equipment for the North-East Region (£320,000). Of the

36 Aid and Inequality in Kenya: British Development Assistance to Kenya five projects financed by the 1970 loan, however, four received aid of over £250,000. They were: Mumias sugar factory (£2.9m.); Mombasa TV (£260,000); Naivasha-Suswa pipeline (£465,000); livestock marketing scheme (£288,000). The fifth project, provision of large hermetically-sealed bins for grain storage, received aid of £170,000. That was still more than the average loan disbursement of £150,000 before 1970. This pattern for larger discrete disbursements continued with the 1973 loan. Apart from continuing land adjudication expenditures (now classed with general development) costing £1.8m., the other projects where aid was settled by end-1975 were the second phase of a livestock marketing scheme receiving over £1m., and a pilot scheme for rural access roads receiving £250,000. Other projects under consideration included construction of a Polytechnic and University colleges, grain storage, and a scheme for rural access roads. All were likely to receive over £1m. each. Over the whole post-independence period the sectoral breakdown by value of British general development aid disbursements was:6 agriculture, 17 per cent; processing agricultural products, 23 per cent; physical infrastructure, 32 per cent; social services, 25 per cent; other, 3 per cent. If the Land Transfer Programme and land adjudication programmes are counted as assistance for agriculture, however, total aid to that sector is 59 per cent of gross capital aid. The trend in the real value of British capital aid to Kenya has been steadily downwards since independence. Table 6, which includes disbursements by the Commonwealth Development Corporation7 and technical assistance (covered separately in Table 8), gives disbursements at current prices. As Britain has been a donor for some years, amortisation and interest repayments on earlier British loans are now quite considerable. In 1972, for example, Britain’s total net aid flow was barely more than technical assistance disbursements; new financial flows were practically wiped out by repayments and interest on earlier loans. For the period 1963–72 Britain received some £50m. from Kenya in amortisation and interest payments. Net and gross aid figures therefore diverge substantially and it is necessary to consider which are appropriate when examining the British aid effort. The concept of ‘net aid’ suggests that, if new aid flows are balanced by amortisation and interest payments, Britain is not aiding Kenya. That is an odd formulation because earlier flows ought to have engendered investments leading to a social return exceeding loan charges. If current and future disbursements. do not do so, aid should not be accepted. If it was right to accept the aid, the subsequent netting of flows for any particular year is quite arbitrary, although it has the useful effect of emphasising the budgetary and balance of payments difficulties which may be encountered by a developing country. It also goes some way to offset the crudeness of statistical series that show total aid flows without discriminating between loan and grant or loans on different terms. Rather than netting aid figures, however, a better procedure is to take account of terms directly and to compute the ‘grant equivalent’ of each aid disbursement showing its net value at disbursement when the present value of future charges has been deducted.8 British net aid disbursement to Kenya 1963–73 was some 60 per cent of gross disbursements. Grants accounted for a third of the latter and the loans have a grant equivalent of at least 80 per cent, depending on discount rate. Net British aid in the superior sense of grant equivalent would thus be over 87 per cent of gross aid. For many purposes, therefore, the gross aid figures give a better picture of the transfer than the net figures.

The Facts of Aid

37

We now attempt to refine the figures for aid disbursements given in official statistics by taking account of important developments that they ignore. We shall adjust the figures for inflation in Britain, parity changes of the pound Sterling and the extent of procurementtying. All figures in Table 6 are in current prices. If they are deflated for changes in the British price level9 they reveal that aid has been falling in real terms, even ignoring the 1967 devaluation, and the float of the pound from 1972. The accumulating burden of Kenya’s repayments to the UK is shown by the rate of decline of the net aid disbursement figure of nearly 20 per cent a year. These figures take account of the changing purchasing power in the UK of aid flows to Kenya but take no account of the changing purchasing power of UK aid in Kenya and elsewhere in the world. Two factors affect that: the extent of the procurement-tying of aid and the value of the pound relative to other currencies. Procurement-tying is an important feature of aid and we consider it in some detail. British aid is tied entirely to the purchase of British goods or goods produced in Kenya (so-called local costs). If the Kenyans wish to spend British aid directly on goods from third countries, they have to secure the approval of ODM which in turn approaches the Department of Trade. Both ministries have to be satisfied that no equivalent British good is available before any special dispensation is given. Such requests and dispensations are rare. The proportion of general development loans that has been tied to purchases of British goods as opposed to local costs has varied between aid agreements.10 A weighted average proportion for all general development loans after 1964 is 51.7 per cent. The proportion of all British capital aid to Kenya tied formally to procurement in the UK is 24.6 per cent, as LTP and land adjudication funds are untied. In addition, all technical assistance is tied to acquisition of personnel in the UK. This was carried to derisory lengths. A British passport holder, for example, if educated and resident in Kenya, might not be eligible for employment by the Kenya government under UK technical assistance. There has been some liberalisation of these conventions owing to the personal intervention of the former Minister for Overseas Development, Judith Hart. Special authority can now be obtained from ‘higher channels’ within ODM for non-British residents to be designated technical assistance personnel. It remains rare. Formal procurement-tying is not the only sort. Much British money has been tied to its use in the purchase of land under the Land Transfer Programme where there is no question of the acquisition of British goods, but some 60 per cent of the money paid to British farmers for their landholdings is brought back from Kenya to Britain, as Kenyan Exchange Control allows, so it too is effectively procurement-tied. The practice of procurementtying clearly involves costs for the recipient country which should be deducted from the nominal value of aid.11 The most obvious cost is the prevention of competitive international tendering for contracts, so the price the Kenyans have to pay British suppliers for goods may consequently be well above the world market price. This is particularly likely where the supplier is in a monopolistic position in the British market. Empirical studies of these effects in many Idcs have produced widely differing estimates of the additional costs. An UNCTAD publication cites a figure of 20 per cent as being quite a reasonable estimate of the higher prices tying entails on average for Idcs.12

13989

12739

net of amortisation

net of amortisation and interest

5679

5024

net of amortisation

net of amortisation and interest 4447

4986

5451

14736

16142

16692

1965

4198

4875

5355

8481

11096

12095

1966

2938

3673

4166

4842

7751

9583

1967

1686

2496

3001

7016

9968

10965

1968

1003

1804

2394

6485

9706

10756

1969

897

1669

2535

6357

9804

11099

1970

2093

2829

3587

4465

7870

9820

1971

520

1209

7787c

4895

8437

24434

1972

b

a

includes £13,797m. to write off loans for civil servants’ compensation and pensions. This does not appear in Table 5. Owing to the establishment of the headquarters of many EAC services in Kenya, much of this aid was actually spent in Kenya. c Includes some £5.8m. write-off of pension loans. Sources: British Aid Statistics 1964–68 to 1968–72.

6039

gross disbursements

To the East African Communityb

14391

gross disbursements

To Kenya

1964

(£’000)

Table 6: Total British aid disbursements to Kenya and the East African Community

1537

2177

2977

5009

8793

11668

1973

24343

31397

43292

75025

103556

131503

1964–73

Total

The Facts of Aid

39

Table 7: British aid disbursements at 1964 prices 1964

1965

Gross 14.4 16.3 disbursements Net transfer (gross disbursements less amortisation and 12.7 14.3 interest payments) a Excluding pension payments.

1966

£m. 1967 1968 1969 1970 1971 1972 1973

11.3

8.6

9.2

8.8

8.5

6.6

7.0a

8.7

7.9

4.4

5.9

5.3

4.7

3.0

3.0

3.7

There are other costs, however, when procurement-tying is combined with project-tying (making aid conditional on the execution of approved projects) which may well be even more considerable. The commitment of aid funds to specific projects might well be delayed and there is a drain of administrative resources. As Clive Gray puts it: The additional costs of tied project aid in terms of administrative resources and delays in project execution are headaches familiar to anyone who has ever been involved in aid administration, on the side of either a donor or a recipient. To begin with, the aid administrators must identify and measure those components of a project which can reasonably be imported from the aid source, rather than leaving it to the more efficient procurement staff of a contractor, or simply to the workings of the free market. One should not under-estimate the drain on aid, and host government staff resources, involved in deciding questions such as what types of the myriads of standard components that go into a building can be purchased out of local distributors stocks as opposed to being ordered and shipped specially from the aid source country. Next, if the local economy is presently importing its requirements of the goods and services in question from sources with are ineligible for purposes of the project, new procurement channels must be developed, and the procurement must be supervised by aid administrators in both the source and recipient countries. The writer has repeatedly been struck by the high opportunity cost of the government staff time required to manipulate commercial trade channels and procedures in order to satisfy tied-aid requirements—staff time that, from an economic point of view would be far better spent evaluating projects more closely and administering their substantive aspects more intensively.13

Tying slows the commitment of funds, while high-import projects are found, then it slows disbursement of funds and the execution of projects. British goods are notoriously subject to long delays in delivery and the Kenyans may have to wait months for British equipment to be supplied when comparable goods are available immediately from the local agents of third-country suppliers. The tradition in Kenya of trading with Britain may mean, however, that some of the difficulties which Gray describes are less severe for tied British aid than for that of other donors. While the agreement to provide a proportion of local costs of any project represents an advance on simply providing the import content of projects, it is still a form of tying, although one that is applied with some flexibility. Local-cost provisions are usually used to finance labour costs and costs of construction. Where they include locally supplied goods it may be that the proportion of Kenyan value-added in the price of the goods

40 Aid and Inequality in Kenya: British Development Assistance to Kenya is low and the third-country import content consequently high. In addition, paying a local agent for goods off the shelf can count as part of the costs of a project tied to British procurement if the agent agrees to re-stock with British goods, but it is clearly hard to check this. Fortunately, ODM realises the difficulties which source-tying creates for recipients. Its maintenance is the result of British Treasury and, to a lesser extent, Department of Trade pressure. Clearly, in valuing British aid, the existence of, and changes in, the intensity of tying have to be considered. Some of the costs mentioned do not necessitate any modification of the figures of Table 7. Delays in disbursement, for example, simply mean that in the absence of tying, the figures would be higher. The important administrative costs are impossible to quantify. The extra paid for British goods over world market prices as a result of tying ought, however, to be deducted.14 Although an average of around half of British aid for general development has been tied to procurement in Britain, this cannot always have been a problem. Kenya naturally takes a high proportion of its merchandise imports from Britain—an average of around 30 per cent in the period we are considering. If British aid had not been project-tied, the procurement-tying might have been no problem at all. The money could perhaps have been used entirely for goods that would have been imported from the UK anyway, as it is hard to enforce stipulations that aid should lead to ‘additional’ imports. When aid is also project-tied, however, it may not be possible to spend all British aid on projects where British imports would have been used. Some might have to be spent on projects where other countries’ products would have been first choice, in which case the combination of project-tying and procurement-tying succeeds in raising British exports to Kenya. If the Kenya government would have imported for projects in a similar way to the economy as a whole—a very strong assumption—tying might result in a surcharge on a proportion of the imports for British-aided projects equal to the difference between the marginal propensity to import from Britain and the proportion of British loans that is procurementtied. It will be less of a problem in so far as Britain can be allocated projects for which its exports are competitive. On the other hand, that ignores the possibility that British firms may tender one price when competing freely with foreigners for an order, but offer quite another when they know they are faced with a captive market owing to aid-tying. With these qualifications in mind, we can guess that tying might be a real constraint for about 15 per cent of gross British general development aid and the goods it buys are about 20 per cent overpriced (the UNCTAD estimate). That would mean the value of the 50 per cent of British capital aid that has been for general development is overstated by some 3 per cent. The overstatement of total gross aid including LTP and land adjudication funds is therefore less still. The overstatement of net aid for the period is greater as reverse flows of interest and amortisation to Britain consist of free foreign exchange. Although technical assistance is tied, this is unlikely to be so much of a problem. It is unlikely that Kenya could get suitable people more cheaply by hiring on the world market. It may well be that people work in Kenya for less when they feel the job has the backing of their own government than they would if hired directly by the Kenya government.15 However, some jobs, especially consultancies, could probably be done more cheaply and efficiently if technical assistance were not tied, and the donor hired anyone qualified.

The Facts of Aid

41

The cost of British aid-tying to Kenya therefore appears very modest—perhaps as low as 1 per cent and very unlikely to be more than a 5 per cent overstatement of the value of the aid. This, however, is not counting the inevitable administrative costs and delays which are probably more important, The effect of the parity of Sterling on the value of British aid cannot be separated from the problem of tying. That proportion of aid that is for local costs is in effect free foreign exchange. The Kenyan government is committed to making payments to residents in Kenyan shillings but it then receives an equivalent amount in Sterling to add to its foreign exchange reserves. The injection of money into the domestic economy gives rise to extra import demand which is financed by the additional foreign reserves, That proportion of local-cost aid (denoted in Sterling) which would finally finance non-Sterling imports is reduced in value by the full amount of the devaluation, but the proportion itself might be altered by any relative price movements caused by the devaluation. The value of that proportion of local-cost aid that is eventually spent on British goods is, however, not affected by the devaluation. (Clearly devaluation is bound to affect the internal British price level and the cost of exports, but this effect is covered by deflating aid flows by a price index.) Similarly, the value of tied aid is not reduced by the devaluation per se. Indeed it should be increased. The reason is that if a devaluation of Sterling makes Kenya want to import more goods from Britain, any given degree of tying of aid may become less onerous following a devaluation. There is, however, no evidence that the Sterling devaluation of 1967 did much to affect the proportion by value of Kenya’s imports coming from Britain,16 and as the direct pecuniary costs of tying (as opposed to the costs of administrative demands and delayed disbursements) are modest anyway in relation to total aid flows, it is doubtful if this beneficial effect of a devaluation is significant. The effect of the devaluation for practical purposes is therefore confined to that proportion of local-cost aid that would eventually have been used to import non-British goods, i.e., we should subtract from British aid flows for 1968 and subsequent years a proportion of the total equal to 1×0.143 (devaluation)×0.45 (approximate proportion of aid that is untied)×0.70 (approximate proportion of imports not from Britain). This is a mere 4.5 per cent. So if capital aid flows are written down by up to 5 per cent to allow for the effect of procurement-tying, it is only necessary to deflate the 1968–73 figures by some four and a half per cent in allowing for devaluation. And for net aid the adjustment should be even less because the devaluation reduces the debt burden of repayments in Sterling. Consideration of devaluation, therefore, fortifies the conclusion that aid has been falling in real terms, but hardly makes a dramatic difference. In 1975, there were two departures in UK aid policy towards Kenya, representing a softening in terms. Owing to Kenya’s balance of payments difficulties following the oil crisis, Britain made a balance of payments support grant of £2.5m. This was nominally for British goods but was untied to any project. Later in the year, ODM announced that future aid to countries with an annual income per head of $200 or under would be entirely in the form of grants. Kenya is among those countries.

42 Aid and Inequality in Kenya: British Development Assistance to Kenya

British Technical Assistance Table 8: British technical assistance disbursements to Kenya and the East African Community

To Kenya To East African Community

1963

1964

(£’000) 1965 1966 4849 3527

2604 3329

1967

1968

1969 2733 2390

1970 3271 2340

1971 2891 2603

1972 4499 1972

1973 4741 2972

Total 1963–73 39688 29639

3887 2823

2323 3123

To Kenya To East African Community Sources: British Aid Statistics 1963–67 to 1968–72.

3815 2205

4075 2355

Table 9: British technical assistance to Kenyaa: by recipient ministry at 31 March 1974 Agriculture Commerce and industry Finance & Planning Education Natural Resources Health Home Affairs (Police, etc.) Information and Broadcasting Labour Attorney General’s Office Judicial Department Power & Communications Office of the President Land and Settlement Tourism & Wildlife Works TOTAL Source: Ministry of Overseas Development.

OSAS 54 1 17 559 5 25 42 4 2 8 14 4 23 13 1 74 846

Advisersb 18c 1 3 4 – 1 2 – – – – 3 – 1 1 – 34

% of total 8.2 0.5 2.3 63.8 0.3 2.9 5.0 0.5 0.2 0.9 1.6 0.8 2.6 1.6 0.2 8.6 100.0

Excluding EAC. There are 362 additional OSAS personnel with EAC, most of them serving in Kenya (Harbours 46, Railways 149, Post and Telecommunications 102, General Financial Services 65), and 3 advisers. b This figure includes some advisers not strictly attached to a ministry but to other organisations, e.g. Tana River Development Authority. In such cases the expert has been imputed to the ministry appropriate to his work. c This figure excludes personnel serving with research schemes, mostly in the field of agriculture and natural resources. There were 66 of these. a

Technical assistance is a very important part of aid to Kenya, which gets some 45 per cent of aid disbursements in this form, and Britain is much the most important supplier. British technical assistance personnel can be classified in three groups. There are volunteers

The Facts of Aid

43

and experts, the latter being further divided into wholly- and partly-funded personnel. Volunteers are not part of the official aid programme, and we do not consider them further. The wholly-funded personnel are usually advisers to the Kenya government or a parastatal organisation whose salary is met entirely by Britain. They work for a contract period and do not have established posts in the Kenyan administration. They are usually specialists with a particular, more or less technical, expertise rather than managers or administrators. The terms and conditions of their service are covered by the Special Commonwealth African Assistance Plan (SCAAP).17 Partly-funded personnel are British nationals who hold established posts in the Kenyan administration. The Kenya government pays them a ‘local’ salary and the British government adds an inducement allowance or ‘topping up’ to UK levels. The terms and conditions of these personnel are covered by the Overseas Service Aid Scheme (OSAS) and the British Expatriates Supplementation Scheme (BESS). At independence, there were large numbers of British people working in the Kenyan administration. The OSAS scheme provided a device whereby they were induced to stay and others were recruited. It provided a means for HMG to support Kenya’s general administration at a time when there would have been grave difficulties for the independent government in finding adequate executives and administrators to man the civil service. In the early independence years, therefore, most British technical assistance was OSAS and many of the people concerned were involved in the general running of government departments. BESS supports British nationals employed outside the central government in local authorities, universities and state corporations. While OSAS people are still more numerous than other technical assistance, their numbers have declined more or less steadily. There were about 1,600 OSAS personnel in 1965 compared with about 1,000 now. The numbers doing general administrative work under the scheme have gone down much faster than OSAS numbers as a whole. One group, teachers, increased in absolute numbers for eight or nine years after independence, and has begun to fall only in recent years. There has been no clear trend in the numbers of expert advisers supplied under the SCAAP scheme, which was roughly the same in 1965 and 1973, although it had been much higher in between. Disbursements on British technical assistance are given in Table 8, and the numbers of British technical assistance personnel by Kenyan ministry in Table 9. Britain still supplied over half of expatriate personnel in Kenya in 1973, but only two years before, the proportion was as high as 64 per cent. The cash value of Britain’s technical assistance effort is, however, a lower proportion of the total as the areas in which British personnel work are rather atypical, There are more relatively low-paid teachers and fewer highly paid expert advisers.

The Commonwealth Development Corporation The statistics in Table 6 include investments by the Commonwealth Development Corporation (CDC). Although this body has come under the authority of the Minister of Overseas Development since 1965 and it lends for projects in less developed countries at concessionary terms, its activities are regarded as separate from the aid programme proper. The CDC was established by Act of Parliament in 1948 as the Colonial Development Corporation to assist in the economic development of dependent territories and took its

44 Aid and Inequality in Kenya: British Development Assistance to Kenya present name in 1963. It operates on commercial lines and has a statutory obligation to pay its way taking one year with another. CDC obtains money on loan, largely from the UK Exehequer from which it can obtain £205m. of a total statutory borrowing capacity of £225m. The Corporation may negotiate any interest rate but generally aims to lend at some two or three per cent above its Treasury borrowing rate. Most of its loans are currently made at interest rates of 8 to 10 per cent although for agricultural projects it can lend at 5 to 6 per cent. It can also take equity in projects. The Corporation’s investment in Kenya has been substantial. The book value of its outstanding investment there in 1973 exceeded £30m. At independence it was just under £10m. in current pounds. The lending of CDC is not subject to the Country Policy Paper18 which guides the allocation of ODM funds, but like the official aid programme CDC has sought to concentrate funds in agriculture. At independence CDC was already involved in mining, electricity and water supply, housing, hotels and engineering. It held equity in hotels and manufacturing concerns; it owned the Kenya building societies and had made loans inter alia to the Kenya Power Company (£3.5m.), the Kenya Meat Commission (£250,000 with £135,000 outstanding) and Nairobi City Council for water works and housing schemes (£250,000). It was committed to providing finance, a manager and accountant-secretary for the Development Finance Company of Kenya.19 CDC was also involved at independence in the beginnings of its most successful venture in Kenya, the development of small-holder tea-growing which, from 1954, had been developed to 1,600 acres and one tea factory by the Kenya government on a proving basis, using funds derived from the British government grant towards the development of African agriculture. In 1960 CDC committed £900,000 to the Special Crops Development Authority, redesignated the Kenya Tea Development Authority (KTDA) in 1964, to develop small-holder tea-growing and lent some £480,000, two-thirds of the cost of six tea factories, to process the small holders’ output, West Germany provided £212,000. A second KTDA scheme increased the number of factories to seventeen by the end of 1970 and a third brought total tea acreage planted to 64,500 acres by the end of 1972. By then CDC had committed some £1.35m. for field development and the IDA had committed £ 1.875m. Neither of these loans was fully drawn as crop yields and cesses exceeded estimates and KTDA costs were kept down. CDC had also contributed a substantial part of the finance for ten of the factories in operation, some £800,000. In 1973 the Corporation agreed support for three more factories and committed nearly £7m. for the Fourth Plan to expand acreage by a further 36,000 acres and for the construction of an additional twenty-two factories. The factories in operation are partnerships between CDC, the KTDA and commercial companies.20 Under the Fourth Plan a factory management training and marketing department has been set up within KTDA which will design, arrange construction and undertake the duties of managing agents for future factories. The success of the scheme in developmental terms lies in its involvement of large numbers of poor African small holders, over 90,000 of whom now supply tea to the factories that CDC has helped finance. Teagrowing brings them an income of perhaps £60 a year compared to about £5 a year that many earned previously. Prior to 1954 tea had been regarded exclusively as a plantation crop and the demonstration that highquality leaf could be grown by small holders was very important. CDC claims that the scheme has become a model for the successful organisation of similar projects in other tea-growing countries.

The Facts of Aid

45

The operation is a valuable source of foreign exchange to Kenya as almost all KTDA output is exported. In 1973 it accounted for 18.1 per cent by value of Kenya’s exports (up from 12.9 per cent in 1963). A social cost-benefit analysis of the Third Plan found that it had very high rates of return—over 38 per cent. The return to the small holders, counting their own and family labour as a cost and imputing a cost to the land, was rather less—13 per cent; but their rate of return in cash terms was very high—over 40 per cent in some areas.21 CDC is now involved in the Mumias sugar factory which also received official British aid.22 Other large schemes include the Bura Bura housing estate of some 5,000 houses in Nairobi. CDC has agreed to commit a £1.5m. revolving fund for construction and £3.5m. for mortgage loans to buyers, for a first phase of the project, and equivalent sums of £1.44m. and £ 1.75m. for a second phase. Part of the CDC interest in agriculture takes the form of investments worth a total of £200,000 in two estates producing largely horticultural crops for export, and owned by British and Kenyan business interests. Older investments include £3.5m. lent in 1969 to the Tana River Development Authority for construction of the Kindaruma power station and a loan of £2.6m. to the East African Power and Lighting Company Limited for transmission lines.

Aid From Other Sources This section gives brief descriptions of the activities of some of the important donors to Kenya shown in Table 3. A discussion of the policies behind these donors’ programmes follows that on British aid policy in Chapter Four.

United States Until quite recently the overwhelming bulk of US aid was either technical assistance or food aid. Between 1962 and 1972 the US Agency for International Development (AID) extended grants of $29.9m. to Kenya. These were almost all for technical assistance, although there were some small ‘commodity’ grants supporting the technical assistance and fully tied to US procurement. Aid was overwhelmingly in the field of agriculture, with a smaller number of advisers in education and vocational training. Only two capital loans were extended during this period: $2.2m. in 1964 to improve the Nairobi water supply and $ 3.5m. in 1970 for vehicles and construction equipment for the Kenya National Youth Service of the Ministry of Labour. Apart from AID’s activities, $30.6m. was extended up to 1972 under ‘Food for Peace’ schemes, just over half of which was classed as emergency relief. The remainder was food loans repayable in US dollars. Volunteer assistance by the Peace Corps was recorded at $12.3m. for the period. Since 1972 the US has become a more active donor. Technical assistance has continued at roughly the same levels with back-up in the form of vehicles and other equipment. Grants totalled $2.2m. in 1973. Teams of advisers are operating in the areas of livestock and range development, agricultural credit administration, rural development and family planning, among others. In 1973 the US also extended a programme loan of $10m.—‘to help fill Kenya’s recently emerged resource gap and maintain development momentum’. The counterpart funds realised by domestic sales of imports financed by the loan are to be used for the Kenyan development budget. A further loan of $ 10m. was extended in 1974

46 Aid and Inequality in Kenya: British Development Assistance to Kenya for a livestock development programme. The money was to finance ranching development and abattoir and cold storage schemes as part (about one-fifth) of a larger livestock development programme to which the World Bank, UK and Canada, among other donors, were contributing.

West Germany Since Kenya’s independence West Germany has committed £23m. (DM 189.9m.) in capital aid to Kenya and succeeded in disbursing £9m. (DM 76.8m.) of it.23 The largest single project, accounting for over 20 per cent of total commitments, was the Chemelil sugar factory which also received UK aid. Provision of tourist roads accounts for almost another fifth. Germany has concentrated more on industrial development than most other donors. About 13 per cent of its commitments have been to the Industrial and Commercial Development Corporation of Kenya to help finance industrial estates at Nairobi, Nakuru and Mombasa. In the agricultural field, nearly eight per cent of total capital aid commitments have been for a rice cultivation project and further sums have been committed for a pilot irrigation scheme and for a scheme to provide credit to small farmers. In sarlier years, Germany provided money for the promotion of tea-growing. The most recent large commitment, nearly 10 per cent of the total, has been for water supply and sewerage to Nakuru, Kisumu and three other provincial towns. All German capital aid is loan and the current terms are 30 years maturity with a 10-year grace period and 2 per cent interest. Technical assistance is grant. It has been worth £4.5m. (DM 36.2m.). It is quite widely spread across economic sectors, often associated with German capital projects. A higher proportion has gone into the agricultural sector than is the case with financial aid. The aid is not source-tied. However, many capital aid projects, especially in earlier years, were import-intensive and used goods which German industry was well able to supply. The Germans were also reluctant to finance local costs out of a desire to involve the Kenyans in all projects and ‘make them feel responsible for them’. More recently a more liberal attitude to local costs has evolved.

Sweden In the period 1964–72 the Swedes disbursed only £10.5m. in aid to Kenya but in 1973/74 and 1974/75 they committed sums of £5m. to projects. For the first ten years the sectoral breakdown of Swedish disbursement was: public administration and other services, 35.3 per cent, agriculture 25 per cent, education 18 per cent, water and power supply 10.7 per cent. The 1974 disbursement pattern was: agriculture 22 per cent, education 19 per cent, health, including family planning, 15 per cent, rural water 13 per cent, import financing 10 per cent. Most Swedish capital aid used to be loan, although on very soft terms with 25 or 50 years to maturity, ten-year grace period and either 0.75 per cent or 2 per cent interest charges. More is now grant; some three-quarters of current commitment is grant and the remainder is interest-free 40-year loan. The money is all untied in respect of procurement. Even technical assistance funds are untied in principle although Swedes do tend to be recruited. Like the US and UK, Sweden made a balance of payments loan to the Kenyans in 1974 to counteract the deterioration in the world economic situation. This is supposed to be for

The Facts of Aid

47

imports for rural development, but the Swedes realise that that is semantics and free foreign exchange is just that. Uniquely among donors to Kenya the Swedes disburse their aid in advance of expenditures by the Kenyans and merely require subsequent documentation to show how the money was spent.24

The World Bank Group The World Bank group has long been one of the largest donors to Kenya. Between independence and 1975 it had committed $ 144.8m. in IBRD loans, $ 122.8m. in IDA credits and $ 25.9m. in IFC investments25 for a grand total of $ 293.5m. The IBRD also lent a further $ 229.8m. to the East African Community. Much of this investment has been in large capital-intensive projects, and continues to be so although the Bank now lays emphasis on rural development in its policy pronouncements.26 An examination of the Bank’s lending to Kenya since 1970 illustrates the difficulty of changing the pattern of loan disbursements away from infrastructure. Less than a fifth of commitments were to agriculture. In 1974, however, the influence of the new policy became clear with most of the money going to rural-based productive enterprises. Table 10: World Bank commitments to Kenya 1970–74 Loan

Education Pulp and Paper Mills Nairobi Water Supply. Highway Maintenance Tana River Development Co. (power) Highways Tourism Promotion Services Nairobi Airport Agricultural Credit—second loan Roads Industrial Development Bank Kenya Hotel Properties Ltd. (Tourism) Pulp and Paper Mills Population Project Tea development—third loan Livestock development—second loan

Year 1970 1970 1970 1970 1971 1972 1972 1972 1972 1973 1973 1973 1974 1974 1974 1974

IBRD 8.3 23.0 29.0 29.0 5.0

10.4

$m

IDA 6.1

IFC 15.81

12.6 22.0

2.42

6.0

12.0

2.83 1.77

21.5

The Costs of Aid It is sometimes argued that much aid is swallowed up in a bureaucracy of its own creation. We consider here some of the costs of being an aid recipient. The focus is quite narrow and we are not, for the moment, concerned with any distortions in policy that aid may be responsible for, but purely with the direct costs of administering the receipt of aid and direct expenditures associated with technical assistance. We consider the latter first. The receipt of technical assistance is associated with expenditures by the Kenya government, which often cites them in an attempt to get donors to foot more of the technical

48 Aid and Inequality in Kenya: British Development Assistance to Kenya assistance bill. Some of these expenditures are real costs of technical assistance in the sense that if there were no technical assistance they would not be incurred. Other expenditures would be incurred to some extent if no technical assistance were received. They cannot really be viewed as costs attributable to technical assistance. Total counterpart expenditures, fresh costs or not, are considerable. By and large, all donors expect similar facilities for their personnel. Advisers get a house or housing allowance or pay no rent. The Kenya government meets internal travel and transport costs and provides office facilities. Advisers can import one car duty-free. For operational personnel the Kenya government pays a local salary which is liable to income tax. OSAS personnel receive part of their gratuity payments, pensions, or compensation from the Kenya government. Operational personnel from Denmark, Norway and UNDP are also paid gratuities. Astudy in 197027 worked out that counterpart expenditures to Kenya were just under £4m. for a total technical assistance disbursement of £9.5m. Eighty per cent of this sum was on local salaries and housing. If anything, housing costs may have been understated because most expatriates stay in Nairobi where costs are higher than the national average. The 20 per cent of expenditures not salaries and housing, however, was overstated because costs were imputed to technical assistance not only for office space, secretarial support, and other back-up facilities but also for their share of the Kenya government’s total social overhead expenditure, for example on public utilities. Both these sets of costs, however, were worked out by taking average spending per head for relevant groups in the country, whereas the marginal cost of the expatriates concerned in the case of most expenditures was almost certainly less than the average cost per head of the services. The receipt of one more expatriate does not require extra water supply for Nairobi, for example. In 1970 the local expenditure on British advisers per head was roughly £2,380 and on OSAS people £2,790. If the Ministry of Education was excluded (mainly teachers), the latter figure rose to £3,740. How much of these expenditures can truly be called costs depends on what arrangements the Kenyan government would have made to fill posts if technical assistance had not been used. If local people would have been recruited, part of the housing expenditure on expatriates is a cost Some Kenyans might arrange their housing privately and in general they would house themselves at less luxurious levels than expatriates. Also in a poor country like Kenya there is a difficulty in raising tax revenue, so public money may be worth more than private. The guesstimate for the pure administrative cost of receiving technical assistance, including preparing job specifications for donors and meeting and entertaining them, was about £10 a head, which seems on the low side. Most of that is presumably agenuine cost, although some job specification would have to be done in any event. At any rate such administrative costs were less than 0.5 per cent of counterpart expenditures. If competent Kenyans are available for posts the Kenya government should not request technical assistance, so the alternative to it is generally doing without the services involved or recruiting abroad. In the latter case practically none of the counterpart expenditures are true costs of technical assistance. However there is some suggestion that sometimes Kenya government departments apply for technical assistance because they do not have adequate budgetary allocations to take on the staff they require even when people are locally available.28 That would presumably apply mostly to jobs done by wholly-funded personnel because if a budgetary allocation is available for the local salary of an OSAS officer it should be available for a Kenyan, although it could apply to established posts if the salary required to

The Facts of Aid

49

recruit a competent Kenyan exceeded the going rate of pay. Surprisingly, in view of Kenya’s school-leaver unemployment problem, we gathered that the Kenya government did indeed have difficulty in recruiting for many jobs at existing pay rates because of private sector competition. Technical assistance might be made more cost-effective in some cases by allowing Kenyans to compete for OSAS posts. The true direct costs of technical assistance in 1970, therefore, were between nil and the cost of housing, administration, and gratuity payments (some £2m.), the precise sum depending on how many posts could have been filled adequately from local manpower. The pecuniary costs to Kenya of administering aid are tiny in relation to the sums received. Total expenditure on the Capital Aid and Technical Assistance Division of the Ministry of Finance and Planning in the Kenyan fiscal year 1973/74 was £41,700 and the estimate for 1974/75 was £53,000.29 These are not the only expenses as quite a lot of the time of senior personnel in the headquarters administration of the Ministry is taken up with aid matters. It is very difficult to evaluate the time of a Permanent Secretary in a country like Kenya where administrative skills are scarce; it would be preposterous to suppose that his salary really measured the opportunity cost of his services. Aid negotiations do take up an appreciable amount of the time of a Permanent Secretary (salary in 1975, £4,940) and two Deputy Permanent Secretaries (combined salaries £9,490). In addition about two-thirds of the headquarters administrative services division’s travel and entertainment expenses are incurred in the course of aid dealings, i.e. about £6,700 in 1973/74 and an estimated £10,000 in 1974/75. Operational ministries, owing to the policy of the Ministry of Finance and Planning, are frequently involved with aid. The total cost of being in receipt of aid could well exceed £100,00030—the cost of a modest project. This is very low in comparison with the sums disbursed under aid.31

Notes 1. Economic Survey 1974, Table 10.5, p. 151. 2. J.R.Nellis, ‘Expatriates in the Government of Kenya’, Journal of Common-wealth Political Studies, Vol. XI, No. 3, November 1973, pp. 251–64. 3. See Chapter Four, pp. 89–90. 4. Economic Survey 1966, Table 8.11. 5. Underspending is considered further in Chapter Four, pp. 88–93. 6. These percentages were compiled by considering projects one by one and allocating them to a sector. They are necessarily somewhat arbitrary and approximate. An agricultural college, for example, could be classified either under agriculture or education (social services). We classified it under education but many similar border-line decisions are involved. 7. See below, pp. 63–5 and 99–100. 8. This is difficult to do for yearly British disbursements because different aid agreements with Kenya have resulted in loans on different terms (see Table 5), and disbursements have been delayed so that money might be extended on various terms in the same year. The terms of the current British loan agreement, 2 per cent interest, 25-year amortisation with an eight-year grace period, have a grant equivalent of over 80 per cent, if future charges are discounted at 10 per cent—the rate used by the Development Assistance Committee (DAC) of OECD. DAC records the grant equivalent of all bilateral aid to Kenya at 85.7 per cent in 1972 (Development Cooperation 1974 Review, Paris, 1974, p. 278). With inflation now above 20 per cent in Britain and likely to stay in double figures for some time, 10 per cent is a very modest discount rate. A

50 Aid and Inequality in Kenya: British Development Assistance to Kenya

9. 10. 11. 12. 13. 14.

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

29. 30. 31.

more appropriate one would give a grant equivalent for current loans of over 90 per cent. Future British aid is to be entirely grant. Aid flow deflator from Table 11, p. 209, Statistical Appendix of Development Co-operation, op. cit. Modified by dividing 1968 and subsequent values by 1.143 to remove effects of devaluation, and changing the base year to 1964. See Table 5. See Proceedings of the UN Conference on Trade and Development, Feb-Mar 1968, Second session, Vol. IV. Problems and Policies of Financing. Bhagwati, in UNCTAD, op. cit., ‘The Tying of Aid’, para. 8, p. 46. Quoted from Clive Gray, ‘Tied versus local cost foreign aid’, East African Economic Review, Vol. 3, No. 2, December 1971. There are the same difficulties in calculating the extent of tying of any disbursement as in calculating its grant equivalent. As the amount of tying has varied between different agreements, it is hard to tell precisely how much tying is involved in any year’s disbursement. Furthermore, the proportions tied are average figures. While, say, 50 per cent of a given loan must be spent in Britain, a single project financed by part of that loan can have less or more than 50 per cent import content. The costs incurred by ODM in advertising for and interviewing personnel are counted as part of the aid programme, so there is no direct subsidy here. The average proportion of Kenyan imports coming from Britain for the three years 1965–67 was 31.7; for the three years 1968–70 it was 30.6, which suggests a price elasticity of less than one, but clearly other factors were at work, and we cannot infer much from that. See Chapter Four, p. 82. See Chapter Four, p. 79. Other subscribers were the Kenya government, the West German development agency, and the Dutch aid agency. There was some public subscription. CDC’s 1973 commitment of £1,278,000 represents about a quarter of the DFC’s finance. Brooke Bond Liebig Kenya Ltd., Eastern Produce Africa Ltd., James Finlay & Co. Ltd., George Williamson Kenya Ltd. N.H.Stern, An Appraisal of Tea Production on Smallholdings in Kenya, OECD Development Centre, 1972. See Chapter Five, Part Two. The Sterling-Deutschemark exchange rate has been moving in favour of the latter for much of the period so Sterling equivalents are approximate. See Chapter Four, p. 97. IFC is the bank subsidiary that undertakes indirect (portfolio) investment. See Chapter Four, p. 98. Conducted by the Commonwealth Secretariat (unpublished). That raises the question of what such people are doing if not employed by the government. The answer probably iles in the phenomenon of ‘bumping’. Qualified people might be hired from the private sector, to leave jobs there to less qualified people (or possibly expatriates brought in by private firms—this cost eventually borne by shareholders). This sets up a chain reaction with people being upgraded, and at the lowest level some of the urban unemployed obtain unskilled work. This and all other figures in this section are from the Government of Kenya Estimates of Recurrent Expenditure 1974/75. Taking the view that the time of senior administrators is worth some multiple of their emoluments. We shall argue (Chapter Six), however, that aid has had a deleterious effect on Kenyan administration, giving rise to heavy costs in a more general sense.

4 BRITISH AID POLICY AND ADMINISTRATION IN KENYA

British aid policy, towards Kenya as elsewhere, exhibits a dichotomy. Decisions about the allocation of aid between countries are strongly influenced by precedent and original allocations were usually subject to political and commercial considerations. These may also be important in determining some of the general features of the aid a country receives, how far, for example, aid is tied to specific projects and how far it is tied to procurement of British goods. When a sum has been agreed, however, if it is to be tied to projects, as it generally is, developmental issues become much more important in determining which projects receive assistance. That is noticeable in the case of Kenya. While commercial and political pressures are sometimes brought to bear on individual project selection they appear to have become less influential at this level over the past decade. British project selection in Kenya has become more oriented to development, as it is conceived within the Ministry for Overseas Development (ODM). The changes reflect the varying importance of different Ministries and Departments of State at the different stages of aid policy formation. The institutional arrangements within the UK government machine have changed since aid to Kenya began and with them has changed the importance of different strands in aid policy. An analysis of aid policy, therefore, must be partly historical.

The Early Years, 1963–69 No great decision was involved when the UK began to aid Kenya on the latter’s independence in 1963. As a colony, Kenya had received funds from the British government. Grants in aid of both cash and technical assistance were annual features of the Kenya budget. Almost £5m.1 was given in 1961/62, excluding payments for military forces—worth a further £1.9m. The Colonial Development and Welfare Act (CD&W) of 1945, and its precursor the Colonial Development Fund Act of 1929, provided the Colonial Office with funds to be spent on development programmes in the colonies, and Kenya received its share of these. In 1961/62 it received development grants and loans of £6.8m. By 1963 the policy of continuing payments to independent Commonwealth countries was well established. At the Commonwealth Finance Ministers’ Conference in Montreal in 1958 a policy statement expressed Britain’s readiness to continue financial aid to poor independent Commonwealth countries. This policy had been applied in the cases of many countries reaching independence before Kenya, such as Ghana in 1957 and Nigeria in 1960. In the circumstances of the time, therefore, a decision not to aid Kenya would have been incomparably more significant than one to do so. Britain did have, however, particular interests in Kenya which dictated some form of aid programme anyway and ensured that it would be larger than in other countries. The

52 Aid and Inequality in Kenya: British Development Assistance to Kenya most obvious was the large number of white settlers of British stock who remained in the country. Many of them were well-connected and found champions on the benches of the Conservative Party in Parliament. The aid which received the greatest attention at the Lancaster House Conferences which preceded independence was that for the Land Transfer Programme (LTP).2 The only other finance discussed at Lancaster House in any detail was assistance for the independent Kenya government in taking over the cost of military forces. Just over £1.1m. was loaned for that purpose. These two preoccupations show the nature of British concern with independent Kenya which was chiefly that law and order should be preserved and British property protected. Entrenched clauses were included in Kenya’s independence constitution against expropriation without compensation. It has been argued3 that aid was one means of luring Kenya into a neo-colonial situation. On most definitions of ‘neo-colonial’that probably credits the British with too much ambition. Their aims were of a limited nature. The Conservative government of the day would certainly have preferred a Kenya government made up of members of KADU—representing in the main the smaller Kenyan tribes with little tradition of involvement in the Mau Mau rebellion and believed to have a more amicable approach to the former colonists than African nationalists in the other party, KANU, which was dominated by the Kikuyu and Luo. Yet when elections in 1961 and 1963 made clear the strength of KANU support, HMG did not attempt to install its own client government. It realised the folly of establishing any government that would be overthrown when the British presence was withdrawn, and political engineering was restricted to trying to commit KANU to the LTP.4 HMG had no great confidence in the KANU leader, Jomo Kenyatta; many people seemed to believe the rhetoric of former Governor Patrick Renison that he was a leader to ‘darkness and death’, and foreign investment plummeted at the approach of independence.5 Anarchy and ethnic strife rather than revolution was the fear. Kenya did not possess great strategic or political significance in Cold War terms. Despite anti-Mau Mau propaganda the British knew there was little evidence of Communist penetration, and although Oginga Odinga, the Luo leader and deputy leader of KANU, was thought to have Communist-bloc sympathies he was not taken seriously as a contender for national leadership. At that time there were British bases near both ends of the Suez Canal, at Aden and in Cyprus, and Kenya did not appear either a promising or a necessary place for a military base.6 It was hoped that the general friendliness of the new Kenyan government would be secured by military assistance (quite significant in a new state with fissiparous tendencies). In 1964, following an army mutiny, the assistance was extended to include arrangements to train Kenyan forces. The explosive issue of land was meanwhile anaesthetised by the LTP. Otherwise capital assistance was rather meagre. The independence settlement included a commitment of £34.2m. of aid. But £12m. was loan and grant for the Land Transfer Programme and almost all of it was the unissued balance of commitments made before independence; £4 3m. was the unissued balance of development grants and loans, again committed before independence; £13,6m. was a loan to enable the Kenya government to pay the pensions of British colonial civil servants and while it was right that the pensions should be paid, it was barely short of scandalous to class that as aid to independent Kenya; the civil servants were the responsibility of the British government and their pensions brought no benefits to Kenya. £1.25m. was a budgetary grant continuing colonial practice, but the last of its kind. Only £3m. was a new loan commitment of money for general

British Aid Policy and Administration in Kenya 53 development. More significant in the running of the new Kenyan state and in helping it to take the course it did was technical assistance. Expatriate advisers and civil servants who by their advice helped to delimit the range of policy options considered as real possibilities were probably more influential than promises of capital aid in shaping independent Kenya.7 At the end of 1965 there were 1,716 publicly financed Britons in Kenya; 1,674 were under the OSAS, of whom nearly half were teachers. Technical assistance accounted for around 40 per cent of gross British aid. The British continued to put small money on the Kenyan horse as it began to show signs of running strongly in the right direction—against the predictions of some political tipsters. The political conservatism of the Kenya government and its hostile attitude to nationalisation and free land distribution became apparent and the Sessional Paper No. 10 of 1965 contained impeccably moderate social-democrat expressions of policy.8 As business confidence revived, Kenya continued to provide a profitable base for foreignowned enterprises. The excess of Sterling Area (largely UK) investment income over private long-term Sterling capital flows to Kenya in 1965 and 1966 was £4.2m. and £3.4m. respectively, and much of the ‘inflow’ was re-invested profits of Kenya-based firms anyway. These were tiny sums seen in the context of the British economy and certainly not sufficient to make any great neo-colonial conspiracy worth while. The general course of the Kenyan economy was, however, modestly favourable to British interests and it was therefore natural for HMG to continue supporting President Kenyatta’s government. LTP remained a major portion of the aid committed in 1966/67 and general development aid did not increase in real terms. The Land Transfer Programme on the capital side and the OSAS scheme in the technical assistance field were the two chief pillars of British aid. In 1963 British aid was under the control of the Foreign Office and the Commonwealth and Colonial Office. There was a Department of Technical Co-operation in charge of technical assistance, jointly responsible to the two ministries. In 1964, the ODM was set up, headed by a Minister who for two years had Cabinet rank, and it took over responsibility for all aid. The effect of this for Kenya was not appreciable for some time. Britain entered a period of continual economic crisis, and the conflicting influences of more concern for development and reduced national circumstances were resolved by the British Treasury insisting that Kenya spend large proportions of British aid on British exports. Money for colonies under the old CD&W Act had not been tied to British goods and had been disbursed in advance against short-run projections of expenditure.9 For independent countries, procurementtying was introduced along with the principle of reimbursement of expenditures after they had been incurred. Britain often complained in negotiations with the Kenyans during those years that the projects they proposed for assistance did not include enough imports and pressed them to find more import-intensive schemes. This pressure was rationalised by pointing to the British balance of payments difficulties and the fact that British aid to other countries (including the other countries in East Africa) was tied two-thirds to British goods, and an exception could not be made for Kenya without creating ill-will. The Kenyans replied with pleas that their development plans should not be distorted in an import-intensive direction, and by pointing out the favourable trade balance that Britain enjoyed with Kenya. They argued that although money for land adjudication and settlement in Kenya was necessarily local-cost, some 60 per cent of money for land purchase found its way back to the UK.

54 Aid and Inequality in Kenya: British Development Assistance to Kenya The Kenya government could also argue with justification that the prevailing trade pattern meant that Britain’s trade benefited substantially from any money lent to Kenya, whether it was tied or not. During the middle of the 1960s almost 40 per cent of Kenya’s visible imports and an even higher proportion of her invisibles were from the UK. It has been estimated10 that, on the assumptions that any foreign exchange received by Kenya would be spent quickly (with no leakages caused by excess domestic saving or budget surpluses leading to an accumulation of foreign reserves), and that average and marginal import patterns were similar, Britain could expect to receive back in payment for exports between 40 and 50 per cent of any foreign exchange accruing to Kenya, and that within three years.11 Other aid donors knew the matrix of trading relationships was in Britain’s favour, and Britain’s insistence on tying her aid, according to the Kenyans, influenced unfavourably the policies of other donors. In fact with hindsight it is possible to see that aid-tying did not have some of the effects that the British hoped and that the Kenyans feared. The lack of competitiveness of British industry has ensured it a declining share of the Kenyan market over the past ten years despite aid-tying and other donors, notably the Swedes and West Germans, have unilaterally untied their aid at least formally.12 In other ways, British practice in the disbursement of the independence settlement grant and loan and the two subsequent loans was very helpful. The Kenyans published a Development Plan and Britain, having agreed in advance a total sum with the Kenyans to be spent over a number of years, lent money for expenditures consistent with the Development Plan. These were usually on equipment for quite small projects which were not scrutinised before aid was allotted in anything like the detail that has subsequently become the practice. There were a few large projects where consultants’ reports were a necessary prerequisite for aid disbursement but in general the technique of aid-giving was very similar to the former practice with CD&W finance. The three aid tranches helped to finance 31, 17 and 56 projects respectively. The smallest disbursement was only £7,200— equipment for Egerton agricultural college. The average disbursement size was only about £150,000. The Kenyans found the readiness of the British to provide finance for small unglamorous projects, to act in a sense as donor of last resort, a distinct advantage. It enabled them to spend money piecemeal that they might not otherwise have been able to spend, especially in view of the import requirements of British aid. Some of the projects were not conspicuously developmental. Police radio equipment, broadcasting equipment and prison staff housing were all aided. The former two, of course, had the advantage of a 100 per cent import content. ODM, however, was anxious to move away from simply giving annual apportionments of aid for equipment for a large number of projects. It was keen to tie aid to specific projects that could be examined in detail by its own experts. This, it thought, would enable it to ensure a sound use of funds. A corollary was that HMG after 1966 began to press the Kenyans to put forward fewer and larger projects in order to ease the administrative burden implied by such an interventionist approach. There also seemed to be a feeling that aid spent on lots of small projects was being dissipated and if it were concentrated it would have more ‘impact’. This policy was to lead to increasing underspending by Kenya which as a result had to identify large projects and prepare proposals including plans and evaluations to a level of detail it would not itself have bothered with, perhaps rightly in view of its shortage of skilled administrative personnel. The import requirement was more onerous

British Aid Policy and Administration in Kenya 55 when aid could not be spent in penny packets but more important were the administrative demands made by large projects. Towards the end of the period of the Number Three loan, 1966–70, and subsequently, the difficulties of finding large projects with a sizeable import content which yet met ODM’s ‘development’ criteria became serious. And the difficulties of finding the administration and expertise to conduct adequate evaluations and make plans for a project submission were worse. This tendency in aid policy wiped out many gains in administrative capacity as Kenyans settled down in posts and acquired experience following the outflow of expatriates in 1963.13

Aid Administration Today The Machinery After 1969 the planning and administration of British aid to particular countries became more systematic with the introduction of the Country Policy Paper (CPP).14 This was drafted mainly by economists attached to the relevant geographical department of ODM. It is an internal working document which describes the economy and policies of the country concerned and discusses the main constraints on development and other relevant factors in its relations with the UK. The CPP sets out the proposed amounts and forms of UK aid for the next five years—‘showing how this fits in with the recipient country’s development priorities and the likely future aid programmes of other donors, taking into account any political and commercial considerations which are relevant.’15 The CPP also indicates the main sectors of the economy towards which the aid should be directed and lays down relative priorities. It indicates, too, the proposed terms and conditions of the aid in relation to the recipient’s debt servicing capacity and the practice of other donors. Between 1969 and 1972, in the case of Kenya, the draft CPP was sent to the High Commission in Nairobi (BHC) for comments, particularly on the political, commercial and investment aspects. The draft was (and is) then circulated to the Treasury, FCO, Departments of Trade and Industry and the Bank of England. Their representatives meet ODM as necessary to settle any differences of view which arise. The commercial departments and the Treasury are thus consulted at an early stage of the formation of aid policy to particular countries, and Kenya is no exception. Such meetings result in the proposed aid programme to a particular country being fitted into the overall Aid Framework—the detailed breakdown of the UK’s expected aid disbursements of all kinds for four years ahead, the years covered by HMG’s Public Expenditure Review. The figures in the aid frame for a particular country showing planned disbursements for a number of years ahead will reflect expected spending of funds already committed in past agreements with the recipient country, and will also influence the extent of future commitments. While policy towards a particular country is devised at ODM in consultation with other government departments, the British Mission in the country concerned has some responsibility for administering the aid programme. BHC Nairobi has two First Secretaries dealing with aid, and up to the end of 1974 one was in charge of technical assistance and one in charge of capital aid including both general development and LTP. There was then a reform which left one First Secretary in charge of aid with the other subordinate to him in charge of aid administration, A third First Secretary is now responsible for LTP,

56 Aid and Inequality in Kenya: British Development Assistance to Kenya reporting directly to the Deputy High Commissioner, The two First Secretaries are directly responsible to the Economic and Commercial Counsellor who at present is also Deputy High Commissioner. Project proposals for capital aid are made by the Kenyan government to the appropriate First Secretary in the High Commission. Up to 1972 the applications were then communicated, with the comments of BHC personnel, to ODM in London. Requests for over £400,000 have to go before the Projects Committee—an internal ODM body chaired by a Deputy Secretary. The Committee also considers smaller applications if they are not in line with policy expressed in CPP. Project submissions to the Committee are circulated to other government departments. It is, apparently, unusual for them to make strong representations at this stage as most of their leverage is exerted at the earlier stage of drafting the CPP. After 1972 the situation was changed by the setting up of a Development Division for East Africa in Nairobi. This is an out-station of ODM and consists of a number of ODM experts in various fields who act as advisers to the High Commissions in their region. They are in a position to carry out on-the-spot appraisal of project proposals rather than these having to be sent to ODM in London, or an ODM team having to fly out to Nairobi. On the advice of the head of the Development Division, the High Commissioner was given delegated authority to commit sums of up to £250,000 to projects without reference to London as long as they are in accord with the policy outlines of the CPP. In 1975 this was increased to £400,000. The economists in the Development Division have also become responsible for preparing the first draft of annual revisions of CPP. The setting up of the Development Division has had a number of very beneficial effects on the Kenya aid programme. It enables projects of less than £400,000 to be authorised with much less delay and introduces more flexibility into the aid programme; it speeds up the whole business of project appraisal; and it greatly strengthens the hand of ODM if there are inter-ministry disputes about the desirability of certain projects. The FCO can no longer claim a monopoly of local up-to-date experience, so the political considerations weigh less heavily compared with the developmental.

Technical Assistance Administration The policy towards technical assistance, as well as capital aid, is set out in the CPP, and the Development Division advises on it. The administration of technical assistance16 is, however, rather different from that for capital aid. There are, in the first place, the two quite separate sets of arrangements for wholly- and partly-funded personnel. To obtain OSAS personnel each Kenyan ministry makes a ‘bid’ to the Kenyan Public Service Commission giving requirements for expatriate officers with projected localisation dates of posts. These are vetted and passed on directly to ODM in London. ODM used to be very permissive about meeting these Kenyan requests but it discovered that the rate of localisation of posts by Kenya was less than it expected. Consequently the OSAS commitment looked like running on, scarcely diminished, indefinitely. In 1970 an annual manpower review was instituted. The Kenya authorities are asked to forecast for three years ahead their needs for expatriates and to explain what they are doing about localisation of each post. This means there is scrutiny of requests both for new officers and the continuation of old ones. Every other year an ODM team from London has visited Kenya to assess the needs and requests.

British Aid Policy and Administration in Kenya 57 The latest review was the most thorough ever, involving considerable discussion with the Kenyans. After reviews, ODM sets a ceiling number of personnel for each ministry or organisation. These ceilings need not be lower than the Kenyan bids; that depends on ODM’s view of requirements. However, ODM is trying to phase OSAS out over a longish period, perhaps ten years.17 The aim is to eliminate lower-level posts first. Teachers, mainly in secondary schools, have been a considerable part of the OSAS complement and ODM has been bearing down on their numbers.18 This is a source of continuing contention between the two governments as the Kenyans wish to retain as many teachers as possible owing to the rate of expansion of their secondary school system. This battle, like the ODM ceilings, is sometimes otiose as the planned numbers cannot be recruited anyway. While the OSAS numbers have been falling, the HMG contribution has been rising. The British supplementation is intended to cover the difference between the Kenya government’s local salary and what the equivalent post would pay in the UK, plus a local cost-of-living allowance, inducement to serve abroad, and an element in lieu of employers’ pension contributions. This supplement is non-taxable and payable in the UK. ODM calculates that its share of the costs of an expatriate under OSAS is now 60–70 per cent. This means they are cheap labour to the Kenya government which, as noted in Chapter Three, might well have trouble filling some posts with Kenyans while paying a local salary perhaps 60 per cent below what the expatriate gets. Up to now BHC and the Development Division have not been directly involved in administering OSAS, at least in theory. In fact, much time of BHC officers has been taken up with OSAS matters, acting as assessors and go-betweens and dealing with personal problems of OSAS people. Now local officers in BHC and the Development Division are to take over responsibility for the manpower planning reviews, and play a greater role in formulating ODM policy towards OSAS in Kenya. They may have some delegated decisionmaking power on individual applications beneath the set ceiling. Such developments would be entirely consistent with the drift of ODM policy towards decentralising decisions and, in our view, could speed up administration and perhaps improve decisions. For advisory posts of limited duration, as opposed to an established post with executive duties, ODM supplies people under the Special Commonwealth African Assistance Plan (SCAAP), the details of which were renegotiated in 1974—after seven years of discussion. Requests for this type of technical assistance are made by Kenyan operational ministries to a special division in the Ministry of Finance and Planning. If acceded to by the Ministry of Finance and Planning, they are sent to the appropriate First Secretary in BHC. ODM operates a yearly financial ceiling for SCAAP. The ceiling, like the CPP, is for internal guidance and is not communicated to the Kenyans. Two reasons are given for this secrecy. One, if the yearly ceiling were announced the Kenyans would invent requests to take up all the funds, resulting in a number of inferior projects, which it would be difficult to turn down; with the ceiling not announced, financial stringency can be used as an excuse for not supporting poor projects. Two, if good projects arise in the course of the year (as they sometimes do without warning) it is good to have money in reserve for them, and if countries are not told the full extent of their SCAAP provision, funds can be switched from one country to another. This adds flexibility and is supposed to reduce under-utilisation. We are completely out of sympathy with the first argument, if not the second. In the first place, if the Kenyans were really determined to take up all available funds they could simply

58 Aid and Inequality in Kenya: British Development Assistance to Kenya invent applications anyway until they exhausted them, thus establishing the ceiling by trial and error. If a false ceiling is established because ODM uses it as an excuse to turn down some projects, that totally negates the second argument as far as it applies to Kenya. It is not very winning to plead financial stringency one month, then announce that it has magically disappeared the next because a better project has appeared. (Although this can happen owing to forecasting errors as well as duplicity.) In addition that sort of diplomacy appears to be quite self-defeating on two counts. Many of the reasons given for project-tying aid, both capital and technical assistance, are invalid.19 One of the valid ones is that it enables the donor to demonstrate rational methods of project selection to the recipient in the hope that he will emulate them. That advantage is lost if the recipient dissembles the reasons for his preferences. Secondly, diplomacy of that sort makes for unpopularity rather than the reverse. It will become clear that we regard ODM as, developmentally, a much better donor than average in Kenya, yet it does not seem to be among the most popular donors with the Kenyans themselves. One reason is that they find the British unpredictable and do not understand why ODM rejects some projects and accepts others. They also suspect ODM of killing projects by prolonged delays while simulating interest, when a flat ‘nothing doing’ would be preferable. For all the harm that, in our opinion, more thrusting donors do, many Kenyans prefer to deal with the Americans, for example, who are blunt and straightforward in saying what they will not support and why. As the framework figures are guidelines, not commitments, it is not even clear that secrecy is necessary for flexibility between countries. Britain as the former colonial power is in a more difficult position diplomatically than other donors. Nevertheless a more open style of operation would pay dividends. There is some indication that it is happening with the establishment of the Development Division. For example, the yearly framework figures for disbursements of capital aid are also supposed to be secret but somehow the Kenyans seem to know what they are.20 It is noteworthy that they seldom succeed in spending them all, although, of course, British appraisal of capital aid projects is usually more stringent than of applications for SCAAP personnel because of the different scale of sums involved. Even though the administration of capital and technical aid is being integrated more, the funds for the two forms of aid are kept separate. The reason given for that has been that technical assistance funds are grant and capital funds are loan. This will not be so in future but in any case the inefficiency of this separation is manifest. Occasionally technical assistance funds have been spent up to the ceiling and a moratorium has had to be declared for the remainder of a financial year on this kind of aid. Mean-while capital funds were underspent. It is an obvious reform to introduce some flexibility so funds can be transferred between the two heads. Better still would be to abolish the distinction, making particular aid agreements for a certain sum to be disbursed in any form agreed between donor and recipient. That would not obviate separate technical assistance agreements being made to establish terms for technical assistance which would apply when the recipient opted to take part of an agreed aid commitment in technical assistance form. Neither would it necessarily involve changes in the administration of British aid with separate sections looking after OSAS, SCAAP and capital aid. Co-ordination already exists, although it might have to be stepped up. Despite these flaws the Kenyans have in general found British administration of SCAAP more to their liking than that for capital aid, because as long as requests for

British Aid Policy and Administration in Kenya 59 technical assistance are roughly in line with the CPP they are processed quickly. One British administrator explained that the general principle was that if the Kenyans asked for someone, they probably needed him, and it was unrealistic for ODM to argue about the merits of filling individual posts. As long as Kenya’s allocation of technical assistance funds for any year was not on the point of exhaustion, therefore, there was a tendency to provide expatriates on request as long as they could be recruited. Extensive cost-benefit analysis is not worth while in view of the small sums involved compared with capital projects. Not only is project approval by the donor quicker, there is also a pro forma for the Kenyans to complete when applying for technical assistance, so they are told precisely what information is required and are used to making requests in a standard form. This does not apply in the case of capital aid. Technical assistance administration has clearly shown the benefits of having a specialist aid agency. Some years ago, when there were nearly 2,000 British expatriates in Kenya, the programmes were administered by a single officer based in London with very junior people on the spot. Now there are more officials in London, more, and more senior, officers in BHC Nairobi, and the Development Division all involved with technical assistance. The programmes are certainly not over-administered now. The additional manpower is necessary if ODM is to form policies about the economic sectors in which technical assistance is to be concentrated and to direct the programmes acccordingly. It is sometimes argued that provision of technical assistance helps to mould recipient government policy in favour of the donor state. We discuss later the effects of various aid forms but here, in discussing British policy and administration, we can say that whatever the effect, there is no sign of such an intention on the part of ODM. There is almost no liaison at all between BHC and technical assistance personnel once they are in post. Many experts working in Kenya have been surprised at how unsupervised their work is. SCAAP personnel make an annual report on their assignment to the Kenya government and BHC. It is copied to ODM. In the case of OSAS officers the only report HMG receives of their assignment is the one they write themselves on completing it. This is on the grounds that it would be invidious for civil servants in post with one government to report to another. These reports are filed and add to the store of knowledge of experienced aid administrators, but there seems to be no procedure for collating and periodically considering them in order to inform future technical assistance policy. Even SCAAP personnel sometimes feel a sense of isolation as it is rare for their reports to lead to a request for a meeting with ODM for any re-assessment or rebriefing. A couple of SCAAP men confessed to being mildly shocked that British money was being spent on them without their feeling accountable to anyone. The most important reason is that the two or three BHC officers dealing with technical assistance at present simply do not have the time to monitor the reports of several hundred British experts, far less the terminal reports of a thousand OSAS personnel. Despite the improvement in the running of technical assistance programmes they are certainly not overadministered—rather the reverse—and proper ex post appraisal would require more staff. Procedures for enforcing a higher general standard of technical assistance reporting, and for assimilating the information would be advantageous in improving technical assistance policy even if they attracted accusations of ‘neo-colonialism’.

60 Aid and Inequality in Kenya: British Development Assistance to Kenya

Policy Issues Perhaps the first issue to be settled in determining aid policy towards any country is how much to give. The UK extends to Kenya slightly more aid per head than its average for Commonwealth Africa and three times the average for all Commonwealth countries. One reason is the importance of historical precedent. Reductions in aid, not agreed with the recipient, can be damaging to political relations, so other British ministries apart from ODM have been quite keen to see the volume of aid to Kenya maintained at least in nominal terms. The FCO, for example, is aware of the problem of Kenyan Asians who are British citizens. The Kenya government is committed to Kenyanisation of the economy. So far it has reached an agreement with the UK for British Asians to leave Kenya and come to the UK in a gradual stream. Yet the possibility remains that Kenya, like Amin’s Uganda, could simply expel all non-citizen Asians, leaving Britain with difficulties in absorbing the refugees without harming domestic race relations. This is a tacit bargaining counter for Kenya in relations between the two governments and provides the UK with an incentive to maintain friendly relations. Yet a number of facts up to 1972 persuaded some British aid administrators that on developmental grounds capital aid to Kenya could be allowed to fall in real terms. The considerations were the slow rate of disbursement, the absence of any very severe balance of payments difficulties, and the readiness of other donors to aid Kenya. Of course, in so far as this last factor indicated foreign penetration of a country and markets that had previously been a British preserve, it provided a reason for stepping up aid on commercial and political grounds. In 1972 a World Bank analysis of the Kenyan economy concluded that while up to then Kenya’s chief growth constraint had been administrative, from then on shortages of capital and foreign exchange would emerge. The economic analyses in ODM’s CPPs are usually greatly influenced by IBRD findings, and these conclusions persuaded aid administrators that aid to Kenya should not be cut, quite apart from the Asian and commercial issues. Since then, the oil price rise has heavily accentuated the deterioration in Kenya’s balance of payments, and Britain agreed to the programme grant of £2.5m. for essential imports. Another issue is whether aid should be concentrated on particular economic sectors. It is now explicit British policy to concentrate more on aiding the rural areas. Other donors are following suit.21 Most donor agencies, including ODM, have been persuaded by many of the arguments of the ILO report22 and subsequent World Bank analyses pointing to the unequal mode of Kenya’s economic growth. The concentration on the mass of rural poor is not simply because they are more numerous than the urban poor but also because relative improvements in urban conditions could cancel themselves by causing more poor people to migrate from country to town. Rural development is necessary to help both groups, therefore. Symptomatic of ODM’s emphasis on rural development is the project to assist in the construction of rural access roads, both for their own sake, and as a pilot scheme for a larger World Bank and bilateral donor consortium project to build many such roads by labour-intensive methods. The pilot project cost £250,000 and was the first authorised by BHC Nairobi under the new delegated authority. The UK’s ability to do this and get the project under way greatly impressed other donors who were unable to respond so quickly, On the other hand ODM will not now finance trunk roads between urban centres or heavy infrastructure primarily serving urban areas. In the rural areas the concentration

British Aid Policy and Administration in Kenya 61 is on agriculture or at least directly productive enterprises. ODM is less eager to assist with social services as the Kenya government is capable of providing these itself and is under some political pressure to do so. These preferences are far from being hard and fast, however. The accent on rural development indicates the predominance of ODM’s conception of development over other considerations in aid-giving. It means a departure from projects that diplomats and British businessmen can see from their Nairobi windows, and appears to go against Britain’s immediate commercial interests, as the import content of rural development projects is often low. Insistence on the policy conflicts with shortterm political interests, because rural projects are less visible, and despite the statements in Kenya’s Development Plan about the priority of rural areas, the Kenyan government still frequently requests aid—but increasingly in vain—for urban-based projects such as airport improvements and trunk roads. Money has been requested for the Industrial Development Bank to invest in large industrial projects—some of them perfectly good and viable ones— but apart from CDC’s seconding a management adviser and ODM putting IDB in touch with commercial credit sources, Britain has refused to aid this body. ODM officials have persuaded other departments in HMG to support the policy by arguing that it is in Britain’s long-term interests. Whatever can be said for aiding projects which will give business to British exporters, or accommodating some shift in Kenya government policy, Britain’s real interest, it is argued, is in helping to build a stable and prosperous Kenya. While this view is uppermost it does not always get a free run with the money. It is not all that long since there were differences of view within the British administration. Certain projects which the Kenya government has wanted strongly have been backed within HMG because of the political returns from doing so but have conflicted with ODM ideas of development priorities. An example may have been a project costing £260,000 in 1971 to extend television to the Coast Province of Kenya. It was known to be greatly favoured by President Kenyatta, and support for the project might therefore have been helpful in maintaining good relations. Aid administrators might also be expected to favour it because of the very high import content. The required import content of an aid agreement loan is an average figure, so if some projects that are almost 100 per cent imports can be found it makes it possible to accept others with a high proportion of local costs, and this facilitates disbursement and removes some of the difficulties of administering the aid programme. This means of course that procurement-tying regulations (favoured by the British Treasury) can deform decisions and modify the commitment to development. The difficulty in the television case was apparently that a great deal of sophistry failed to convince some ODM advisers that the project had much to do with development, for it was accepted only after going before the Projects Committee. Ironically, it may be this or similar incidents which have led some diplomats to describe ODM’s readiness to oppose its conception of development to that of the Kenya government as ‘neocolonial’, although they admit the legitimacy of project selection in order to ensure Value for money’. Other donors as well as ODM are concerned that the mode of economic progress achieved in Kenya since independence is not viable in the long run, and a more egalitarian course should be pursued if political tensions within Kenya are not to grow. It may seem odd that donor agencies, including ODM, should take a longer view than the Kenyan government of what is in Kenya’s interests. If there really is a threat of serious political instability, for

62 Aid and Inequality in Kenya: British Development Assistance to Kenya example from the present economic and social policies, that would appear to concern the Kenya government before anyone else. However, perhaps the situation may be compared to that in Britain, where governments have been notoriously prone to inflate the economy at the approach of an election, irrespective of the long-run wisdom of doing so. In Kenya the need to maintain a short-run balance on the political scene may preclude taking a long-run view, at least without authoritarian measures to suppress interest groups that benefit from the status quo. And it is not clear which interest groups would provide the will for such measures and the means to carry them out.23 A question which immediately arises is, what is the point of having strong sectoral preferences and attempting to act on them, if they are not really shared by the recipient government? The Kenya government has declared itself for more poverty-oriented rural programmes. And there is no doubt the commitment is taken seriously by many Kenyan civil servants. It is fair to say, however, that there is some doubt among the donors how far the Kenya government as a whole does share such preferences, and how far it merely pays lip service to them. Many claim to observe a discrepancy between policy statements and the pattern of project applications they receive. To insist on one’s own priorities in such a situation and to try to enforce them by project-tying aid may achieve little except to alienate the recipient. The usefulness of project-tying aid to secure a certain use of resources depends entirely on how fungible24 the resources in question are. We analyse fungibility and discuss its application to Kenya later,25 and here merely note ODM’s attitude towards it. The feeling in BHC and the Development Division is that one has to support good projects irrespective of political pressures or the possibility of fungibility, because doing so sets a good example to the recipient and persuades him to evaluate possibilities rationally, and ultimately it is necessary if the aid agency is to maintain morale and self-respect. The cynical would observe that this argument tends to vindicate the view that many arguments for project-tying are founded in the organisational interests of aid agencies, In the case of Kenya, however, ODM personnel believe that the recipient’s scope for shunting funds around is limited. At this point all we would say is that they appear to consider fungibility as largely a phenomenon concerning British and Kenyan funds. But where there are other donors British funds may be displacing other donors’ aid. If the other donors then aid projects of which ODM disapproves, some (not necessarily all) of the point of projecttying is lost. There is a donor practice, shared by Britain, of insisting that Kenya puts some money into any project aided by the donor. This ties up much Kenyan money, reducing the scope for fungibility. UK administrators and those of other countries, however, insist that such is not the intention; they simply want to ensure a genuine Kenyan commitment to anything they undertake.

Underspending Unquestionably, the British practice, shared by other donors, of tying aid both to projects and partially to procurement has reduced the rate of capital aid disbursement. This means that a lot of aid which Kenya is going to the trouble of negotiating for is not being received. British funds not spent in one year cannot be automatically carried over and added on to the framework figure for disbursements the following year. Money extended under an aid agreement that is not taken up during the period of the agreement is added on to the value

British Aid Policy and Administration in Kenya 63 of the new agreement. So in 1973, for example, Britain signed an agreement to extend to Kenya £17m. of ‘new’ money to be disbursed in the period 1973–77 plus £7m. to be carried over from earlier agreements (mainly the 1970 one) which had not been taken up. This division into ‘old and new’ money is, however, fairly clearly a matter of semantics; the amount of new commitment can vary depending on how much remains to be spent of a previous tranche. ODM officials insist that no crude ‘netting’ takes place, but they also admit that over the years Kenyan underspending has been a factor in determining the size of aid agreements. If underspending continues, as the aid programme is continuous, the money not being spent is pushed continually into the future and is lost in some real sense. Even if spent eventually, its real value is eroded by inflation. There are two components of underspending. One is a slowness in the Kenya Treasury in making disbursements against commitments already made. The second is a slowness in committing to specific projects funds assigned to Kenya in an aid agreement. To eliminate the first problem the Development Division would be prepared to see Kenya’s Ministry of Finance and Planning use ‘overcommitment’, that is, authorise more expenditure in any year on ‘British’ projects than Britain’s framework figure for disbursements. Then natural slippage of projects behind schedule would mean that the available aid would get taken up, probably without much, if any, exceeding of the framework figure, rather than it being underspent as at present. The Ministry does occasionally commit in advance of an aid agreement but generally it displays a very conservative, bank-like attitude to finance. Indeed, one project supported by another donor was held up because the Kenya Treasury refused to authorise expenditure on certain items, although they were fully covered by promised aid. That was because the aid had not actually been received, and there was a general proscription on expenditure from Kenyan funds on the items in question as part of an economy measure.26 Much the more serious underspending problem, however, is the one of getting enough project applications to which to commit all the funds that have been agreed. This is emphatically not a problem peculiar to British aid. All donors share it. The most important constraints are the shortage of ideas for sizeable projects in some of Kenya’s important operational ministries, and a lack of capacity to prepare project proposal and evaluation documentation to the level of detail required by many donors, including ODM. Hence the paradox, not unique to Kenya, that in a poor country lacking just about everything required to raise its people’s material standard of living there appears to be a surfeit of capital aid. The problem in Kenya will be intensified by the growing fashion among donors for favouring rural development. Productive investments in rural areas are harder to plan than simply putting up a factory or building a trunk road, and in truth when worth-while opportunities for development can be identified, they probably do not require very much money anyway, but rather improved administration and specialised skills and perhaps the release of some constraints imposed by adverse government policy, such as unfavourable agricultural pricing. Apart from these intrinsic difficulties is the fact that the Ministry of Agriculture is one part of the Kenyan administration that finds it hardest to plan projects, unlike, for example, the Ministry of Works.27 Donors’ reactions to this situation take a number of forms. One is to diagnose the problem as a shortage of human capital and to attempt to inject technical assistance into the Kenya administration in a general way. A second is to take over some functions that

64 Aid and Inequality in Kenya: British Development Assistance to Kenya strictly are those of the Kenya government; these include identification of projects and, usually in collaboration with a Kenyan operational ministry, preparing the project to implementation. This is done either by aid agency officials or by consultants or technical assistance teams specially engaged. A third is to relax requirements in some way, perhaps accepting vaguer or otherwise poorer project submissions, or easing the donor agencies’ own rules, accepting higher local-cost elements in projects, or more recurrent expenditure, or perhaps abandoning sectoral preferences to some extent. The ultimate in this line is to abandon project-tying and give balance of payments support grants or loans. A fourth is to give up and reduce capital aid allocations. Nearly all aid agencies in Kenya have done one or more of these at some time or other. ODM has shown some sympathy for giving general technical assistance. Despite its running down of OSAS it has, as noted, been permissive in its attitude towards SCAAP personnel. ODM has been shy in the past of the option of preparing projects itself, being much less interventionist in that sense than many other donors, although that policy is not changing. We suspect the option of lowering standards must have been taken from time to time although any moderating of ODM’s own rules has been only marginal. The import requirement provisions are no less steep than they were, and the aversion to recurrent costs remains. In some cases, however, non-preferred projects have been undertaken to spend money. The recent programme grant may have been exceptional, rather than an extension of a tendency to compromise with principles, but future aid agreements between the two countries may include programme aid as part of the package. As for the option of reducing capital aid, we are confident that Kenyan requests for more aid would have been more successful if disbursement rates had been higher in the past. We believe that an inability to spend capital aid is more likely to mean that technical assistance cannot be used properly either, than to mean it is bound to do good. Many donors have attempted to release the bottleneck in the Ministry of Agriculture by giving technical assistance. The UK, Canada and the World Bank have all done so at various times. The Ministry has a Project Planning Unit, which is largely staffed by expatriates, and it is beginning to jib at taking more. Some of the older advisers with a good knowledge of the country have done excellent work, but in general expatriates sitting in Nairobi, even if skilled in writing up project applications, are unlikely to get fertile ideas for projects or to be able to evaluate information they receive without the local knowledge and experience that many of them lack. The fate of the Unit provides an object lesson in the limitations of technical assistance when there is a weakness in the administration of a recipient country. The expatriates work and plan in vacuo, a tendency accentuated because there are clusters of them. The lack of political drive, which meant that indigenous Kenyan administrative resources and recurrent expenditure were not directed into agriculture, nullifies the efforts of expatriates even when those turn out to be more than jejune and inexperienced. They are reduced to justifying their position as fulfilling an educative function—demonstrating certain management and planning techniques—because they know that at present little is coming of their work. Even that benefit is reduced if a unit consists largely of expatriates, as then much of their educating is of each other. ODM does not have such a penchant as some other donors for putting in teams of planners,28 although it has more than one expatriate planner in Education and has provided a team for the Tana River Development Authority. Both of these ventures may well prove

British Aid Policy and Administration in Kenya 65 fruitful, but in general experience does not support the view that Kenya’s absorptive capacity can be usefully improved by inserting cliques of expatriates. In this context two may be company but three is certainly a crowd.29 There are perhaps two exceptions. One is in the field of research, which by its nature is an activity which can be more isolated from other government functions, and which often requires close co-operation of similarly oriented people. ODM has supported successful research teams in agriculture, especially in maize-breeding and agronomy. The second is the provision of short-term teams to tackle technical problems relating to the preparation of a single project. Consultancies may be the best means of doing this. ODM appears to have a reputation for being more difficult than most other donors in the amount of information and documentation it requires before supporting any project. In ODM’s view Kenyan operational ministries are supposed to evaluate capital projects on the lines of ODM’s Guide to Project Appraisal30 and to send applications to the Treasury to forward to appropriate donors. The ministries do not do that much, and when they do, Britain is not a favourite donor to approach first. A common informed view is that ODM asks for information that it is impossible to supply in advance of a project being implemented; the assessment of a construction project, for example, may call for building plans. ODM asks for such details simply in considering whether to support a project. However, the Kenyan government does not have the manpower resources to justify preparing a queue of thorough project applications. It is only worth working up a project to the implementation stage if the Kenyan government is sure it is going ahead. British practice, it is alleged, necessitates the Kenyans sinking precious administrative and planning resources and bearing the risk that the British will then turn the project down or insist on unacceptable modifications of the plans. Other donors are thought to be easier to deal with in that they accept a project in principle and then, if they want further details and plans, offer a technical assistance team to work with the Kenyan ministry concerned in preparing these. While this view is very widely held, it does seem that the Kenyans misunderstand the procedures of some other donors. The Germans, for example, ‘earmark’ projects for support at the time of general aid negotiations. The Kenyans regard this as a commitment but in fact the Germans do not and could, in principle, refuse to support a project which emerged in a form they disapproved. In fact such a politically embarrassing situation has never arisen, evidence perhaps of the influence foreign experts can have over the form of any particular project. The British approach, however, does have its positive side. ODM claims that its policy is to strengthen the Kenyan Treasury in its dealings with operational ministries. So ODM in principle expresses interest only in projects proposed by the Kenyans, either directly in negotiations, or more commonly by publication in Part Two of the Development Plan,31 or in a Public Sector Projects List circulated to donors. Traditionally, expression of interest is made to the Kenya Treasury, usually to the External Aid Division which is asked to obtain an application from the appropriate ministry. Delays frequently ensue, punctuated by British appeals to the External Aid Division and by Treasury appeals to the operational ministry for expedition. Applications are produced, and in the past they have been referred back by HMG or ODM as inadequate. Eventually an application is produced that is regarded as adequate, It is examined by the staff of the Development Division and, for sums over £400,000, sent to London with their recommendations. In the early months

66 Aid and Inequality in Kenya: British Development Assistance to Kenya of the Development Division further delays occurred because ODM experts in London on the Economic Planning Staff queried facts and recommendations. Gradually the worst duplication of this kind is falling away as the Development Division becomes established. If the project is a large one, however, it has to wait for the next meeting of the Projects Committee. Even then it may conceivably be referred back. Eventually the project may go through but not always before mutual irritation has been engendered between British and Kenyans. This contrasts markedly with the procedure of some donors, not, as Kenyan Treasury officials may be excused for believing, because their requirements for project write-ups are less onerous than those of ODM—we do not think they are much—but because their procedures are less formal. Other donors may see a project they fancy in the Kenyan literature or they may dream up their own. An approach is then made to contacts in the operational ministry concerned. Promises are made of money for the project and support is won. The donor may well practically write the application he wants to receive on behalf of the operational ministry and get the latter to send it to the Treasury with the information that the donor is prepared to finance the project. (No wonder the Treasury is not plagued by requests for more information.) The application is sent off to the donor country for approval. When the Kenya Treasury gets around to forwarding the application to the donor agency in Kenya, it is accepted and the project can be put in hand. Clearly this procedure leads to more rapid project preparation and implementation than the British method, but equally clearly it has dangers of its own. It means that only a minimal attempt is made to devise the projects the Kenyans want, rather than those the donor wants to support. The Kenya Treasury finds itself receiving applications for projects that may or may not have the same names as those in the Development Plan, but if there is donor money behind them, it is hard to turn them down. The Planning Department in the Ministry of Finance and Planning has no opportunity to establish any kind of priority ordering of projects, and to direct donors’ attentions to the top-ranking ones; it is reduced to saying yes or no to what donors propose. British administrators claim that some other donors have bitten off more than they care to chew by using this technique, and that some projects forced through the Kenyan machine have been flops on any criteria. Pressure can also increase friction and one administrator summed up: ‘Aid must be partly about goodwill, so what is the point of a procedure that alienates the Kenyans straight away?’ Unfortunately, that argument does not accord with the views of Kenyan civil servants. Treasury men in particular are only too well aware of the limitations of some operational ministries. Being constantly chivvied by the British and having to chivvy other ministries in turn is not a comfortable position, and they would welcome the British doing some of their own pressing. Furthermore they are concerned with disbursement and they know that dealing directly with operational ministries speeds a donor’s programme. They recognise the argument for not undermining Treasury control but, strangely, give it much less weight than BHC and ODM. In fact Treasury civil servants do not seem excessively concerned to establish priorities. As long as projects are in the Development Plan, they are content to handle them on a first-come, first-served basis. Evidence of this is a practice, enshrined in a number of intra-ministry circulars, that Treasury approval shall not be given for any project—even if in the Development Plan—unless some aid finance is assured. This amounts, in effect, to an admission that the Ministry of Finance and Planning is content

British Aid Policy and Administration in Kenya 67 to let donors do some pre-selecting for it. It would thus like to see the British take a more active role in aid-giving and make their own contacts with the operational ministries. There is evidence now that, since the setting-up of the Development Division, the British have realised this, and are doing so. The staff of the Development Division have discussed applications with an originating ministry before they were submitted, to ensure that the application arrived in a suitable form, and have actually helped in drafting applications. What is comfortable for Treasury officials, of course, is not necessarily best for the country in the longer run. ODM is therefore in a considerable dilemma. The desire to disburse aid faster and to promote better projects has to be balanced against the further weakening of Treasury control and the implicit infringement of Kenya’s sovereignty.32 The trouble is that a self-denying ordinance by ODM does little good if other donors are piling in.

Donor co-operation Projects in which a number of national donor agencies co-operate are still a tiny minority in Kenya. Multi-donor consortia, when they occur, are usually under the auspices of the World Bank. An example is a livestock promotion scheme where the UK is aiding the marketing component. The exception is the co-operation of Sweden, Denmark and Norway in ‘Nordic’ schemes, The UK is not antipathetic to co-operating with other donors and has done so occasionally for years, although it is fair to say that consortia are not its preferred mode of operation. ODM keeps in touch with the World Bank, and in one case turned down a project for a teacher training college after discovering an IBRD appraisal had gone against it. In general, however, the UK follows rather than initiates moves towards donor co-operation, leaving those to be started by the World Bank and UNDP which organises a monthly get-together lunch for donor agencies in Nairobi. ODM’s considered view is that a bilateral agency, especially that of the former colonial power, is ill-placed to set about engineering joint action by donors, especially as the Kenya government is thought to be suspicious of any signs that donors are ‘ganging-up’ on it. There is a Consultative Group for East Africa which holds periodic meetings of donors, chaired by IBRD, in respect of individual recipient countries. It is attended by bureaucrats of British Under-Secretary level and below from donor agencies’ headquarters. Some discreet carving up of donor areas of interest goes on but disputes as such are not resolved. The Kenyans who attend emphasise those aspects of their policy which please the donors, and diplomacy reigns supreme. The occasions appear to be of no practical importance in shaping anyone’s policy, although they are said to improve the atmosphere in which donors operate. However, senior staff of the Kenyan Ministry of Finance and Planning are said to have devoted most of two months to preparing documentation for a Consultative Group Meeting in 1972. The present state of donor co-operation appears to be an advance over the former situation. Some years ago the UK seemed excessively suspicious when the USA, for example, noticed that the Special Rural Development Projects33 were likely to founder on the indifference of important ministries in Nairobi. UK representatives refused to take part in joint donor representations to the Ministry of Finance and Planning. The US had made its own mistakes in the field of rural development and these were notorious but it should have been clear they were irrelevant to the point at issue. Although the US was quite right

68 Aid and Inequality in Kenya: British Development Assistance to Kenya in its forebodings, some UK administrators wanted to assure the Kenya government of their continued confidence. This may have been an instance of the pervasiveness of ODM and FCO ‘diplomacy’ as much as suspicion of other donors.

Other Donors’ Policies So far we have considered British aid policy, referring in passing to other donors. In Kenya there are many such, a fact that greatly affects British policy, and arguably should do so still more. The popularity of Kenya among other donors is probably best explained on organisational rather than political grounds.34 The decision to ‘have an aid programme’, like the decision to operate a national airline, is taken in many countries as evidence that they have graduated to a higher stage of political and economic maturity. An early decision may be taken to spend some of the money in Africa. (Why not?) The country that gets the money will then have certain features. It will not have a desperately objectionable government to Western capitalist countries; it will not be too poor or small (it is impossible to spend much money in countries like Chad or Mauritania without practically taking them over; things are hard enough in Kenya), although publicity given to the plight of the 25 least developed’ countries, which are generally small, may change that slightly. Unless the prospective donor is France or Belgium it will prefer an Anglophone country as more of its technical personnel are likely to speak English; the scale of French aid, however, provides a more reasonable rationale for this tendency. It will be a useful bonus if the recipient provides an acceptable locality for European administrators and technical advisers in terms of climate and social life. It is evident that Kenya meets these requirements—perhaps best of all African countries, In addition, Kenyan politicians and civil servants have acquired considerable skill in dealing with donors. Their skill in this direction outstrips that of the rest of the government machine in finding productive uses for the aid on offer. In this section we give brief descriptions of the policies of a few of the more important donors.

United States The United States Agency for International Development (AID) is an agency of the State Department of the world’s largest economic and military power, and as such is conscious that its actions have implications other than the purely humanitarian. AID personnel, however, insist, with plausibility, that military, political and commercial considerations are much less important in US aid to Kenya than they are elsewhere like Latin America, where US interests are more directly involved.35 Although there is considerable US investment in Kenya it is small compared to that of the UK, and the area has no particular strategic significance. In Kenya humanitarian motives for aid are the ones stressed by AID personnel. In its emphasis on agriculture the US has anticipated the common fashion among other donors and more recently it has rationalised this emphasis and declared it as a policy for US aid in general of developing ‘poverty focused aid’.36 The AID mission in Kenya is currently planning and negotiating a sector loan of $10–15m. for agriculture. If the US has been a leader among donors in terms of sectoral emphasis, however, its methods and procedures have been among the most backward. Each technical assistance project is the subject of separate agreement that has to be signed afresh each year.

British Aid Policy and Administration in Kenya 69 Capital loans are also subject to separate agreements that lay down strict guidelines for disbursement. Funds are withdrawn if an agreement is not signed within six months of funds being earmarkéd, if prior conditions are not met, or if a project is not completed in three years. The chief difficulty is that AID gets a country-by-country annual vote in Congress and does not know beforehand how much aid will be available for each country. Activity is assumed to go on at the previous year’s level until the Bill is passed and new commitments can be made.37 The need for each loan or grant to be negotiated independently meant in the past that the US was rarely approached for funds by the Kenyans, who preferred to submit projects to other donors. In recent years the US has been getting around this by being the most initiatory of aid donors. Sometimes the AID Mission in Nairobi writes project submissions of behalf of Kenyan operational ministries and sends them almost simultaneously to the Kenya Treasury and AID in Washington. The US does not restrict itself to projects in the Kenya Development Plan. Project ideas ‘consistent with Kenya and US policy’ can emerge in informal dialogue and then they may be taken up by AID who put technical assistance in to assist with project preparation. For example, the US is ‘concerned to help the Kenyans with their equity problem and help the marginal farmer’.38 The Kenyans approached AID for a loan for small-farm credit, but the Americans did not consider credit to be the binding constraint on rural development. They got Kenyan approval for a technical assistance project to work up a general agricultural programme dealing with many factors including credit, and that is to be the basis of their sectoral loan. As an AID man put it: ‘Once we decide to get involved, we push,’

West Germany West Germany has as much difficulty in spending its aid as ODM, At the end of 1974 not a Mark had been spent on some projects agreed two and a half years earlier. This is partly owing to Kenyan administrative delays and partly because there is little local delegation by the Germans, so project applications have to go to Germany or teams from there have to visit Kenya for the assessment of projects. The Germans are now interested in switching more to programme or sectoral aid partly to help disbursement. German aid is fixed in two-yearly negotiations with the Kenyans at which individual projects as well as the total sum to be given are discussed and funds are earmarked for projects up to the level of the total commitment. Agreements are made for the two-year period without even ‘planning’ figures being given for longer periods. One advantage of the German approach, however, is that the programming is done entirely in terms of commitment without stipulation as to rates of disbursement. The Germans have a very high tolerance of deviations in disbursements for any given year. They claim that experience is enough to guide them in estimating the rate at which commitments will be disbursed, and there is no need to be more rigid. The Kenyans do not have to worry about how much or how little they are spending in any year; they can just draw on a commitment until it is exhausted. Capital aid and technical assistance are discussed at the same negotiations although there is a separate commitment framework for each. Unlike the British the Germans tell the Kenyans what the TA commitment is. They are clear-sighted about the division of aid into

70 Aid and Inequality in Kenya: British Development Assistance to Kenya two exclusive funds. German officials admit it has no rationale and explain it as historical accident: capital aid was the responsibility of the Treasury while technical assistance came under the Ministry for Economic Co-operation. Vested interests are involved and it is not likely that the two frameworks will be amalgamated in the near future. At present capital aid is loan and technical assistance is grant.39 It has been suggested40 that in the early years of the aid programme the Germans were intoxicated with the post-war success of their own market economy and anxious to proselytise others. There is little indication of it currently in Kenya, although West Germany is not among those donors that prefer aiding Tanzania to Kenya. The Germans disavow the approach of giving aid primarily to regimes with whose ideology the donor agrees—‘as long as the recipient government is a reasonable medium for assisting poor people it’s all right’. A move to programme aid would, however, involve some political interventionism in the style of the IBRD. This is acknowledged within the German aid agency.

Sweden In many ways the Swedish administration of aid is more liberal than that of any other donor. Like the UK the Swedes agree a block sum of aid for disbursement over a period (of three years in their case) before discussing particular projects. A firm figure is announced for disbursement in the imminent financial year with planning figures for the two following years. The Kenyans can apply for aid for projects at any time against the sums agreed. Unlike the UK Sweden has full carry-over of funds to following financial years if they are not disbursed on time. Uniquely among donors in Kenya the Swedes also disburse their aid in advance of expenditure by the Kenyans. Detailed estimates are requested for spending over the next six months, and disbursement is made against the estimate. At the end of six months the Swedes request reporting on the expenditure. They are primarily concerned that the balances they have advanced have been committed by the Kenyans and are not so concerned about whether the money has actually been paid out. Then they make further advances against spending estimates for the next six months. There is an eventual check on expenditure in the Kenya government’s own audited reports. This is necessary to compute the following year’s budgetary allocation. The Swedes are not neurotic about financing recurrent expenditure. A number of their projects, an artificial insemination programme, a foot-and-mouth control programme, and the Kenya Science Teachers Training College, included provision for recurrent costs to be met, then to be gradually taken over by the Kenya government. The combination of three-year aid agreements, carry-over of undisbursed funds, advance disbursement, no procurement-tying, recurrent cost financing and soft terms makes the Swedes popular donors but has not prevented underspending. Undisbursed aid balances have built up, and when advance disbursement began in 1973 the Kenyans were asking for the next tranche while being unable to show they had fully committed the previous one. However the system is said to be working more smoothly now. The Swedes are explicit about their conception of development and their greater willingness to aid governments they believe to be pursuing it. The Swedish aid agency, SIDA, prepares an equivalent of the ODM Country Policy Paper stating a view of recipient government development policy, and it is shown to the recipient government. SIDA’s policy

British Aid Policy and Administration in Kenya 71 is to promote economic equality and economic independence. In recent years in Kenya this has led SIDA to favour rural development projects. It has agreed to contribute to the Kenya government’s Rural Development Fund.41 However, many Kenyan civil servants believe Sweden favours Tanzania on ideological grounds, preferring that government’s approach to economic development.

The World Bank In his Nairobi speech in September 1973, Bank President Robert McNamara repeated the view that the tendency to concentrate resources on a ‘modern sector’ using capitalintensive technology had encouraged dual economies with few benefits percolating down to the rural poor.42 A new strategy was required, spreading benefits among rural masses even at the expense of some economic growth as conventionally measured. The Bank has encountered problems in beginning to implement such a strategy in Kenya.43 Many of its projects are still for heavy infrastructure, the reason being that the Bank has been doing what it can. Proposals for productive rural sector investment have not been forthcoming, The Bank regards its interest in livestock development in co-operation with other donors as typical of the way it wants to go and is taking a less accommodating view of requests to give large sums of money for trunk roads. As an international agency the Bank is less inhibited than most bilateral agencies in making recommendations on general policy to the Kenya government. It has conducted a series of reviews of the Kenya economy complete with recommendations on the pattern of public sector investment and fiscal and other issues and incorporating IMF views on exchange rate policy. Its views are very similar to those outlined in Part One of the current Kenya Development Plan but they are not really reflected in the pattern of projects for which the Kenya government requests foreign assistance or the pattern of projected recurrent expenditure. In 1974 the Bank was negotiating to give a programme loan to Kenya, and a prelude to doing so was a close examination of Kenya’s macro-economic policies. As well as extending programme loans and intervening in economic policy, the Bank is among the most punctilious of donors when assessing single projects. There are seven different procedures for making loan disbursements, one of which has to be followed to the letter in any particular case. Disbursements are usually in respect of expenditures already incurred by the Kenyans, although sometimes the Bank can supply letters of credit direct to suppliers in other countries on provision of shipping documents. High standards of accountability are enforced for Bank projects, including three-monthly reports on their progress. Bank loans are, however, not procurement-tied to any particular country, although in the past they have been for ‘offshore’ expenses. Despite this the Bank has been quite a popular donor with the Kenyans. There seem to be two reasons. Most important is the fact that it is prepared to advance larger sums for a single project than many bilateral donors and if necessary, in the case of very large projects, can organise consortia better than bilateral donors. Secondly, whatever the difficulties imposed by rigid administrative procedures, they do leave recipients knowing where they stand. The Bank is very straightforward and is always ready to put in consultants to help with difficulties in project preparation or implementation. IDA loans are also softer than those of many bilateral donors.

72 Aid and Inequality in Kenya: British Development Assistance to Kenya

The Commonwealth Development Corporation The rationale for having a body like CDC as well as an aid programme proper is not perhaps obvious at first sight. It appears to involve two administrations instead of one. And while the CDC uses what is in effect aid money, the requirement that it should pay its way44 would appear to conflict, at least potentially, with the aim of development. For one thing it entails somewhat harder terms which have to be borne by the recipient. For another, given the price distortions in Idcs, projects which are socially beneficial in a macro-economic sense and those that make a profit are not necessarily the same thing. It might also appear strange, if the British government takes seriously the policy priorities devised in ODM, that these are not reflected by CDC. The Corporation does meet the Minister of Overseas Development and takes account of his views as to which countries are to be regarded as high priority for lending. It can go into new countries only with his approval. Yet on terms and project selection it goes its own way. The arrangement does have advantages. The fact that it is not a government department gives CDC a number of important freedoms. It can insist upon preconditions before supporting a project, with a directness that government agencies do not find easy. Its requirement to pay its way gives it, in effect, a locus standi for being forthright with recipient governments. In this its demeanour, where necessary, can resemble that of the World Bank rather than that of a bilateral donor agency. This also applies to the nature of technical assistance supplied. As an investor, rather than simply an aid-giver, CDC monitors projects after they have been launched, as noted in the case of tea projects.45 While perhaps irritating on occasion to the Kenyans, this after-care can be a valuable form of technical assistance. Technical assistance supplied by ODM to operate capital projects works for the Kenya government, and if its advice is ignored has little recourse but to write a gloomy report to ODM at the end of the year. The CDC technical assistance is often from people still working for that organisation and their recommendations, if ignored, can be taken up at a higher level. There is very little of this ex post evaluation and discipline in the case of projects financed by ODM. CDC funds are not attended by all the rules that have circumscribed ODM. Money, for example, is not partly grant and partly loan, partly ‘capital’ aid and partly technical assistance. If CDC wishes to appoint consultants it can do so without concern as to which heading the expenditure should come under. In addition its funds are not sourcetied. Commonly CDC will be financing a venture in conjunction with a sponsoring commercial company. That company will generally obtain supplies from the best source. CDC often tries via its Supply Department to find a competitive British supplier, but does not insist upon it. Most of the tea factory equipment in Kenya is British, but so are the commercial companies concerned. CDC is also involved in textiles in Kenya with 10 per cent of the equity (£44,000) and £80,000 in loans committed to Kisumu Cotton Mills Ltd. The Development Finance Co. of Kenya holds another £10,000 equity. Private interests in the company are Indian and most of the equipment is Indian, although there is also some Dutch and German machinery. The Corporation’s closer contacts with the private sector mean that there seem to be more project ideas at a concrete stage competing for its funds than is the case with ODM, which has to rely solely on the Kenya government bureaucracy. In 1974, for example, CDC

British Aid Policy and Administration in Kenya 73 was fully committed world-wide and was turning down projects in Kenya. The majority of approaches for funds come from the private sector but the trend, in Kenya at least, is increasingly for the government or para-statals to make the approach. CDC is aware that many areas of the economy are so sensitive that it is preferable for the government to be involved from the start. Anything involving land is an obvious example. State involvement also means that the government is less likely subsequently to introduce policies that are detrimental to the interests of CDC ventures or to argue about the interpretation of agreements made with such ventures. Any tension in government policies or objectives can perhaps be resolved at an early stage. CDC activities are thus being drawn more into the public sector. On the other hand in 1971 the Conservative British government attempted to integrate official aid more closely with British private investment. In a White Paper46 a number of measures were announced to encourage British private investment in Idcs. The most important were: Encouragement for joint ventures involving British private capital. This was to be done after a request from the Idc government. HMG could then supply aid funds, through it, to local investment banks and development corporations for use in British joint ventures. ODM could also help British firms to make the necessary contacts. Encouragement for British firms to investigate investment opportunities in Idcs. British aid funds were to be made available for part of the cost of pre-investment studies by private firms considering either a financial investment or participation in management of Idc enterprises. Provision of government-to-government aid for infrastructure—road or rail communications, electricity, water, sewerage or drainage etc.—where facilities would otherwise be inadequate for a British private investment or joint venture.

These provisions are still in existence, but with the change of British government in 1974 little or no emphasis is given to them in Kenya or elsewhere. The one outstanding case of HMG providing aid funds for the Kenyans to invest in a joint venture with British private enterprise was the Mumias Sugar project and that pre-dated the provisions.47

Notes 1. The figure was somewhat higher than usual as some money was given for flood relief. Kenya 1962, London, HMSO, Ch. 3, p. 25. 2. See Chapter Five, Part One. 3. E.g., by Colin Leys in Underdevelopment in Kenya: the Political Economy of Neo-Colonialism 1964–1971, London, Heinemann, 1975. 4. See Chapter Five, pp. 106–7. 5. Private long-term capital movements: 1961, £2.2m.; 1962, −£0.2m.; 1963, −£9.4m.; 1964, −£15.0m.; 1965, −£1.5m. Kenya Statistical Abstract 1967. 6. Duncan Sandys made this clear in the House of Commons on 22 November 1963 (Hansard, Col 1396). 7. However, an ODM official makes an interesting point: in the early years of independence there were as many advisers in Tanganyika but they did not constrain policy there. The influence of advisers is proportional to the congeniality of their advice. The greater complexity of the Kenyan economy and administration, however, may have made expatriates less dispensable.

74 Aid and Inequality in Kenya: British Development Assistance to Kenya 8. See J.R.Nellis, ‘Expatriates in the Government of Kenya’ in Journal of Commonwealth Political Studies, Vol. XI, No. 3, November 1973, who records that this seminal expression of policy was framed by a foreign adviser. 9. However, some CD & W balances are still unspent, and ODM has stated that their use is ‘limited to British goods unless strong reasons can be given for relaxation of this rule’. 10. Clive S.Gray, ‘Impact of Foreign Aid for Local Costs on the Donor’s International Receipts’, Eastern Africa Economic Review, Vol. 3, No. 1, June 1971. 11. The Kenyans were not committing the Mercantilist fallacy in pointing this out. The British Treasury’s insistence that aid should be at least partly tied was rational only if Sterling were overvalued. If it is so, however, exports are overpriced and it is good to promote them. Of course, the argument had more force when the Kenya shilling was fully convertible. 12. Their currencies, of course, were not overvalued. 13. The backlog in spending capital aid has always been a problem. In 1970 only two-thirds of the £18m. committed in 1966 had been spent and £0.5m. was still unallocated to any project. The trend in ODM policy precluded any improvement. See p. 88. 14. See R.B.M.King, ‘The Planning of the British Aid Programme’, Journal of Administration Overseas, January 1972. 15. King, op. cit. 16. Some confusion is possible because within ODM the term technical assistance is sometimes used to refer only to wholly-financed advisers as distinct from those partly-financed people in established posts with the Kenya government covered by OSAS. We use technical assistance in the ‘outsider’s’ sense as a general term covering both groups. 17. The difficulty in doing it more quickly, even if desirable, is the scale of the programme. In 1974/75 OSAS expenditure in Kenya, including EAC, was £4m. SCAAP expenditure was £½m. and capital disbursements £3½m. 18. See Chapter Six, pp. 206–7. 19. See Chapter Six, pp. 184–9. 20. Other agencies, notably the Swedes, tell the Kenyans what their disbursement framework figures are for technical assistance and capital aid and indeed communicate the contents of their equivalent of the CPP. As one Swedish administrator said: ‘If you are only going to give so much aid in a certain way because, for example, you dislike certain policies of a recipient government it seems only fair to tell it so and give it a right of reply.’ The UK does not do so. This seems to be another example of the obsessive and malfunctional secrecy of British government departments. 21. See pp. 94–9. Rural development has always been a strand in HMG thinking since the Swynnerton report. 22. ILO, Employment, Incomes and Equality: A Strategy for Increasing Productive Employment in Kenya, Geneva, 1972. 23. See Chapter Six, pp. 217–22. 24. Fungible is defined in the Hamlyn Encyclopedia World Dictionary as: ‘of such a nature that one instance or portion may be replaced by another in respect of function, office or use: usually confined to goods…as money or grain’. Hence, as one bit of money is as good as another, the effect of aid may be to displace domestic 01 other funds from a project to be used for another project altogether. 25. See Chapter Six. 26. Information from an official of the Ministry. The Swedes are the only donors who disburse aid in advance of expenditure by the Kenyans. One possible advantage of doing this is its psychologically liberating effect on the Kenya Treasury. 27. As one Treasury man ruefully put it: ‘Works…you tell them how much money they’ve got and they’ll tell you where the road stops’.

British Aid Policy and Administration in Kenya 75 28. It did try to infiltrate the Ministry of Lands and Settlement which was spending much British money, but without success. See Chapter Five, pp. 119–20. 29. For further discussion see Chapter Six, pp. 209–14. 30. This proposes cost-benefit analysis on the general principles of I.M.D.Little and J.A.Mirrlees, Manual of Industrial Project Analysis for Developing Countries, Vol. II, Social Cost-Benefit Analysis, Paris, OECD Development Centre, 1968, but with many simplifications in procedures. Use of the guidelines is not obligatory for all sorts of projects. 31. Kenya’s four-year development plans are in two parts. Part One gives general policies and overall expenditure projections of different ministries. Part Two is a list of projects grouped by ministry which are supposed to make up the development expenditure. 32. See Chapter Six. 33. See Chapter Five, Part Three. 34. Another popular recipient is Tanzania, whose ideological attitude is rather different. Of course it could be that aid has a supporting role here, a subverting role there; it could be that such considerations are simply over-rated. 35. This is a source of relief to them; as one said, ‘It’s nice not to be the big kid on the block for a change’. 36. AID Program Budget Submission Guidance to Missions FY 1976. 37. See G.Cunningham, The Management of Aid Agencies, for a general account of US and other country aid procedures. London, Croom Helm, 1974. 38. As told to one of the authors in an interview with USAID personnel. 39. See Chapter Three, p. 67. 40. Paul Streeten, ‘Aid to Africa’, Geneva, UN Economic and Social Council, mimeo., 1970, p. 5, para. 10. The change in German aid policy has been credited to the period when the Social Democrat Dr Eppler was Minister of Economic Co-operation. See Cunningham, op. cit. 41. A form of programme assistance for decentralised rural development. See Chapter Five, p. 181. 42. For a summary and discussion of this speech which itself summarised earlier ones on similar lines, see A.Bottrall, ‘The McNamara Strategy: Putting Precept into Practice’, ODI Review 1–1974. 43. See Chapter Three, p. 68. 44. See Chapter Three, p. 64. 45. See Chapter Three, p. 65. 46. British Private Investment in Developing Countries, Cmnd 4656, April 1971. 47. ‘Joint venture’ may give a somewhat wrong impression as the British company involved, Booker McConnell Technical Services, has a management contract to run the project and only a very small equity stake of 5 per cent. See Chapter Five, Part Two.

5 AID IN ACTION

It is a familiar fact that the influence of aid on a country, even when it is completely project-tied, cannot be assessed by observing individual aided projects and cumulating their effects. That is because of the fungibility of resources. Aid, although it might be provided for specific uses, is an addition to total resources whose effect could be to free existing resources from the uses to which aid is then put. The freed resources are then available for other uses. If the aid had not been provided therefore, it is perfectly possible that the projects it ostensibly financed would have been undertaken anyway and it is other expenditures that would not have occurred. Of course, development projects are not born fully formed. It may be that a project would have been carried out in some form, even without the aid that ostensibly financed it. Yet because this aid was offered, a donor may have come to exercise influence over the specification of the project. It is, therefore, possible to divide the effects of aid into two: the micro-level effects which consist of the donor’s influence over the details of particular expenditures, whether these expenditures depend on him or not, and the macro-level effects which consist of the influence of his aid on the allocation of resources in a more general sense. We postpone discussion of this distinction to Chapter Six and in this chapter we give a detailed account of three important British aid programmes or projects. In the case of two of the studies, the Mumias Sugar Project and the Kwale Special Rural Development Project, the focus is largely on micro-level effects. In the case of the first study, the Land Transfer Programme, the expenditures were on such a scale, the issue was so politically important, and the policy options so diverse, that decisions about the Programme more clearly had considerable macro-level implications.

Part One: The Land Transfer Programme Almost 40 per cent of all Britain’s aid commitments to Kenya since independence have been for the Land Transfer Programme, under which British farmers in the Kenya Highlands were bought out and the land was transferred to African ownership, The Programme was an example of a relatively large land reform, yet the finance for it, some £33m., represents perhaps the most controversial element of British aid. British diplomats and aid administrators commonly regard it as the most farreaching aid Britain has given to Kenya, claiming that it ‘saved political stability and democracy’ and ‘preserved law and order’. Yet many Kenyans dispute whether much of this money can be called aid at all. Parliamentary and press statements have referred to it simply as compensation for British farmers, and generous compensation at that. Sessional Paper No. 10 of 1965, On African Socialism and its Application to Planning in Kenya, hinted at a similar interpretation,

Aid in Action

77

and pointed out that as much of the money was loan, it was Kenya that was doing the compensating. Paragraph 103 stated: The settlement process was designed more to aid those Europeans who wanted to leave than the Africans who received the land. Our land problems should not be settled on terms decided in the United Kingdom… It is unlikely that Kenya, in accepting the debt burden, has obtained economic benefits of anywhere near the amount of debt incurred.

This section tests the latter assertion and considers what British policy interests were served by the Programme, An account of the Land Transfer Programme is followed by a suggested framework for evaluating it and some judgements. The account falls conveniently into two parts: 1960–65 and 1966–74.

1960–65 The Land Transfer Programme was instituted by the colonial administration.1 In 1960 the State of Emergency which had existed for seven years was lifted, and at a constitutional conference at Lancaster House, London, in February of that year agreement was reached on a new constitution with common roll elections and progress to majority rule. Later in the year an Order in Council abolished the Scheduled Areas. These events stimulated considerable unease among European farmers who began to demand that the British government, which so recently had been encouraging them to settle, should guarantee their land rights under the African government which was now seen to be inevitable. While the British government had abandoned any notion of Kenya as a white man’s country, it hoped to preserve a multiracial state where whites still had a large role to play in the spheres of politics and agriculture, among others. HMG therefore concurred with white liberals in the Kenya government in setting up the first Land Transfer Programme, the ‘Yeoman and Peasant’ scheme. The plan was to settle 6,000 African families on Peasant holdings designed to produce £100 a year net income above subsistence, and a further 1,800 families on Yeoman schemes designed to provide £250 a year net income above subsistence. This scheme, introduced in January 1961, was projected to cost £7.5m. Finance was to be provided by HMG, CDC and the World Bank. The Land Development and Settlement Board was set up to administer the scheme. The intention was to ease tension over the land issue, and by taking over underutilised land on a willing buyer, willing seller basis to maintain or raise output.2 There were, of course, settler communities in other British colonies, notably the Rhodesias, so HMG, with its eye on the wider colonial implications, wished to defuse the fears of whites while avoiding any suggestion that it was prepared to compensate settlers or guarantee their property. Compensation would establish an expensive precedent and might encourage expropriation by new governments, To this day British civil servants are at pains to stress that the Land Transfer Programme is not compensation, The forms of the transactions are designed to make that credible. Payments are made by HMG to the Kenya Treasury, after the purchase has been effected and the documentation complete. British farmers deal entirely with officers of the Kenya government in the Ministry of Lands and Settlement.

78 Aid and Inequality in Kenya: British Development Assistance to Kenya A general election in April 1961 led to the formation of a government dominated by the ‘moderate’ African party KADU, but including many white liberals. KADU made it a condition of taking office that it should receive more financial aid to appease African land hunger and so prevent the party being outflanked by more militant African nationalists in KANU.3 To meet the requirements the ‘New Scheme’ was devised, doubling the acreage to be transferred with the intention of settling 12,000 families on higher density plots with net annual cash income targets of from £25 to £40. The programme was principally designed to meet the problems of unemployed Kikuyu families but it also showed the effects of pressure by European farmers. Allowance was made for one payment in cash to vendors, instead of a one-third cash payment and the balance in three equal interest-bearing instalments over five years, as planned for the original settlement scheme.4 This scheme, too, proved inadequate. British farmers continued to write in to the government offering their farms for sale while African squatters were actually staking out areas to occupy on some European farms, By the time that the second Lancaster House Conference was held in February and March 1962, investment by Europeans and Asians had virtually stopped and it was feared that the economy… ‘still greatly dependent on European farming, might collapse’.5 The market in farm land had already done so. As a result of the Conference, KANU members came into a coalition government with KADU, the Majimbo or regional constitution was promulgated, and the ‘Million-Acre’ scheme was born. This encompassed the ‘New Scheme’ and allowed for the purchase of 200,000 acres of land a year for five years at a cost of over £15m. The land, which included highly developed agricultural areas, was to be subdivided to provide net cash incomes in the range £25-£70 a year. Land was purchased in large blocks and settled according to tribal spheres of influence.6 A few farms were also offered to ‘freedom fighters’ still in the forests to induce them to come out and accept the authority of the African government.7 KANU accepted the scheme, although there was some feeling in the party against it as appearing to accept the legitimacy of European landholding. Some leaders believed the land should be acquired free after independence.8 There was also uneasiness because the scheme was being implemented while KANU was not running the government. The party accepted the scheme on balance because its main ethnic support was receiving the bulk of the land, and because it did not wish to delay independence by disputes over the land issue.9 While the settlement was regarded as imperfect it offered some resolution of the land problem which was explosive enough to threaten any government, even an independent African one. Another reason for acceptance which has transpired more recently is the conservatism of Kenyatta and other KANU leaders and their attachment to the notion of the sanctity of property, not to mention property itself. A point at issue between KANU and KADU was the composition of the Central Land Board, set up to take charge of land purchase and conveyancing to new settlers. KADU and the European settlers, both fearing a Kikuyu-dominated government, wanted a board of regional representatives and representatives of the vendors, while KANU wanted a board under central government control. Initially, KADU had its way and Presidents of the Regional Assemblies set up under the Majimbo constitution were to nominate settlers. Settlement money became the responsibility of the Settlement Fund Trustees, a committee of the Ministers of Finance, Agriculture and Settlement. The Land

Aid in Action

79

Development and Settlement Board was wound up and its functions were taken over by the Department of Settlement in the Ministry of Lands and Settlement. However, the Republican constitution which became effective in 1964 abolished Regional Assemblies. The Minister of Lands and Settlement then had final authority in selecting candidates for all schemes. The Central Land Board was wound up in 1965, and its residual powers went to the Settlement Fund Trustees. The pattern of land settlement was thus established for five years and policy did not change until after the reports of a British government mission to Kenya in 1965, led by Maxwell Stamp, which gave an account of the progress of settlement and made recommendations for future policy. The Schemes in existence up to that time were: High Density Settlement: Planned to cover 987,000 acres, the settlers to be landless and unemployed, required to provide only £5 to £7 to include stamp duties, registration and conveyancing fees. Target incomes ranged between £25 and £70 a year. Low Density Settlement: Planned to cover 180,000 acres of under-developed land of high potential. Settlers were required to have agricultural experience and a significant amount of their own capital. Target income was £100 a year. The earlier ‘Yeoman’ scheme with target income of £250 a year was absorbed into this programme. On both these schemes settlers were given Development Loans at 6½ per cent yearly interest for investment and as working capital. They were repayable over 10 years by 20 equal halfyearly instalments. Large-scale Farming Units: Not all the land purchased within the Million-Acre programme was suitable for subdivision, being either too dry or having too poor soil. A small number of co-operative farms was set up in such areas, mainly to undertake ranching. Assisted Owner Schemes: The Land Development and Settlement Board sold some farms to buyers with capital, intending to finance the scheme with IBRD and CDC money lent for the ‘Yeoman’ schemes. The lenders vetoed the scheme as not conforming to the loan agreements, but only after 125 settlers had been settled on 34,000 acres. The Nandi Salient: The Central Land Board bought 17,008 acres for £179,563 and handed it back to the Nandi tribe as having been wrongly alienated. Compassionate Farms: The Central Land Board began a programme of purchasing farms owned by aged and disabled people who because of their infirmity or location were thought to be in danger from squatters but who were unable to sell up and retire because there was no land market. The farms were either re-sold to Africans or included within settlement schemes. The Ol Kalou Salient: Here the Kenyan government asked Britain to finance purchase of some farms that were being abandoned and were overrun by squatters. The request was agreed because both governments feared a precedent from squatters establishing themselves freely on alienated land, and the risk to law and order of a forcible eviction was considerable. The Department of Settlement has since carried out largescale farming as the area has so far been considered unsuitable for small holdings. Finance for the schemes was provided as follows:10 The British government provided a grant for one-third of the value of land purchase and a loan for the remaining two-thirds repayable over 30 years at 6½ per cent interest. It also provided an administration grant and development loans for the settlers.

80 Aid and Inequality in Kenya: British Development Assistance to Kenya Table 11: HMG Finance for Million-Acre Scheme

High Density Scheme Land purchase Development loans Other costs Low Density Scheme Land Purchase Other costs Nandi Salient Total HMG funds

Loan

£’000 Grant

6507 4101 –

3283 – 4320

9789 4101 4320 18210

1263 –

632 1221



195

1894 1221 3115 195 21520

Total

Of the total sum, slightly more than £10m. had been disbursed before independence. At independence and in 1965–66, a further £2m. was lent to the Kenya Land and Agriculture Bank to make commercial loans to Africans for purchase of European farms on the open market without subdivision. Another loan of £1,275,000 was made in August 1964 for land purchase by the Land Bank in the 01 Kalou salient. The Department spent some £490,000 in 1962/63 for Compassionate Farm purchases, also funded by HMG.11 IBRD and CDC committed £2,471,000 for development loans on the Low Density Schemes in the ratio one-third from CDC and two-thirds from IBRD. The organisations imposed stringent conditions on the supervision and administration of the schemes before extending the loans. IBRD was not interested in any plan that did not allow for a net cash income above subsistence of at least £100, so reinforcing the pressure for low density schemes with greater agricultural extension services. Participation from international organisations was said to be desired by the colonial and British governments for political reasons—the leverage they could exert on an independent Kenyan government.12 The full IBRD and CDC commitment was never spent, and in 1972 the IBRD and CDC had disbursed K£l,080,000 and K£676,000 respectively. The West German government also committed development loans for the High Density Schemes amounting to £1,218,000. The Settlement Schemes had a number of objectives which sometimes contradicted each other. The Van Arkadie Report, commissioned by the Kenya government, identified five policy objectives13 which can be condensed into three: to redistribute income and wealth in favour of Africans, especially the destitute; to increase agricultural production by more intensive working of the land; to bring about a structural transformation of the economy. The first objective was politically necessary, and was the dominant principle behind the High Density Schemes. British farmers were paid for the land at 1959 prices—the last year in which the land market was regarded as normal. Valuations were sympathetically carried out by expatriate personnel. Only 141 farms changed hands in 1959 and those transactions became the basis for hundreds of later purchases.14 Part of the British finance was grant because it was realised that many of the permanent improvements and installations on the land, although affecting its valuation, would be of no use in the small-holder agriculture

Aid in Action

81

that was to succeed the British farmers. For the rest, the Kenya government had to repay the loans, and the new settlers were required to pay the Kenya government for their freehold, on the same terms as the British loans to Kenya. Instead of the African settlers receiving subsistence wages as labourers while the European farmer took the profits of the enterprise, the European farmer now took the profits in a capital sum, financed by HMG, which was then partly repaid via the government of Kenya from the profits of the new settlers working the same land. The Africans were theoretically to get subsistence still, plus a cash income of between £25 and £250 a year. It is not hard to see why the scheme commended itself to the British government: the European farmers could be kept reasonably happy, being paid for their land partly by the new settlers and partly by the British government. In the event it was the latter which did most of the paying because the implicit discount rate that capitalised the flow of future farm profits at the 1959 market price greatly exceeded the real interest rate on the British loan. The prices averaged roughly eight times annual profit, implying a discount rate of about 12½ per cent, while the loan was at 6½ per cent over 30 years and no allowance was made for inflation. Even without inflation the grant equivalent of the loan for land purchase was about one-third of the value of the land. Inflation changes that radically. If inflation of the money profits of the land averages 4–5 per cent a year over 30 years of the loan—which looks a very modest estimate at present—the grant equivalent of the loan equals almost two-thirds of the price of the land. The second objective was the one that exercised the Kenya civil service most and made it favour low density settlement which was thought much the more likely to further development. The then Director of Settlement, in his Department’s Annual Report of 1962/63, argued strongly that the optimum size of holding allowed for subsistence plus £100 a year, and nothing less, and the government should accept the selection of small holders and ‘not draw from a cadre of landless and unemployed’.15 Criticism of the schemes came from what might conveniently, if loosely, be termed the left and right. From the left it was argued that settlement in the form it was taking was ‘Rolls Royce in nature’,16 and free distribution to the landless should be made with smaller plots if necessary. This view was current in a section of KANU, and articulated increasingly in 1964 and 1965 by several politicians, notably B.Kaggia.17 From the right, after making ritual obeisance to the political and social imperatives, it was argued that subdivision was already going too far and too fast, that land was Kenya’s most precious asset, and systems of tenure should be such as to ensure its optimum economic use, within the political constraints.18 This was the ‘developmental’ view current among civil servants and the IBRD. This view, of course, interpreted the third objective of structural change narrowly to mean Africanisation of much agriculture with smaller holdings and more labour-intensive methods, but basically producing much the same crops for the market. During the middle sixties, the ‘right-wing’ view prevailed and has since dominated policy. It received definitive expression in the Stamp Report, published in October 1965.19 The Report emphasised that the transfer of land did not lead, of itself, to increased output and indeed caused a foreign exchange drain as it was a condition of the programme that the farmers could take their payment in Sterling. Kenya’s problems, including unemployment exacerbated by a population explosion, could only be solved by general economic expansion. It stated that great success in expanding output had been achieved where consolidation and registration of land-holdings had been effected and concluded

82 Aid and Inequality in Kenya: British Development Assistance to Kenya that an expanded programme of land consolidation should be undertaken. The returns to investment were thought to be greater in areas already farmed by Africans than in Settlement areas. A programme of land purchase ‘at a much reduced level’ was recommended as there was ‘an immediate need…for the maintenance of a market in land’. However, a pause in settlement was recommended: Although the settlement has done much to alleviate a difficult social problem of landlessness, the high density settlement schemes have not so far produced the levels of cash income for farmers which are necessary for the progress of the Kenyan economy. A pause in settlement for at least two years should be made to enable past results to be analysed and present and future practice to be improved.20

1966–74 The Stamp Mission report was one of the bases for negotiations between British and Kenyan Ministers in London in November 1965. The Kenyan background to these talks was one of increasing political dissension in KANU between the ‘radicals’ led by Oginga Odinga and Kaggia and the ‘conservatives’ led by Tom Mboya and, as it later turned out, enjoying the support of Kenyatta. Only four months later Odinga was to leave KANU and form a new party—the Kenya People’s Union (KPU)—one of whose platforms was the need for free land distribution.21 It would be surprising therefore if the Kenya government did not argue strongly the political necessity of continued land settlement, and claim that Stamp had underestimated the political factor, although the view that land consolidation and registration had a lower cost-benefit ratio in purely economic terms was accepted and expressd in Sessional Paper No. 10. Political unrest was also a factor in making European farmers eager to leave at an undiminished rate. Kenya government estimates were that in the event of a collapse in the land market some 100,000 acres a year would be abandoned, although this was probably a ‘negotiating’ estimate rather than a serious projection. In the event a compromise was reached. An interest-free loan of £6 3m. was agreed for Land Transfer, including £0.55m. for further settlement schemes. Stamp had recommended further land transfer at a maximum of 85,000 acres a year, but a rather higher rate of 100,000 acres a year was agreed. In April 1965, the Agricultural Development Corporation (ADC) was set up to undertake agricultural operations in the national interest on the direction of the Ministry of Agriculture. An ODM Mission went to Kenya as a result of Stamp’s conclusion that a review and subsequent strengthening of the Kenyan administration of land transfer was required. The Mission recommended that the ADC should be the vehicle for further land transfer. The ADC acquired farms and leased the majority to appropriate tenants (often farms were leased back to the European farmer from whom they had been bought) or farmed them itself until they could be sold to African private buyers. Some farms were retained and operated as National Farms where a specialised activity was involved such as maintaining a pedigree herd or cultivating improved seed varieties. The money which HMG lent to the Kenya government interest-free was lent on to ADC at 5 per cent. ADC tenants in turn paid rent of 6½ per cent of the farm’s valuation and 7½ per cent interest on ADC farm improvement loans. Kenyan demands for some sort of subdivisional settlement scheme were met by the Harambee programme of low density settlement to be carried out at the rate of 20,000

Aid in Action

83

acres a year for four years, a tenth of the rate projected for the Million-Acre schemes. The British were content that this should continue under the administration of the Department of Settlement only because the ADC was judged not ready to take it on.22 Officers in the Ministry of Lands and Settlement regarded the ADC as the brainchild of the Minister of Agriculture, B.R.McKenzie, and his attempt to get in on the settlement business as he was becoming resentful of the growing power of the Ministry of Lands and Settlement. British support for the ADC signified a loss of confidence in the latter ministry and in settlement schemes as such. The Africanisation of the Department of Settlement did seem to coincide with a decline in administrative efficiency. While the Million-Acre programme was carried through quite successfully, the much smaller Harambee programme lagged from the first, and by the time it was overtaken by a policy change, little of it had been completed. To that extent, Stamp’s belief that the Kenyan settlement administration was at full stretch was vindicated. Some of Stamp’s other conclusions, however, were subsequently falsified by events. An ODM team evaluating the Harambee settlements proposals in 1968 observed that there was little evidence to justify the widespread belief in the success of low density and in the failure of high density schemes in agricultural terms. Confirmation that in certain respects high density settlement had in fact turned out the more successful came in 1971 with the publication of the Ministry of Finance’s Farm Economic Survey.23 The Survey had sampled both types of farm on settlement schemes from 1964 to 1968. While the level of output per acre was higher on low density farms than on high density for three years out of four, the rate of growth of output per acre was much higher on the high density schemes. Net farm profits per acre were higher on the low density schemes for the first two years but were higher on high density schemes for the following two years. High density also showed a greater return on capital invested. And while most of the settlement farms failed to reach their target incomes, the shortfall was greater on the low density farm with targets of £100 and £250, than on the high density with targets of £25 and £40. The intermediate income group with a target of £70 a year did worst on this reckoning. Ironically, the Harambee schemes fell into this latter range. Later work attempting a cost-benefit analysis of the two forms of subdivisional settlement has tended to support the conclusion that the high density schemes have often given a higher economic return to investment than the low density schemes.24 The conventional wisdom came full circle in 1972 when the ILO report on Kenya quoted this survey and urged the further subdivision of settlement holdings, particularly on low density schemes, to increase labour absorption and the intensity of land use and thus raise output per acre.25 There were further changes in settlement policy in 1970. After McKenzie’s retirement the ADC was phased out of settlement by the Kenya government and restricted to managing existing state farms. The practice of leasing farms back to their former owners was stopped, and all settlement and land to be settled became the responsibility of the Settlement Fund Trustees. The Harambee scheme was stopped and only 431 families were settled on about 16,000 acres. A survival of the Stamp period, however, is the definition of a mixed farm, an enterprise involving mixed arable and livestock production as opposed to either a ranch or plantation. Britain, for the earlier schemes, agreed to finance the purchase of mixed farms as these are suitable for subdivision. Ranches are in general too dry to provide a livelihood

84 Aid and Inequality in Kenya: British Development Assistance to Kenya for small holders, and estate plantations growing permanent crops such as tea and coffee are usually on such fertile soil as to make their purchase prohibitively expensive. The MillionAcre Scheme was carried out on an ad hoc basis without ‘mixed farm’ being defined. A definition was laid down in 1967 that is still used. Table 12: Britain’s commitments to land transfer since 1970

1970 Land purchase for settlement: ADC and Agricultural Finance Corporation Special Scheme Total 1973 Land purchase for settlement Agricultural Finance Corporation Total a 2 per cent interest.

Grant

£m.

– 1.0a

2.5 0.25

Loan

3.75

6.0 7.0

1.0a

The loans to AFC are for on-lending to private buyers making free-market purchases of European-owned farms. The Special Scheme was for the purchase of farms which were becoming dilapidated or over-run by squatters, to prevent the need for the Kenya government to impose a management order.26 The land bought with these later British funds has been settled differently from the earlier schemes—on the Shirika or co-operative principle. Landless and unemployed people are settled on large farms along with the original farm employees. Each family is allocated a plot of about 2.5 acres while the rest of the farm continues to operate on a large-farm basis with a manager supplied by the Settlement Fund Trustees.27 Theoretically the responsibility for management is transferred progressively to the cooperative society of all the settlers. Originally they were supposed to be buying the land with a purchase loan but the policy may be adapted so that they hold the land on a lease, paying an annual rent of 5 per cent of the land value. The advantages of this form of settlement are continuity, the realisation of economies of scale, including the maintenance of large livestock herds and the fact that capital equipment on the farm, useless in the event of subdivision, can continue to be used and maintained. The co-operatives get development loans for the purchase of the farm’s loose assets and for other investment. Each settler is required to be resident on the scheme, and the annual target income for each family is K£60 plus subsistence, the money coming partly from part-time employment on the large farm and the remainder and subsistence requirements coming from the small holdings. While the scheme seems reasonable in principle, in many cases it appears not to be working as intended. One effect of providing the additional land for the subsistence plots is that the area of the original farm is reduced by perhaps a quarter. Labour requirements, far from being such as to provide work for the increased numbers of people on the farm, are therefore reduced. The majority of people who get jobs are those with specialised skills, like tractor drivers or mechanics, so the available work cannot be shared. The settlers are thus left on plots smaller than any on high density schemes and without the concentrated

Aid in Action

85

agricultural extension services available on the older conventional schemes. The farm manager has no time or inclination to concern himself with small-holder cultivation, being a paid official with no incentive to do so. When the development loan is paid off, the settlers should theoretically be in a position to ask the Settlement Fund Trustees to withdraw their manager so that they can run the farm themselves and take decisions about what proportion of the farm’s net profits should be re-invested and what proportion they will take in the form of dividends. This is not a situation that has yet arisen, however, and it is one that is regarded with considerable trepidation by many settlement officers. On some farms the lack of competence of the manager will maintain the farm in debt. One more able Shirika manager said that he just would not tell the settlers what the loan repayment position was until he thought they were ‘ready’. Yet the settlers took no part in the running of the farm at all. As the land area had been reduced, the manager was laying workers off, not using more. Almost all the men on the settlement left to find work in a nearby town where they earned wages that made it uneconomic to employ them on the farm. In fact the manager insisted that mechanisation would remain more economic and efficient on a farm of that size (some 4,000 acres net of the settlement) specialising in dairy production, even if wage rates were much lower. In the circumstances there seemed no reason why the settlers should ever be ‘ready’. In practice, therefore, the Shirika farms are run much like ADC state farms with bits carved off to provide small holders with the barest subsistence. At present the system works reasonably smoothly because the new settlers, being previously landless, are grateful enough for the small plots. There is a fear, however, in the Department of Settlement that they may in a few years’ time demand either a share of profits of the farm or an active share in its running, to both of which they are entitled, although both may be inimical to the farm’s profitability as a large unit. We received conflicting impressions of what would happen then. Some civil servants seem committed to Shirika as a way of preventing much of Kenya’s mixed farming land going to subsistence agriculture. Others maintain that the Department has an open mind and should Shirika prove unsuccessful, the farms will be subdivided. It is the IBRD view that subdivision should occur if Shirika encounters major problems, and it is our view that the schemes might very well do so, especially where they operate in areas where the most economic crop is not labour-intensive. The implications of Shirika do not appear to have been thought through to the point of reconciling the desire to maintain large-scale farming with the claims of the landless who have been given, so far notional, rights on these farms. The other important form of government-organised settlement did not use British funds. In fact it led to conflict with the British which, the Kenyans allege, caused a delay in the negotiation of general development aid in 1970. The Haraka settlement schemes (formerly called Squatter Settlements) were carried out on farms from which the European owners had b een excluded by a management order under the Agriculture Act.28 This Act gives the Minister of Agriculture power to take over inadequately managed or supervised land. It obliges him to run the farm and pay compensation to the owner or sell it and give the owner the proceeds minus a deduction for the costs of sale, but neither was done in fact. Instead, in 1965, a Commissioner of Squatters was appointed to settle the land so taken over. People who could prove they were landless registered with the Commissioner who could provide land for them. About 45,000 people registered. Settlement began on the farms taken over,

86 Aid and Inequality in Kenya: British Development Assistance to Kenya and it was done for the most part rapidly and cheaply with settler families getting small plots of about ten acres. In July 1971 the Department of Settlement took over the schemes from the Commissioner. The government s action drew protests from HMG and eventually was contested in the courts where it was ruled that acquisition had taken place and the Kenya Land Acquisition Act of 1968 applied.29 The Kenya government was required to pay the market value of the test-case farm plus a 15 per cent surcharge. Following negotiations it seems likely that the British government will allow the use of land transfer funds to pay for these farms, but not the extra 15 per cent above market price or interest payments owing because of the delay in compensation. Originally, it was planned to settle some 45,000 families on Haraka schemes, but later this target was reduced and the programme was ended after 14,635 families had been settled on 35 schemes totalling nearly 140,000 acres. One reason was agricultural fastidiousness; the schemes were unplanned and looked scruffy. More important, the political situation cooled down with the successful suppression of the KPU (whose genesis had almost coincided with that of the Commissioner for Squatters). HMG, which held up subsequent aid negotiations to discuss compensation for the land-owners concerned, did not help and the transfer of the schemes to the Department of Settlement, where they were not popular, doomed them. Nevertheless, Haraka was a cheap form of settlement as no farm planning was done before settlement took place; the settlers got no development loans and little in the way of extension services. Marketing co-operatives, fostered from the start on MillionAcre settlement schemes, are only now being started on Haraka schemes. The Director of Settlement commented in 1972, ‘these schemes continue to be a problem to my department.’30 Yet Haraka is not without its proponents. J.D.MacArthur, formerly of the Ministry of Agriculture, in an evaluation of settlement schemes,31 states that in view of their cheapness they may well have a higher benefit-cost ratio than other schemes. In effect the administration and other inputs lavished on other settlement schemes have not had very high returns, so Haraka shows up well. As the Haraka type of settlement might represent what the Kenya government would have done on a larger scale if British funds for land transfer had not been available, that is expropriate the land against some promise of future payment and settle it rapidly by subdivision into small plots with little planning, it is tempting to regard Haraka as a yardstick, an indication of what might have been on a larger scale. However it would not be appropriate to take yield estimates or estimated patterns of production on Haraka settlement and compare them with the results on other sorts of settlement scheme, because in general the land taken over for Haraka settlement was not mixed farming land. Many of the 50 or so farms taken over were formerly ranches or sisal estates. Haraka land is usually extremely marginal for small-holder arable cultivation, and as no development money has been available, the settlers have not been able to buy livestock. Haraka schemes in general seem to be less well developed than high density schemes of the same vintage but that is attributable less to lower initial expenditure and planning than to a harsher environment. On one scheme at Munyu, near Thika, Central Province, the rains had failed for four years out of five and the settlers had had to leave to get work to live. Not surprisingly most of the ground in the plots was uncultivated. It was the opinion of a good many settlement and agricultural officers interviewed, however, that when conditions were favourable, Haraka

Aid in Action

87

settlers achieved just as high returns per acre as conventional high density settlers, even with no high-grade cattle as supplied elsewhere, less extension advice, no special development finance, and inferior roads and other infrastructure. However a high proportion of output was consumed on the holding and less was sold for cash. These impressions do have implications for an assessment of the effect of British funds on post-settlement farming. There are no reliable output figures from settlement schemes later than those given in the 1964–68 Farm Survey, although there are figures showing sales through settlers’ co-operatives.32 These are an imperfect guide to output as they obviously exclude subsistence output and private sales, which are important for many products. With these reservations, they appear to reveal increasing output which exceeds pre-settlement levels in many areas. This is partly due to technical advances, notably improved seed strains for maize, in particular. It is quite possible that output per acre would have increased even faster had there been no transfer and settlement.

The Effects of British Aid on Land Settlement The Land Transfer Programme to many people simply means the Million-Acre Scheme. Yet over 40 per cent of British money for LTP and nearly 60 per cent of money for land purchase has gone on subsequent programmes, many of which involve little or no subdivision of landholdings. Quite a lot of the money therefore has financed not a land reform whereby the system of land tenure was changed, but a simple transfer of ownership. Large-scale farming has not been a great success in Kenya.33 The transfer of large farms is also open to more serious corruption than the allocation of small holdings. In retrospect it is probable that British policy took a wrong turning with the Stamp Report. There is clearly a role for a large-farm sector but given the shortage of experienced entrepreneurial farmers and the abundance of peasants, the balance between subdivision and ADC farms was too much in favour of the latter after 1966. While the Million-Acre Scheme was not a perfect land reform it was more thorough-going and probably more successful, socially and economically, than those aided programmes that have followed. It cannot be concluded however that HMG is entirely to blame for the systems of tenure that have evolved in the former Scheduled Areas. British aid has had a declining effect on the form of settlement that has taken place on transferred lands since independence. The Million-Acre Scheme was devised and largely executed by Europeans in the Kenyan administration in close consultation with the British government. The form of settlement in the Stamp period was partly determined by the unwillingness at that time of the British government to make more funds available for subdivisional settlement out of a belief, still regarded as justified in some circles, that settlement had gone too fast and strained the planning, administrative, and extension services available, and another belief, much more controversial, that high density cultivation was not a success, especially in the sense of not providing a basis for loan repayments. The Kenya government itself, however, was ready enough to curtail Haraka and Harambee schemes once the political goad of the KPU had been removed, and Shirika is its own conception. Many influential Kenyans prefer large-scale farming because of its appearance of modernity, the supposed greater contribution to marketed output, and perhaps because of its consistency with an elitist political and economic philosophy. ODM

88 Aid and Inequality in Kenya: British Development Assistance to Kenya has in any case increasingly opted out of intervening inland settlement as such. Since 1970, funds have been entirely for land transfer and no money has been available for settlement. One reason given is that this is consistent with the general policy of insisting that recipient governments have a financial stake in any project undertaken using British aid funds, as a guarantee of continued interest in the project. It has been argued that transfer and settlement are a single project, and Kenyan finance for settlement is an appropriate division of financial responsibility. This argument appears to be a rationalisation. HMG decided that it was impolitic to interfere in the form and institutions of land settlement because of the high political sensitivity of land. LTP was regarded as different from other aid and ODM handed it over and asked no questions. Having relinquished influence over settlement, however, ODM did not trust the Kenyan institutions to administer development finance, so did not provide any. There was a feeling that the Department of Settlement had become a self-perpetuating body which continued to grow while its proper functions diminished. In fact that was partly a consequence of the way finance was provided for the Million-Acre Scheme, as we argue below. ODM eliminated the grant for administration because, as subdivisional settlement came to an end, it was thought that the administrative expenses should also fall. The progressively softer terms of British aid for LTP—interestfree loan in 1966 and grants after 1970—were also seen as making a contribution to Kenya’s administrative costs. It is true too that HMG did not have the capacity to monitor at all closely the progress of settlement schemes. British aid is administered in Kenya by relatively few people and there used to be fewer. The First Secretary at the High Commission responsible for administering the Land Transfer funds also had, for years, to administer the general development funds. Much of his time was spent acting in a consular capacity to British farmers who came in to Nairobi to offer their farms for sale, or to complain about pressure being put on them to sell, or the prices they were being offered. It was administratively impossible for him to check up on the fate of farms after settlement, and all monitoring of settlement was done by a single Agricultural Attaché, who also had other things to do. HMG’s desire to distance itself from the operations of the Department of Settlement is understandable. British recommendations that the Department should employ more economic expertise, and offers of technical assistance, were coolly received; no technical assistance was used by the Department from the time that its established posts were Africanised and it still employs no economist. However, it was the British government’s insistence on financing land transfer for the Million-Acre Scheme largely by loan, and the financial structure devised by the colonial administration, consisting largely of British citizens, that was partly to blame for some of the subsequent problems. While the nonprovision of funds for settlement, as opposed to purchase, can be justified, it does give renewed credibility to the claim that Britain’s interest in land transfer is purely a concern for the welfare of European farmers—for all British money is now going to them. The cost of subsequent settlement is borne by the Kenyans alone. On most schemes land purchase accounts for nearly half of all expenditure; development loans and administrative expenses account for a quarter each. The principal problem with the Million-Acre Scheme was that, although mainly inexperienced farmers, the new Kenyan settlers were required to pay for the land. Development loans were therefore considered essential if enough cash surplus was to be

Aid in Action

89

generated to make that possible. The development loans were made in kind, not cash, but they too had to be protected, hence for fiscal reasons the scale of extension services in the settlement area greatly exceeded that for Kenya as a whole. That remains true up to this day. The last count of officers in the Department of Land and Settlement in 1971/2 revealed there were some 560 agricultural extension officers of the rank of Assistant Agricultural Officer and above (including equivalent livestock and animal health officers) serving 35,560 settlers in ten Settlement areas, a ratio of just under 1:64. This did not include Settlement Officers proper or co-operative officers, also on the strength of the department. The establishment has not changed since. For the rest of the country the ratio of Assistant Agricultural Officers to farmers was 1:700–1,000. There were in addition some 5,000 Junior Agricultural Assistants in the field, many of them untrained. As they were serving some 1.2m. small farmers, that still meant a ratio of 1:240.34 Originally the Department of Settlement was supposed to administer the Settlement Schemes for two and a half years. It became clear that a longer period of intense supervision would be required to supervise the loans. It was believed that as the Department had extended the loans it had to retain control of supervision and agricultural extension on the schemes, and if the Ministry of Agriculture took over, sufficient priority for the schemes could not be ensured. The period of the Department’s administration was thus progressively extended to five then seven years. It now seems to be accepted that the Department’s suzerainty over the Settlement areas is indefinite. While the Van Arkadie report accepted the need for continued supervision, it recognised that the nature of the administration of the Settlement Schemes would change and agricultural extension would become more important than settlement per se. Consequently it recommended transferring the Department from the Ministry of Lands to the Ministry of Agriculture.35 This never happened, probably because the Minister of Lands and Settlement, Mr. Jackson Angaine, was too well-established politically. It became increasingly necessary for the Department to obtain technical staff on secondment from the Ministry of Agriculture and other ministries. This was unpopular with the staff themselves, however, as they considered that they were being overlooked for promotion in their parent ministry while on secondment. The inevitable upshot was that they were established in the Ministry of Lands and Settlement to provide them with a career structure and promotion prospects in that ministry. The result is that rural administration and agricultural support in Kenya is sharply divided between Settlement areas and the rest of the country, with parallel pyramidical structures for the different areas. The Ministry of Lands and Settlement has become a quasi state-within-a-state, duplicating the functions carried out by a range of ministries elsewhere. HMG rightly noted and deplored this tendency but failed to realise, at any rate early enough, that the financial structure of support for the Million-Acre Scheme was partly to blame. Of course, it was only partly to blame. The provision of loan rather than grant finance posed a fiscal problem for the Kenya government but it was not bound to attempt to solve it in the way that it did, although in doing so it was continuing on a course set before independence. Strictly speaking, the problem of where and how to apply agricultural extension services to achieve the greatest gains in output is a separate problem from how to tax output so as to repay an external national debt. It is highly unlikely that such an uneven provision of extension services as Kenya has is economically optimal, and if it is, that is no more than a lucky accident, for the allocation was not made on the grounds

90 Aid and Inequality in Kenya: British Development Assistance to Kenya of economic rationality. Furthermore it seems unreasonable to attempt to fund the whole British debt by what amounts to a tax on the output of the new settlers. Although they obtained the land at subsidised rates they were far from being the only beneficiaries of LTP, as we shall argue in greater detail below. Charging the settlers much less, allocating rural services and credit on agricultural economic criteria, and using the administrative resources saved by winding up the Department of Settlement when its work was done, to expand the activities of the Department of Inland Revenue and later the Exchange Control division of the Central Bank, to reduce tax evasion of various kinds, would have been a better strategy economically and fiscally.36 That things were not done in a more sensible way must be explained on political grounds. The new Kenya government was just as anxious as the colonial administration to bring home to the Kenya masses that there would be no free land after independence. There was in any case considerable competition to get on to Settlement Schemes and had the cost to the settler been lower no doubt that would have intensified. By charging the settlers the government reduced the grievance of those not allocated plots and emphasised that land was not a birthright but an economic asset to be paid for. This also fulfilled the ideological function of suggesting that when it was paid for it was legitimately occupied by the payer, a matter of some importance to leading politicians, many of whom became large land-owners. Even after British funds became interest-free loans in 1966, subsequent land purchase and development loans to settlers carried interest at the old rate (e.g. on the Harambee scheme). And loans to Shirika farms have also carried interest although British finance has been on a grant basis since 1970. A possible justification is the need to relieve the fiscal burden caused by defaulting on the earlier loans. The Kenya government’s insistence on the sanctity of private property and the need for land to be bought, however, was a plank in its campaign against the KPU in 1965, and it may well be that this was the more important consideration. Nevertheless, perhaps these political and ideological objectives could still have been achieved with lower charges on settlers. The weaknesses of the strategy adopted are demonstrated by the poor repayment record of the settlers. Land and development loans to the value of about £16.7m. had been issued in 1972 (the latest available figures from the Annual Report of the Department of Settlement). From the first, the rate of repayment on bills due has been low. In 1964/65 it was47 per cent. The Department, however, charges interest at the 6½ per cent rate on the arrears, hence increasing the indebtedness of the settlers. The total amount which had been billed to settlers at the end of 1971 was about K£10.4m., i.e. over 60 per cent of the full value of loans, although about 60 per cent of the loans were repayable over 30 years and the rest over 10 years. The settlers are still paying bills at the rate of around 50 per cent. In 1973 they were some K£5.3m. in arrears.37 The Van Arkadie report38 estimated that if settlers could make annual loan repayments of only K£800,000, by 1989 the interest owing on arrears would exceed their repayments and the scheme would be financially unviable— the settlers’ debts would be increasing. Although the report regarded this as the outcome of pessimistic assumptions, reality has been even worse. Total repayments in 1971 were K£642,000 and in 1972 K£683,000. President Kenyatta’s two-year moratorium on loans for new settlers, announced in 1967, has complicated the calculation but some observers, including a World Bank team, believe the scheme is already financially unviable. As a result, the 1974–78 Development Plan said the government was considering not charging

Aid in Action

91

interest on arrears, an overdue reform urged by the Van Arkadie Commission in 1966.39 It is also proposed to offer settlers in arrears the option of converting their holding to leasehold, but that would probably be extremely unpopular. The Commission also urged a tougher line on defaulters, with increased evictions. Despite the political difficulties involved, a number of evictions subsequently took place. The Department’s Annual Report in 1966/67 recorded 76, compared with two up to the time of the Van Arkadie report. This policy was based on the premise that when all allowance had been made for the difficulty of settlers in meeting their loan repayments there was an element of refusal to pay based either on the belief that the land should be free or on a shrewd calculation that the government would be politically unable to evict. There were however quite sufficient grounds for predicting widespread payments arrears even if all the settlers had been anxious to repay. Billing was six-monthly, with the first bill arriving six months after the settler moved on to his plot. Few had by then generated enough cash income to be able to pay, so they started off in arrears, Even more important was the low level of production. This was owing partly to the inexperience of settlers and also to shortages in the early years of certain inputs and assets. Farm budgets were based on planned stocking rates, but in the early years of settlement there were fewer grade cattle available than the farm plans called for.40 Not all the farmers therefore had the cattle they needed to generate the target income. When the target cash income above loan repayment and subsistence is only £25 to £40 a year with high levels of husbandry required, it is clear that a shortfall in production, which is likely to occur, will immediately jeopardise the loan repayment capability of the farmer. Certain household expenditures such as minimal food, clothing and, in Kenya, school fees, are likely to be given precedence over loan repayment in any case, so in the circumstances it is surprising that loan repayment was as high as it was on high density schemes in the early years, especially as in the case of loan repayment, defaults by below-average farmers are not offset by extra payment by the above-average. On the other hand it is perhaps surprising that repayment rates have not risen more since. There have been other distortions. One of the channels for loan repayment was through a levy on earnings from sales through the settlers’ marketing co-operatives. While this was administratively convenient, it probably had the undesirable effect of retarding the growth of the cooperatives by encouraging settlers to sell their produce privately. Some extension officers believe that this effect is smaller now that settlers are used to the idea of co-operatives and appreciate their benefits. As a few farmers approach the point where they receive their freehold, this is expected to provide an incentive to the others, and once the inevitability of repaying the loans is accepted, the farmers find the co-operatives a convenient P.A.Y.E, way of doing it.

The Evaluation of Aid for Land Transfer Most of the British money has been given purely for land purchase and it is those funds we consider first here. Money for the administration of settlement or to finance development loans is considered further later. We consider the costs and benefits of the British finance for land transfer under two heads: the value and cost to Kenya; the value and cost to Britain.

92 Aid and Inequality in Kenya: British Development Assistance to Kenya

The Value and Cost to Kenya The net value to Kenya is the sum of the net extra benefits accruing to Kenya citizens as a result of the provision for land transfer of the finance, compared to the situation where it was not forthcoming. In order to assess these benefits, it is necessary to guess how the problem of foreign farmers would have been dealt with by the Kenya government in the absence of British finance. One thing the independent government could have done in theory is nothing. This appears to be the assumption of those on the British side who argue that LTP ‘saved law and order’. It would seem to be so that had the government stood by inactive, illegal squatting involving violence would have occurred on a large scale and, had the government attempted to resist it while offering no quid pro quo, the government might well have collapsed. For that very reason it seems completely at odds with the realities to assume the Kenya government would have done nothing, and the possibility can be dismissed. Assuming therefore that the Kenya government realised that it faced an absolute necessity to replace European farmers with Africans, there was more than one course open to it: (i) It could have undertaken to buy farms from their owners at what they considered a fair price based on historical market values. This appears as a simple transfer payment, but the real cost to the economy is likely to have been high. Many of the farmers would have left the country (as they did) and taken the money, leading to a foreign exchange drain. Without labouring the point further, therefore, we can say that if this were the policy adopted, British funds were worth their full grant equivalent—the full value of the finance minus the present value of interest and loan repayments. It might be objected that the Kenya government could have imposed exchange control or otherwise restricted the ways in which the purchase money could be spent. The market price of the farm was, however, determined under the colonial system where there was no exchange control between Kenya and the UK from where most of the farmers came. Imposition of exchange control or any other restriction on their use of the proceeds of sale amounts simply to paying less than the old market price and is ex hypothesi excluded from case (i). (ii) The Kenya government could have expropriated. Expropriation is here defined as the compulsory acquisition of assets at less than their historical market price (i.e. that before the process began). The expropriation of the farms would have reduced the costs to Kenya under (i) but would have given rise to other costs. Without proving the statement we can plausibly assume that these other costs would vary directly, but probably not proportionately, with the degree of expropriation—which becomes 100 per cent when the government kicks all the farmers off and gives them no compensation.41 The costs of expropriation would take the form of certain flows of production forgone over time which can be reduced to a single figure by discounting at an appropriate interest rate, In practice the government will not be indifferent to the interpersonal distribution of those flows. For example, the consumption of European farmers can be subtracted from the flow of production forgone as the Kenyan government can be assumed not to value it. There are a number of expropriation costs one can posit, but there may also be some benefits. The costs of expropriation are the net present value at the time of expropriation of the following effects:

Aid in Action

93

(a) A loss of production owing to the possible accelerated departure of expert farmers at a rate exceeding that at which they can be replaced, whatever the organisation of land tenure, i.e. a managerial input is being lost. (b) A loss of production owing to the running down and perhaps the deliberate destruction by farmers of productive assets during the time that they as a group are being removed. (c) A more indirect cost, the acquisition by the Kenya government of a reputation for arbitrary action inimical to business confidence and foreign investment. (d) A saving of administrative resources: if nothing had been paid for the land the government might have adopted a different pattern of settlement, e.g. it might not have attempted to retrieve the money from new settlers and so saved administrative resources spent on debt accounting and collection. (e) An effect on the political popularity and hence authority of the government: the morale of Kenyans might have been improved by the spectacle of their government taking a firm line, but against this is the vestigial capacity of the white settlers to make political trouble. In fact a UDI by white settlers on the Rhodesian pattern is sometimes said by British civil servants to have been averted by LTP. The need for metropolitan troops to combat Mau Mau surely demonstrates that a UDI was not possible. A Kenya government would have had to disguise intentions to expropriate, however, if independence were not to be delayed. Clearly if the Kenya government had done the best it could in the absence of foreign finance, it would have taken over the farms with just that degree of expropriation that balanced the marginal values of costs under headings (i) and (ii), i.e. when the loss in the present value of production owing to a bit more expropriation equalled the cash value of that bit of expropriation. The situation with which we might compare British-financed purchase at market prices, therefore, is the acquisition of the farms using the optimal degree of expropriation. The value of the aid for land purchase then is the difference between what transfer would have cost the government of Kenya expropriating optimally, and what the existing arrangements have cost.42 Unfortunately, we do not know what this optimal degree would have been, and it is fair to doubt whether the Kenya government would have known it either. At this range it is impossible to say what sort of deal a Kenya government acting in the majority interest would have offered the settlers. One Kenya civil servant insists that the full market price would have been paid in line with the constitution and the government’s own ideology. That is perhaps the most diplomatic thing to say when one is relieved of having to make the choice. His superior guessed some fraction of the 1959 market value might have been offered—perhaps a half, with payment by a medium or long-term bond. Prices of the farms that did change hands on the free market in 1960–61 were at half the 1958 prices anyway. The provisions of the Agriculture Act might have been liberally applied to farms which it was possible to claim were underutilised or mismanaged, perhaps with an amendment to make subsequent settlement legally possible. What would have been the expropriation costs of such an arrangement? If farms were simply taken over on being abandoned, while some payment were made on compulsory acquisition, it makes sense for all the owners to stay as long as possible in order to minimise their losses. In practice that does not appear to be the way people always behave. In Kenya in the early 1960s, there were some desertions anyway—for example, from many of the

94 Aid and Inequality in Kenya: British Development Assistance to Kenya farms on which management orders were placed. If a farm is not making much money or requires new capital investment, the owners may not be prepared to make the necessary outlays in order to recoup what the government has promised, because their confidence has been shaken sufficiently to doubt the reliability of any deal proposed. Settlement Officers, usually white, ran the farms taken over for the Million-Acre Scheme before the land was settled. Such an arrangement on a larger scale could have stretched managerial resources. The competence and even honesty of the early Settlement Officers was very varied anyway. Another source of loss is likely to have been the running down of the farm—not replacing capital equipment, allowing permanent improvements to deteriorate, and not using fertiliser, so reducing soil fertility. Some Kenyans have argued that the effect of expropriation on running down would not have been great because European farmers did it anyway. Every district settlement or agricultural officer has some story of a departing farmer who deceived valuers about the age of cattle or replaced his best stock with inferior animals after the valuers had called, and there are also stories of malicious damage by farmers. Apparently some allegations of malicious damage, especially in the Machakos area, have their origin in the floods of 1961 which naturally destroyed many groundworks like dams. In fact of all the cases of alleged stripping after the visit of the valuer, only two were established.43 Usually the allegations related to farm houses rather than the farms themselves. This is not surprising as there was a ceiling price on farm houses that was below the market rate. The Settlement Officer taking over any farm had the right and duty to query anything on the ground which differed from the valuers’ report of which he had a copy. Nevertheless, it can be inferred that some cheating occurred from the fact that for transfer schemes after 1966 valuation procedures were tightened up. Veterinary inspection of herds was introduced along with a second inspection by the valuer who now returned to the farm with Settlement Officers on the day of the handover. Whatever the scale of cheating, most of it occurred in the early years of settlement. A more important reason why the costs of running down could probably have been small in settlement areas is that much of the capital of farms was in any case useless for the kind of farming that followed: fencing and piping were usually in the wrong place for small holdings and could not be moved. Machinery was inappropriate and needed too much maintenance. A graphic example of this is provided by the experience of a farmers’ cooperative at Endarasha in the Kieni West Division of Nyeri District, Central Province. The European farm which was taken over had an irrigation system depending on a motorised pump operating from a river dam. No one on the settlement scheme could maintain the pump, which fell into disuse. For ten years many of the small holders have walked several miles for their water. Now the co-operative is attempting to save Ksh3m. by a levy on its members to install a reticulated irrigation system depending on a hilltop reservoir and gravity feed.44 The uselessness of much on-farm capital for settlement purposes was indeed the reason why Britain gave one-third of the Million-Acre land purchase money in the form of a grant. The point that emerges is that the loss of production owing to expropriation would have been small or negligible after settlement. The effect would have been felt in the large-farm sector—in those areas not settled at all or in areas settled later, before they were settled.

Aid in Action

95

Without the land purchase money there would have been no recovery in land prices and no investment in the large-farm sector. In any event it took until about 1967 for money prices to reach the 1957 level again. The ways the Kenya government could have countered such an agricultural depression would have been to raise food prices considerably— with political dangers—or to increase the rate of settlement—Haraka style. More rapid subdivision might have involved breaking up breeding herds and could have led to other longer-run dangers such as increased soil erosion, owing to more intensive cultivation and less supervision of output. More rapid take-over of large farms by Africans could also have led to declining yields—as it did even at the rate experienced. Faster take-over might well have resulted in an output performance even worse than that of discouraged white farmers running their holdings down. We cannot proceed rigorously and quantify the direct costs of expropriation,45 but we can estimate the order of the value of production that would have been endangered by expropriation—the output of white farms for, say, seven years after independence—and compare it with the market price of the farms purchased. The first difficulty is knowing where to draw the line. The main effect of expropriation would have been on those mixed farms not bought up. Would it also have depressed output in the rest of the large-scale agricultural sector and on plantations and ranches not immediately threatened, and even expatriate business generally? It is highly probable, but we shall ignore it for the moment by assuming the government could have made clear its expropriations were confined to a ‘special case’ sector. This clearly imparts a considerably conservative bias to our evaluation. The Van Arkadie report estimated the value of output from the mixed farming sector at £15m. in 1961.46 We have used a similar method to estimate the value of output from the large mixed farm sector for the years 1963–71. (See Table 13.) Only a part of this came from European-owned farms, and it is necessary to estimate that proportion. The 1968 Economic Survey and the 1970–74 Development Plan47 give two estimates of the proportion of the mixed farming area still in white hands, and using these as benchmarks we have assumed that, of the remaining large farm areas after deductions for settlement, 7 per cent a year was taken into African ownership up to 1967 and 5 per cent a year thereafter. It would not be reasonable, however, to take these estimates of the proportion of the mixed farm acreage in white hands in each year as estimates of the proportion of output coming from white-owned farms, as the latter were undoubtedly more efficient than their African-owned counterparts during this period.48 In the first place, it tended to be the more efficient white farmers who held on while the less efficient were content to be bought out and as surveys showed, African farmers who took over large farms often ran into difficulties. On the basis of farm surveys we have assumed that the value of output per acre was 50 per cent higher on white farms than black. This assumption, combined with the acreage estimate, gives estimates of the proportion of total mixed-farm output coming from white farms.49 This output, however, does not represent the social value of continued production; a number of adjustments have to be made. The price of agricultural commodities does not always represent their true social opportunity cost. Many agricultural products are marketed in Kenya by Marketing Boards to which the large farmers often have privileged access. The prices charged to consumers often included a subsidy element to the producer

96 Aid and Inequality in Kenya: British Development Assistance to Kenya to the extent that overproduction resulted and was often exported at a lower price than domestic output. Certain important commodities have been revalued at social accounting prices based on the import or export prices of traded goods in an attempt to get nearer to the true social value of output.50 Table 13: Estimated value of production by large mixed farms in Kenya, 1963–71 (current prices) £m. 1963–1967 K£m.e 1968–1971 a a a b b 1963 1964 1965 1966 1967 1968c 1969c 1970c Wheat 2.95 Maize 2.05 Other cereals 0.51 Total cereals 5.51 Cattle and calvesd 0.61 Sheep 0.22 Pigs 0.50 Poultry, eggs 0.27 Pyrethrum 0.77 Other temperate cropsf 0.44 Milk 1.86 Butterfat 1.41 Butter, etc. 0.01 Total dairy 3.38 Total 12.08

3.50 0.98 0.47 4.94 0.56 0.19 0.48 0.25 0.49

4.16 1.03 0.49 5.68 0.53 0.17 0.47 0.22 0.71

3.17 1.24 0.32 4.73 0.55 0.19 0.54 0.20 0.74

4.08 2.87 0.30 7.25 0.54 0.22 0.44 0.20 0.42

6.50 2.92 0.30 9.72 0.60 0.20 0.40 0.20

6.50 4.90 2.00 1.50 0.30 0.30 8.80 6.80 0.60 0.60 0.20 0.20 0.35 0.30 0.20 0.60 Immaterial

0.37 1.78 1.16 0.01 3.07 10.69

0.38 1.92 1.10 0.01 3.15 11.64

0.33 2.70 1.01 0.01 3.83 11.53

0.31 2.88 2.03 0.01 5.13 14.97

0.30

Immaterial Immaterial immaterial Immaterial 4.00 5.00 14.40 13.65

5.00 16.57

1971c 5.10 2.00 0.30 7.30 0.60 0.20 0.25 0.60

6.50 15.60

Sources: a 1967 Abstract of Statistics, Table 74, p. 77. b Economic Survey 1968. Table 4.7, p. 58. c Estimated from total agricultural output figures, various Economic Surveys; large farm proportion in total taken as average of 1963–67 proportions for wheat, other cereals and livestock except for cattle; a slight decline in the large farm share of maize, cattle and dairy products was allowed for by rounding down. d Van Arkadie’s method (op. cit., Chapter 2, para. 113) was followed by assuming one-quarter of cattle sales came from mixed farms—the remainder from ranches. e Average exchange rate: £1=K£0.85. f Pulses, potatoes, fruit, vegetables and flowers, plus oilseeds.

A deduction has to be made for the cost of inputs. The input-output table for Kenya for 1967 was used. Value added was taken as 0.82 per cent of the value of marketed output.51 Much of the value added, however, accrued to the European farmer in the form of profits. Input-output data show that in the agricultural sector some 60 per cent of value added is profits. This figure was assumed to rule for the European mixed-farm sector. About 13 per cent of these profits were taken by the government in direct taxation and are therefore of full value as a social benefit.52 The rest of value added consists largely of wage payments and

Aid in Action

97

these, too, were valued at their full amount. Net profits were treated differently. In the first place the consumption of European farmers could be regarded as valueless from a Kenyan point of view. Their savings, however, especially after exchange control was introduced in 1967, probably financed some investment of use to the country. No data exist on the savings propensity of European farmers as a group. As self-employed businessmen they might be expected to save quite a lot in the ordinary course of events, but it is doubtful whether they were doing so in the early 1960s and if they were, they were probably remitting their savings abroad. In the end we assumed one-quarter of net profits went into investment useful for Kenya, so post-tax profits were valued at a quarter of their sum at accounting prices. This is clearly a rather arbitrary procedure but has to do in the absence of fuller information. This adjustment completed the derivation of what might be called the ‘social surpluses’ realised in each year. These were deflated to 1963 prices using the middle-income consumer price index,53 then discounted to a present value, as of 1963, using an accounting rate of interest of 10 per cent. Land purchase expenditure for 1963–73 was similarly deflated and discounted (see Table 14). The calculations indicate that the present value of land purchase money expended was some £13m. in 1963. The 1963 present value to the country of European farm production during the forthcoming nine years was about £27m. Obviously both these figures are subject to wide margins of error. They indicate that the output safeguarded was worth about twice the cost of doing so, hence the saving from a 10 per cent degree of expropriation would have been wiped out by a fall in the 1963 present value of output of slightly less than 5 per cent. Of course, this says nothing about how farmers would have reacted to any degree of expropriation; we have no means of estimating any behavioural equation of that kind. This merely shows roughly what sort of response would have been necessary to wipe out the gain from any expropriation. It seems that paying the market price or something like it might well have been good tactics for the Kenyans, even in the absence of British finance. There are many plausible scenarios. If, for example, the output of the white farms had fallen to half its actual figure for eight years after independence (owing either to the lack of interest of the whites or the inexperience of those succeeding them), the gain from a complete expropriation would have been wiped out even if output thereafter rose to the level it has actually reached. The optimal degree of expropriation, therefore, might well have been small.54 Some of the assumptions made in reaching the estimate can be varied to test sensitivity. If productivity on white farms had been merely average for the large-farm sector, for example, 1963 present social value of European farm production would fall to about £23m. Table 14: Costs and potential benefits of land transfer A Social surplus from white mixed farms, 1963 prices 1963 1964 1965 1966 1967 1968 1969 1970 5.47 4.42 4.36 3.58 3.91 5.02 4.05 3.27 Present value in 1962 of these flows with a discount rate of 10 per cent: £27.03m.

1971 2.96

Source: Output as in Table 13: (i) each line multiplied by accounting ratios (see note 50, p. 143); (ii) yearly totals multiplied by a factor (as in note 49, p. 143), to obtain share of white farms; (iii) multiplied by 0.82 (proportion of value added in output) to give value added at accounting price; (iv) multiplied by 0.59 to derive social surpluses at accounting prices derived as follows:

98 Aid and Inequality in Kenya: British Development Assistance to Kenya Social surplus=0.39(VA)+0.13(0.61)(VA)+0.25(1–0.13)(0.61)(VA)=0.59(VA) non-profit component tax on profits 25 per cent of net profits of value added (v) deflated by consumer price index (as in note 53, p. 144) to obtain social surplus. B Land purchase expenditure, at 1963 prices 1963 1964 1965 1966 1967 1968 1969 1970 1971 3.1 3.91 2.788 0.897 0.884 0.925 0.914 0.893 0.840 Present value in 1963 of these flows with a 10 per cent discount rate: £ 13.24m.

1972 1.008

1973 1.653

Source: Basic statistics from Department of Settlement Annual Report; deflator price index: as above.

If European farmers’ savings, as well as their consumption, are valued at zero, the 1963 present value of production falls to £21m. If both of these new assumptions are made together, a figure of £18m. is obtained. All of this merely considers possible production losses, but what about the effect on foreign and other investment? On independence foreign investment fell away but then recovered (see Table 15). Table 15: Private capital flows in the independence period

Net Long-term private capital movements International investment income

K£m. 1961 1962 1963 1964 1965 1966 2.2

−0.2

–9.4

−15.0

1.5

1.0

−6.1

−7.0

−9.2

−9.0

−9.0

−12.5

1967 1968 1969 8.0

9.1

16.0

−14.0 −14.1 −16.2

Source: Government of Kenya Statistical Abstract, 1967 and 1973.

The so-called ‘direct effect’ on the balance of payments of foreign enterprise is usually negative in Kenya because profit repatriation (international investment income in the national accounts) usually exceeds the capital inflow. However, to look at this net figure is to make the same error as in the netting of aid flows.55 There is nothing more ‘direct’ about these effects of foreign investment, that show up as monetary flows in the balance of payments, than about those that are implicit in the trade figures. Foreign-owned enterprises are producing output that is either exported or would have to be imported or done without. While it is true that if the government acts competently in the national interest, it can usually prevent harmful foreign investment, in practice some suboptimal and perhaps harmful foreign investment has probably occurred in Kenya—that in enterprises which are heavily protected and have a negative value added at world prices so that their influence on the trade figures is also negative. But it is not just that sort of investment that would have been discouraged by a reputation for expropriation. More than capital, Kenya lacks managerial and entrepreneurial expertise which is supplied jointly with capital by foreign investors. It is an open question whether Kenya has obtained these inputs as cheaply as it

Aid in Action

99

might have done in view of the adverse financial flows and the transfer pricing associated with foreign investment. It is also an open question whether the management obtained was employed in the best activities from a social point of view. Much management was, however, obtained and Kenya could have dispensed with foreign investment only with a totally different kind of economy. Now we consider finance for the administration of settlement and for development loans. These cannot really be divorced from the land purchase money because they would certainly not have been forthcoming if Kenya had taken over the mixed farms with more than a small degree of expropriation. The development loans and administrative grant could quite simply be valued at their grant equivalent. However, that would probably be incorrect. The sums cannot be evaluated separately from Land Transfer money, for all the money helped to induce the Kenya government to a misallocation of resources, as we have argued. If it had not been forthcoming, and a more rational system of rural administration and agricultural extension had been adopted in the country, there would almost certainly have been considerable net savings to the Kenyan exchequer and possibly gains in agricultural output. These ought to be subtracted from the grant element of all the funds for the LTP in assessing their value to Kenya. While that is easy to state in principle, it is impossible to do in practice. Nevertheless the implication of considering all likely costs and benefits is that Kenya would probably have been well advised to pay close to the market price for farms taken over and British aid for land purchase was actually worth something near to the grant equivalent of the sums advanced. To this must be added the grant equivalent of the development and administration funds, minus something for a misallocation of resources. Even if we make the strong assumption that these cancel out and the development and administrative funds were worthless, we are left with a substantial sum. Over £12m. of the £20m. for land purchase has been grant anyway, and inflation has greatly reduced the real interest and repayment burden of British loans. Even assuming average inflation of only 4–5 per cent a year for the period of the loans, the £20m. of British funds were worth £18–19m. as aid to Kenya. The contention of Sessional Paper No. 10 is false. The benefits of these funds to Kenya greatly outweighed the debt burden. There is an objection to the foregoing analysis which takes as its starting point the fact that the rate at which settlement could proceed safely was itself a decreasing function of the money paid to the European farmers. The more compensation they received and repatriated, the more the Kenya government had to tax future output in order to pay the compensation. Had the European farmers received nothing at all, the land could have been parcelled out free to new settlers with a minimum of planning. Then, it is argued, the whole of the mixed-farming area could have been settled very rapidly. Marketed output would have fallen, and Kenyan agriculture would have moved largely to a subsistence basis, but the majority of the population would not necessarily have been worse off. There is, of course, nothing inherently illogical about this. In 1961 Kenya had a typically colonial economy and in many ways the structure of the economy has hardly changed. Expropriation of all land and its free distribution would have destroyed this economy and led to the atrophying of many industries that existed in Kenya. Such a policy was not even considered by the colonial civil servants who framed the LTP. More recently much thought has been expended on how the existing economy might be made to serve the interests of the

100 Aid and Inequality in Kenya: British Development Assistance to Kenya majority better, perhaps in the teeth of the interest groups spawned and maintained by it.56 Of course an agriculture given over to peasant subsistence, while it might have supported the population as a whole at better nutritional levels than the export enclave agriculture has done, equally well might not. It would have been less diverse than the current economy and perhaps less resilient in the face of natural disasters like drought. It would have posed grave developmental problems of its own. Notably, it would have been difficult to tax such production or realise the savings necessary for productive investment or industrialisation. The effect on the price of foodstuffs to the small but politically vital urban population would also have been critical. But perhaps it is not necessary to speculate here whether it would have been better from the majority viewpoint to thus ‘go back and start again’ rather than attempt to develop along the existing lines, because this was not really a live political issue in Kenya at the time of independence. What we are prepared to say is that once the line of existing development was chosen, the optimal economic policy in the early years of independence, if not later, was to pay expatriate farmers something close to a market price for their holdings. If, as we suspect, Kenya would have developed on similar lines even without the money provided for land purchase, then it is not necessary to judge how far British interests were in conflict with those of the Kenyan masses (i.e. to evaluate the suitability of a capitalist form of development for Kenya) in order to conclude that LTP funds were a useful form of aid.

The Value and Cost to Britain Much of the money that Britain forwarded for land purchase was of course repatriated. The Exchange Control Office of the Central Bank of Kenya has a record of remittances of proceeds of sale of farms bought with British aid funds. In the period January 1967 to August 1974, K£5.15m. was remitted, of payments by Settlement and Trustees, ADC and AFC. British disbursements in the period 1967/68 and 1973/74 (British financial years) on the Land Transfer Programme were £7.83m. (ODM figures). At an average exchange rate of £1=K£0.85, this gives a repatriation rate of 77.4 per cent. However, this figure assumes disbursements from Britain are in respect of purchases during roughly the same period, whereas there is usually a lag between the British national taking his proceeds in Sterling and repatriating them and the Kenya government requesting reimbursement from HMG, so some of the repatriations have been covered by subsequent disbursements. A more accurate figure is probably obtained, therefore, by omitting remittances made in 1974 as those were almost certainly in respect of disbursements not registered in the statistics. That reduces total remittances to K£3.93m. and the repatriation rate to the region of 60 per cent. It would not be accurate to say that the money repatriated was not a net cost to Britain, especially if the Kenyan government would have bought the farms anyway. In any case some has been subsequently re-exported to Australia, South Africa and other places. If the costs of the programme have been modest, the advantages to Britain are largely political and psychological. A Conservative government was relieved of an embarrassing situation involving land-owners who had strong connections with its own supporters, and some of whom could justifiably claim to have been misled by the British government.

Aid in Action

101

It could also be argued that by making the money available, Britain influenced the political and economic direction that an independent Kenya would take. It was certainly in British manufacturing, ranching and plantation interests that the Kenya economy should continue on its predetermined path, but it is impossible to know what the importance was of these funds in themselves. Aid is only one of a complex of relations between developed and emergent states and is usually given too much importance by commentators. As the ILO report says, ‘…Kenyan attitudes and aspirations had perhaps been moulded more than was realised by the style and ethos of the divided economy, by the colonial experience of having had to accommodate oneself and to work within the existing structure of the economy rather than to change it. Thus, when national independence was achieved, the political aim of taking over the economy became merged almost imperceptibly with the individual aspirations to take over the jobs, positions and life-styles which the economy made possible.’57 That acculturisation of the Kenyan elite had already occurred. At the time of independence, was there a politically significant group aware of the implications of a radical land policy and prepared to accept them? There was certainly a populist group in favour of free land distribution, but there was not much debate on the issue before independence and it was not until 1965 that the populists left KANU to form the KPU. This may have been a result of the LTP, of course. Even then, however, the KPU policy had not been thought through to the extent of accepting the revolutionary social and political changes it entailed. Consequently it was easily stigmatised as irresponsible.58 While it is clear therefore that the provision of LTP finance on concessional terms was absolutely consistent with a policy of protecting British interests in general, it is not really clear that the provision was critically necessary for the protection of those interests. We cannot know that the degree of expropriation would have been high in the absence of aid, or if it were, that this would have permanently undermined British capitalism in Kenya.

Appendix: A Model of Expropriation In this appendix we justify, rather more formally, the assertion that the optimal degree of expropriation was probably low in Kenya’s mixed-farm sector in view of the output at stake. What follows may have application to other land reforms or expropriations. Information in markets is never perfect so the notion of a market price is usually vague. It seems plausible to assume, therefore, that farmers’ reactions in terms of reducing investment, running down their holdings or even leaving, would be moderate in response to a small degree of expropriation. As the degree of expropriation increased, the reactions would intensify as more farmers reacted to the low prices being paid and as the reactions of individual farmers became more violent. At a certain degree of expropriation almost all the farmers would be running their holdings in an extractive manner if they had not left altogether, so further increases in the degree of expropriation could give rise to only a small further reaction, We define e, the degree of expropriation, as (M—P)/M, where M is the market price and P is the price actually paid for farms, and we define r, the reaction, as measured by the loss of farm output compared with normal running. Given the pattern of response hypothesised, plotting e against r diagrammatically would give a rising curve with a point of inflexion.

102 Aid and Inequality in Kenya: British Development Assistance to Kenya

Figure 1 So far we have said nothing about time. The curve could be drawn in Figure 1 on the assumption that the government took over farms at a constant rate of acres per year. That would not be very intelligent, however, because the optimal speed of expropriation from the government’s point of view almost certainly varies with the degree of expropriation. It is well known, for example, that if a complete expropriation is planned (e=1) it is better to do it very quickly. If the government varies the speed of taking over land so that the speed of expropriation is always best, given the degree of expropriation, then the curve in Figure 1 will be rather flatter. Then the reaction curve would indicate the loss of output owing both to the discouragement of the farmers being expropriated and the inexperience of those succeeding them. In what follows we consider only the degree of expropriation and assume the government expropriates at the speed appropriate to the degree. Our reaction curve therefore charts the smallest possible output losses for any degree of expropriation. Having introduced time, we have to work in present value terms; savings on purchase owing to expropriation and losses owing to reduced farm output are considered as the present value of these cash flows at the start of the expropriation, In the context of land transfer in Kenya the maximum value of e (=1) is worth the 1963 present value of the Land Purchase funds, some £13m. (as the take-over was in fact at market prices). The maximum value of r is where all output from farms to be expropriated is lost and we have calculated that as having a 1963 present value of about £27m. Normalising the axes of Figure 1, therefore, with e=1=£13m., the maximum value of r=2 approximately. The simplest form of reaction function having a shape as we have drawn is a cubic form. To identify it further we can specify the following boundary conditions consistent with our behavioural assumptions: (a) r = ae3 + be2 + ce i.e. there is no intercept term—the reaction begins with the expropriation.

Aid in Action

103

i.e. e=0 and 1 are minimum and maximum points respectively of the function. This defines a set of cubic functions of the form: (1) where x is an unknown parameter. It is constrained to give a value of r between 0 and 2, as e ranges between 0 and 1, so it must lie between 0 and 6. As noted, in effect it is a function of the speed of expropriation, but we assume that that is chosen to minimise x. The task of the Kenya government in the absence of LTP would have been to minimise E, the cost of taking over the farms, where: E = (1−e) + r

(2)

To find the minimum value of E, set the first derivative of E with respect to e=0 (the usual first order condition for identifying an extremum). This gives: (3) where e* is the value of e that satisfies the necessary condition for a minimum. The second order condition for a minimum is: d2E/de2>0, from which we derive that:

These results are, however, inadequate to determine the best policy which may be a corner solution, not shown up by calculus. In Figure 2 we plot reaction functions (1) with their associated expenditure functions (2) for three values of x. Note E=r when e=1, so the curves meet at that point. For values of x=x2 and x3 the minimum points on the expenditure function are at a2 and a3, equivalent to low rates of expropriation (e*2 and e*3). For the lower value of x=X1, however, the point a1, although a local minimum, has a higher value of E than when e=1. For values of x < 2, (3) has no real solution—the expenditure curve falls monotonically in the relevant range and complete expropriation is optimal. Higher values of x still give a corner solution although real solutions to (3) exist. To determine the critical value of x below which a corner solution is optimal, reflect at e=1:

This must equal the value for E when e=e*. Hence substituting e* for e, and (1) into (2):

104 Aid and Inequality in Kenya: British Development Assistance to Kenya (e* −1) is a factor of this cubic equation which can be written:

Figure 2

Aid in Action

105

(e*−1) (2xe*2/3−xe*/3+[1−x/3])=0. Ignoring the solution e*=1, we can treat the quadratic expression in e* as=0. The solutions to the quadratic equation are an expression in x which must, however, equal the expression in x in (3). Equating the two expressions and solving for x gives the result that x=2.66. That is therefore the critical value for x, below which complete expropriation becomes optimal.

Conclusion If x in the farmers’ reaction function is below 2.66, it pays to expropriate them completely, making no payment. If x lies in the possible range 2.67–6.0 the optimal degree of expropriation is low. Where x=2.67, for example, the value for e*, the optimal degree of expropriation, is 0.25, i.e. the price paid to farmers should be 25 per cent below the market price. The savings entailed by such a course of action on paying the full price is, of course, the money saved by not paying the full price minus the value of lost output, or e−r. Where e is 0.25, r=0.139, or almost 14 per cent of the market price and hence the saving is 11 per cent of the market price. As x increases, optimal expropriation falls until at x=6, e*=0.09. That is, it is best to pay 91 per cent of the market price. In the Kenyan case, however, given the empirically calculated limit on r it could not have been worth paying more than that. Savings then would have been about 4½ per cent of the market price. So, while the precise figure depends on farmers’ reactions, it would have made sense to take between 9 and 25 per cent off the market price or to pay nothing at all. It could not have been optimal to do anything else. The savings on paying the market price achieved by any degree of expropriation are always less than the degree of expropriation by the amount of the reaction. When expropriation is total, the reaction becomes the whole cost of taking over the farms. Even then the savings over paying the market price will not be very great unless x is quite a lot below 2.67; when it is 2.66, and r=0.89, for example, total expropriation, while the best course, is only about 11 per cent cheaper than paying the full price. Now, British aid was worth the difference between the cost to Kenya of an optimal expropriation and the cost of the existing arrangements (viz. the market price of farms minus the grant equivalent of the aid). The value of the aid could have been very low or even negative if x was low or even=0 (i.e. if an expropriation would have been costless). To assert that total expropriation would have been much cheaper than paying the market price, so the aid was damaging or worth little at best, is to assert in effect that x was very low. There seems to be no empirical warrant for this assertion. That is especially true when we consider that reactions to this expropriation might have extended outside the mixedfarm sector (which removes the limit on r). It may well be, then, that the best Kenya could have done without aid would have been to save some small percentage of the market value of the farms. British aid was then worth its grant equivalent—some 90 per cent of nominal value—minus this small percentage of land purchase costs. If the Kenya government had got its policy wrong and tried paying, say, half the market price, the operation might have ended up costing more than paying the full market price (a situation that arises when the expenditure functions in Figure 2 have points above r=1, and the e chosen results in one of these points being reached), so British aid also prevented the possibility of expensive mistakes.

106 Aid and Inequality in Kenya: British Development Assistance to Kenya This model of expropriation is deficient in at least two ways. We have not treated the problem of the speed of take-over at all thoroughly. And we have assumed in effect that the Kenya government had to select e at the start of the programme and maintain it throughout the programme. In fact the best strategy may be to change e over time, paying different percentages of the market price in different years. Specifically, as fewer expatriate farmers remain to cause trouble and as more Kenyans acquire experience, the optimal degree of expropriation may rise. However, two considerations limit the importance of this. One is that if e is changing, what will determine reactions is not the present value of e set in any year, but farmers’ expectations about its future course. The farmers would have spotted any tendency for e to rise and perhaps extrapolated on that trend in forming their expectations. Raising e steadily only works if the farmers are very stupid. Secondly, if the expropriation affects confidence outside the mixed-farm sector and expatriates remain important in other sectors, raising e over time may lead to greater losses elsewhere in the economy than it saves on the farms.

Notes 1. See Chapter One, p. 15. 2. C.P.R.Nottidge and J.R.Goldsack, The Million-Acre Settlement Scheme 1962–1966, Nairobi, Department of Settlement, 1966, p. 1. 3. G.Wasserman, ‘The Independence Bargain: Kenya Europeans and the Land Issue 1960–1962’, Journal of Commonwealth Political Studies, Vol. XI, No. 2, July 1973. The party initials stand for the Kenya African Democratic Union and the Kenya African National Union. 4. Nottidge and Goldsack, loc. cit. 5. Ibid. 6. Report of the Mission on Land Settlement in Kenya (Van Arkadie Report), Nairobi, December 1966, mimeo, Chapter 2, para. 32. 7. Warhuiu Itote, Mau Mau General, Nairobi, 1967, pp. 257–60. Also Van Arkadie, op. cit., Chapter 13, para. 543. Kenya 1962 (HMSO), p. 130, relates: ‘kikuyu subversive activity came to the fore with the discovery of the activities of the Kenya Land Freedom Army.’ 8. J.W.Harbeson, ‘Land Resettlement and the Politics of Rural Development’, Institute for Development Studies, Nairobi, IDS Discussion Paper No. 28, 9. Wasserman, op. cit., pp. 104, 105 and 116. 10. Nottidge and Goldsack, op. cit., p. 9. 11. Department of Settlement, Annual Report 1962/63, p. 7. 12. Wasserman, op. cit., pp. 106 and 111. 13. Van Arkadie, op. cit., Chapter 3, paras. 91–109. 14. Ibid., Chapter 6, para. 182. 15. Department of Settlement, op. cit. 16. Ibid., p. 2. 17. C.Gertzel, The Politics of Independent Kenya, London, Heinemann, 1970, p.45. 18. Department of Settlement, op. cit., p. 3. 19. See ODA, Aid Policy in One Country: Britain’s Aid to Kenya 1964–68, 1970. 20. Ibid., pp. 4–5. 21. Gertzel, op. cit., pp. 73, 84–85. 22. The source of this and other information in this section is interviews with British and Kenyan civil servants.

Aid in Action

107

23. Statistics Division, Kenya Ministry of Finance and Economic Planning, An Economic Appraisal of the Settlement Schemes, 1964/65–1967/68. Farm Economic Survey Report No. 27, Nairobi, 1971. 24. M.F.Scott, J.D.MacArthur and D.M.G.Newbery, Project Appraisal in Practice: The Little/ Mirrlees Method Applied in Kenya. Forthcoming. 25. International Labour Office, Employment, Incomes and Equality: A Strategy for Increasing Productive Employment in Kenya, Geneva, 1972, Chapter 10, pp. 165–72. These suggestions of the report were not well received by some agriculturalists who thought they underemphasised that land is not homogeneous and that many less fertile holdings would be non-viable if subdivided. 26. Seep. 116. 27. Development Plan 1974–1978, Nairobi, Government Printer, p. 229. 28. Agriculture Act, No. 8, 1955, section 187. 29. New Mery Sisal Estates v. Attorney General, 1972. 30. Department of Settlement, Annual Report 1972. 31. Scott, MacArthur and Newbery, op. cit. 32. Department of Settlement, Annual Report. 33. See Chapter Two, which refers to the need for a rehabilitation programme financed by the World Bank. 34. Figures supplied by the Departments concerned. 35. Van Arkadie, op. cit., Chapter 16, para. 663. 36. Over half the employees in Kenya potentially liable were not assessed for income tax and nearly half the liability to graduated personal tax up to 1971 was evaded, The situation is worse in respect of the self-employed. ILO, op cit., Chapter 16, p. 272. 37. Figures taken or calculated from Annual Report of the Department of Settlement, various issues, 1962–72. 38. Van Arkadie, op. cit., Appendix G, Table V. 39. Development Plan 1974–78, Chapter 10, para. 116, p. 228. 40. Van Arkadie, op. cit., Chapter 9, para. 313. 41. The degree of expropriation expressed as a percentage is defined as: (M—P) 100/M. Where M=market price and P=price actually paid. Note that the market price has no ‘economic’ significance; it is simply the price that the farmers’ experience makes them consider ‘just’ and at which they will maintain their farms. 42. The latter sum presents no difficulties, being the sum paid out for the land minus the grant element of British finance. 43. Information supplied by the Department of Lands. 44. The Nation, 18 September 1974, and as told to one of the authors on a visit. 45. But see the Appendix to this section, pp. 137–4. 46. Van Arkadie, op. cit., Chapter 2, para. 113. 47. In 1961 all mixed farms were in European ownership. Economic Survey 1968 (data relate to 1967), p. 55: Africans own 59 per cent of the mixed-farm areas in the ratio 14:12 large farms to settlement plots. Whites therefore still owned 41 per cent of the mixed-farm sector, and 56 per cent of the existing large mixed farms in 1967. Development Plan 1970–74, p. 200. Two-thirds of the mixedfarm sector had been transferred, half-and-half large farm and settlement, in mid-1968. 48. Yields per acre for white-owned farms in Trans Nzoia in 1959/60 were similar for maize to yields on African farms in the same area ten years later, despite the introduction of higher yielding varieties. Profits of white farms in the earlier year which were in the size range 800–999 acres (the average size of farm in the sample of African farms ten years later) were about 42sh an acre, double the figure for African farms ten years later, at current prices. Output per acre was similar, but white costs per acre were only 2/3 of those on the later African farms. See Economic Survey

108 Aid and Inequality in Kenya: British Development Assistance to Kenya of African-owned Large Farms in Trans Nzoia 1967/68–1970/71, Farm Economic Survey Report No. 28, November 1972, Table 3.15. A Report on an Economic Survey of Farming in the Trans Nzoia Area 1959/60; Farm Economic Survey Report No. 5, April 1961, Table IV. 49. White farms were assumed to produce 50 per cent more output per acre than black farms. Proportion of output from white farms was: Oe/Ot= 1/(1+2la/3le) where Oe=output of European mixed farms, Ot=total large mixed-farm output, la=percentage of mixed-farm acreage under African control and le=percentage of mixed-farm acreage under European control. 50. Accounting ratios were supplied by M.F.Scott, who calculated them for Scott, MacArthur and Newbery, op. cit., as: Wheat Maize Dairy cattle Milk and butterfat All others Pyrethrum

All years 0.84

1963

1964

1965

1966

1967

1.23

1.69

1.33

0.84

0.58

Succeeding years 0.87

1.41

1.32

1.15

1.07

1.08

0.00

1.30 0.84 1.00

We assumed the accounting ratio for pyrethrum to be zero after 1968 on the grounds that the Pyrethrum Board allocates quotas and after that date all quotas allocated to large farms could have been filled by small holders. The share of large farms in pyrethrum production was down to some 30 per cent by 1966 compared with two-thirds in 1963. 51. Central Bureau of Statistics: Input-Output Table for Kenya 1967, Nairobi, 1972, Table 1, column 3; value added is 0.80 of gross output The ratio of output of mixed farms at accounting prices, to that at market prices, is 0.9. Using a standard conversion factor of 0.80 for the accounting price of farm inputs, and assuming that the mixed farm sector is typical of the agricultural sector as a whole, gives a ratio of value added to gross output, both at accounting prices, of 0.82. 52. M.J.Westlake, ‘Tax Evasion, Tax Incidence and the Distribution of Income in Kenya’, IDS, Nairobi, 1971, mimeo. 53. Price Index; 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 111.9 119.0 123.9 136.0 100 101.0 102.2 105.9 107.4 108.1 109.3 Source: Abstract of Statistics 1967, Table 179, p. 157; Abstract of Statistics 1973, Table 247, p. 258. 54. On plausible assumptions about the pattern of farmers’ reactions it is possible to draw firmly such a conclusion about the optimal expropriation policy. See Appendix to this section. 55. See Chapter Three, p. 55. 56. This we take to be one of the themes of the ILO report, op. cit. 57. ILO, op. cit., Chapter 6, pp. 87–8. 58. Leys, op. cit., p. 225. While Leys regards the Land Transfer Programme as an important instrument in smoothing the transition of the colonial economy to a nominally independent one favourable to British capitalist interests, he also brings out the political weakness of the ‘radicals’ in Kenya, before and after independence.

Aid in Action

109

Part Two: The Mumias Sugar Company The setting up of the Mumias Sugar Company (MSC) is the biggest ODM-aided project Kenya has undertaken since independence, excluding the Land Transfer Programme. The British aid contribution was a soft loan of £2.9m. in 1971 towards a total investment of some £7.5m. The Kenya government is now considering further cane-growing and processing developments in Western Kenya so the lessons of the project have an immediate as well as a general importance. We give a summary of the main features of the project and an account of its institution, followed by a description and evaluation of the outgrowers scheme, whereby much of the cane processed in the Mumias factory is produced on small-holder plots. There is a discussion of the lessons of the project for the question of appropriate technology, costbenefit evaluation techniques, and the practice of aiding commercial ventures. The Mumias sugar scheme consists of a factory processing cane, a nucleus estate of some 8,000 acres where roughly half of the factory’s cane requirements are grown and the outgrowers scheme. The factory and nucleus estate employ directly over 2,500 workers. Operations began in July 1973 and in 1974 55,000 tonnes of mill white sugar were produced. Production is planned to increase to 75,000 tonnes over the next two years. The outgrowers scheme is scheduled to grow and provide all the additional cane required. There are currently some 3,500 outgrowers with about 14,000 acres under cane, in an area within an eight-mile radius of the factory. Eventually there will be 5,000 outgrowers with 20,000 acres under cane over a wider area. There are also about 1,000 people on a canecutters register, of whom about 600 are employed cutting outgrowers’ cane at any time. Many of these workers are, however, outgrowers or their dependants doing extra work for cash wages. Of the project’s total capital cost of £7.5m., the factory itself cost £4.5m. The British government’s loan extended at 2 per cent interest included £2.5m. ‘for the factory’ which was supplied by a British firm, Fletcher and Stewart, a Booker McConnell subsidiary. The money was used by the Kenya government to acquire equity in the Mumias Sugar Company, which was set up to own the scheme. The Kenya government holds 69 per cent of the equity. Other equity holders are: Commonwealth Development Corporation 12 per cent, Kenya Commercial Bank 9 per cent, East African Development Bank 5 per cent and Booker McConnell Ltd. 5 per cent. Bookers Agricultural and Technical Services have a management contract to operate the scheme.

Sugar in the Kenya economy The scheme has importance for the Kenyan economy as a whole. Kenya’s total production of sugar in 1973, when Mumias produced only some 20,000 tonnes, was 133,000 tonnes, or just over half of consumption requirements. Kenya has long been an importer of sugar. The setting up of two new sugar factories in Nyanza in 1967 was followed by a once-andfor-all drop in the volume of imports. However those factories have experienced teething troubles and the demand for sugar has increased more strongly than expected with rising national income. Table 16 shows how imports have risen after a fall.

110 Aid and Inequality in Kenya: British Development Assistance to Kenya Table 16: Imports of sugar 1968

1969

24147

8540

624

324

1970 tonnes 21223 K£’000 935

1971

1972

71747

103816

3768

7053

Source: Kenya Government, Statistical Abstract 1973.

The rising price of sugar on world markets in recent years has meant that imports have become an increasing foreign exchange burden. In 1972 sugar imports were over half by value of all food imports (K£12,836) and almost 4 per cent of total imports. Consumption is expected to reach 285,000 tonnes by 1978 and with Mumias and the older factories at full stretch that would cut the import requirement to some 50,000 tonnes.1 Thereafter rising demand would again increase imports, hence the plans for further sugar development. This demand is partly the result of the Kenya government’s own policy. It fixes the price which it pays to producers for sugar and adds an excise duty and a margin to allow for distribution costs via the Kenya National Trading Company. The price of sugar in Kenya has been considerably below the world market price during the upheavals on the world market over the past couple of years; it was briefly only a tenth of the world price. The difference between the import price and the consumer price is met from a sugar price stabilisation fund. Sugar is a very important consumer good in Kenya. Variations in its price, colour and texture arouse strong feelings among the Kenyan public. Because of this the government has maintained a price that does not always reflect the cost of the commodity to the economy, and is planning to meet a domestic demand predicated on that price. Sugar makes little direct contribution to the diet although it is argued that it makes more nutritious foods palatable and increases their consumption. Subsidisation is based, however, on its political importance rather than its dietary attributes.

History of the Project The original idea for a sugar industry in Western Kenya came from the Swynnerton Plan in 1954. The modern Mumias development was the brainchild of the then Minister of Agriculture, B.R.McKenzie in 1965. The area north of Lake Victoria in the Western Province of Kenya was selected because it seemed climatically reasonable and although quite densely populated, with 222 people per square mile, it was very underdeveloped. The Abaluhya population depended on subsistence farming, and a number of state initiatives to introduce cash crops, including coffee and sisal, had all failed for a variety of reasons. Some cane was, however, being grown in the area for jaggery production. After a preliminary study Bookers Agricultural and Technical Services was approached.2 It declined to begin operations immediately but insisted on a pilot scheme to test the viability of commercial sugar production in the area. Eventually it was agreed to have a feasibility study run concurrently with a preliminary stage of the main project so that if the pilot project were successful there could be a rapid transition to commercial production. This pilot project was largely financed by the Kenya government with £450,000, and Bookers

Aid in Action

111

put up £50,000 to be repaid on presentation of the feasibility reports. The pilot project began in 1967 and a final feasibility report and social cost-benefit analysis appeared in 1970. They concluded that the project was a good one and the Kenya Cabinet gave the goahead at the end of that year. There was some dispute over the financing of the project. The Kenya government wanted Bookers to put up a substantial equity stake while the company wanted simply a contract to supply the factory via their subsidiary Fletcher and Stewart and a management contract for their Technical Services Division to run it. HMG was keen to provide aid for the project which was an attractive one from the donor viewpoint; it was large and highly visible, the commercial involvement meant it was likely to be well run; and as it included a factory it had a high import content. Far from the Kenyans going cap in hand to HMG, therefore, they could afford to haggle. An attempt was made to obtain British funds for Mumias outside the agreed aid framework figure. Some reservations were felt about Mumias within ODM because of its effect on East African economic co-operation, At that time Kenya imported its sugar requirements from Uganda, which had some advantages as a sugar-producer. At the price paid for Uganda sugar, the Mumias project did not have a very high projected rate of return (see below), and there seemed to be a case for leaving sugar production with the sister-state in the East African Community. This view did not prevail because it was clear that Kenya intended to go ahead with sugar production in any case. An attempt at leverage would have failed and would furthermore have been fiercely resented by the Kenyans. Finally there was the consideration that if Britain did not support the project some other donor would, and perhaps German management and equipment would be used. HMG succeeded in lending money to Kenya for the project within the aid framework, and the Mumias Sugar Company was formed in 1971. Bookers obtained the management contract but was obliged to take a small equity stake—about half what the Kenyans had pressed on them. The British position in these negotiations was strengthened because the Kenyan government had been impressed with Bookers’ handling of the pilot scheme. And the Chemelil factory which had been supplied from West Germany and supported by German as well as British aid funds had rum into technical difficulties. As a quid pro quo which Bookers had to pay for the management contract at Mumias, it took over the management of Chemelil, which had been operating at a loss and running at well below capacity. Equity in the Mumias company amounted to £3.5m. and loans of another £5.2m. were arranged, although only £3.8m. of loan finance was eventually used, as contingency funds built into the estimates were not in fact required.3 The Kenya government had intended to on-lend the 2 per cent British aid loan to the Mumias Sugar Company at 3 per cent and put in about £1¼m. in equity. On ODM advice, however, it put in more equity—some £2.5m. This reduced the Kenya government’s subsidisation of the project in the long run, as 3 per cent was much below the commercial lending rate, and it reduced the gearing and hence cash flow problems of the project in the early years. The Chemelil sugar factory, financed by German aid loans in Deutschmarks, had had a gearing ratio of 3:1, and this was one of the factors which caused it to show a loss in operation.

112 Aid and Inequality in Kenya: British Development Assistance to Kenya

The Project in Prospect and Practice Sugar production in Kenya had not been a success when Bookers began its pilot scheme at Mumias and the company approached the project with considerable care and circumspection. This, by general agreement, has resulted in a sugar complex that operates efficiently and outperforms all others in Kenya. The experience of Bookers in cane production elsewhere in the world was important. The Chemelil sugar factory, supplied from Germany, was based on designs that were insufficiently flexible for local conditions, and the factory suffered repeated breakdowns. The techniques and equipment at Mumias, drawing on experience in other tropical areas, were better adapted to its situation. The Wanga area of Western Province where the factory is situated is also technically better for sugar production than Nyanza. The rainfall is higher and more regular and the soil is more suitable, being more porous and draining more easily than the black cotton soils of Chemelil. These characteristics obviated the need for either irrigation or much drainage and reduced considerably capital costs of the nucleus as well as making for success with outgrowers. While Western Province is in general densely populated, the Wanga area is much less so than surrounding districts. The land on which the factory and nucleus estate stand is rented by the Mumias Sugar Company from the government which had to acquire it from its owners, the area having been previously adjudicated and registered to private ownership. Much of the area for the nucleus estate was taken from lower-lying land in the Nzoia valley which requires extensive drainage and was in consequence little-used. The price which the government had to pay for the land, and the social cost of putting it under sugar, were lower than they would have been elsewhere in Western Province. There was however some difficulty when the land was being bought. The local MP urged small holders to hold out for a higher price, and the personal intervention of President Kenyatta was required at one stage to keep the project on schedule. In carrying out their commercial feasibility study, Bookers worked with 1970 prices including a controlled sugar price of K£45.25 a tonne and controlled cane price to outgrowers of K£2.34 a tonne less transport and other eharges.4 They computed that the expected rate of return over 25 years would be 12 per cent. This incorporated conservative estimates of technical coefficients such as yields of cane per acre and tons of sugar to tons of cane. If a 5 per cent increase in production was assumed, it gave a rate of return of 13.1 per cent. The project was expected to be in deficit from 1970/71 to 1973, with a cumulative deficit of K£5.53m. Allowing for contingencies, therefore, over K£6m. was thought to be required to finance it. These estimates were amended in various ways in calculation of the social costs and benefits. There were additional necessary costs incurred not by Mumias Sugar Company but by the government, of K£1.45m. for land acquisition, road construction, and township housing and finance of outgrowers. Bookers followed a policy of adjusting market prices as little as possible. This was reasonable as shortage of reliable data reduced adjustments to guesses anyway. Yet some prices were so clearly out of line with true social costs that some adjustments were required. Sugar was valued at its c.i.f. import price, not the domestic consumer price. At that time Kenya imported sugar from Uganda at a price of K£40.4 a tonne. This was taken as the value of output. Since it was K£5 a tonne less than the Kenya price the benefits of the project had to be written down by some K£250,000 in an average year.

Aid in Action

113

On the other hand the cost of untrained labour was marked down. The Kenya Union of Sugar Workers applied a minimum wage of 3.85s5 a day although casual agricultural labourers in the district worked for between 1.50s and 2.00s a day. Bookers claimed that ‘some research’ suggested the opportunity cost of an unskilled man-year of employment was between 400s and 450s, less than a third of the wage it was paying, some 1,600s a man-year. To be on the safe side it fixed the shadow wage at 40 per cent of the market rate—some 1.70s a day or 640s a man-year. This adjustment reduced costs by K£96,000 in an average year. No adjustment was made to the wages of skilled workers which were also counted gross of income tax. Some government-incurred costs were included, but road and housing construction were shadow-priced at 90 per cent of market cost to allow for the unskilled labour component and indirect taxes and duties. Compensation for standing crops and houses was included but not the cost of land for the nucleus estate. Nor was rent imputed for outgrowers’ land being put under cane. This was on the grounds that there was no land shortage in the district, over half being uncultivated. Part of the nucleus estate area, being swampy, was particularly little-used, and the assumption was made that the purchase receipts would be used to buy unutilised land elsewhere in the area. Finally the cash surpluses of outgrowers were included among the social benefits of the scheme. These added an annual amount, rising to K£452,000 in 1981, to the scheme’s benefits. Other benefits arising from road construction were not included, and costs and benefits were discounted over 25 years, making the conservative assumption that the project would have no terminal value after 25 years. The internal rate of return was found to be 12.6 per cent. A 5 per cent fluctuation in output gave internal rates of return of 11.2 per cent and 13.3 per cent for declines and increases respectively. Obviously many of the estimates used in this cost-benefit analysis could be varied. As noted, a lot of carefully conservative estimates were made—of technical coefficients, of the reduction appropriate for the shadow wage rate; the extraneous benefits of road and housing construction were not counted. There is, however, one outstanding instance of an omission in the other direction. This occurred in the treatment of the outgrowers’ surpluses. The social benefit was taken to be the outgrowers’ cash surplus, making no allowance for the cost of their labour or that of their families. It may be that working on sugar, especially after initial planting and weeding have been carried out and the cane is big enough to look after itself, does not conflict much with growing other crops. Yet people’s time usually has some value to them, and hard work has some disutility. The fact that small-holder land is so little utilised, which was used to justify its negligible shadow price, must suggest that labour might well be a scarce factor of production. If so, the labour of the small holder and his family cannot be ignored. Adequate input-output data are lacking to assess this cost rigorously. However, Bookers calculated that the return small holders could expect was between 152s and 136s an acre for a fiveyear cycle, taking no account of the farmer’s labour. It added that a farmer making 152s an acre for a five-year cycle, if he used hired labour, would realise some 28s an acre less, i.e. the cost of hired labour would reduce his surplus by about a fifth.6 There seems no reason why the farmer’s own labour should be less valuable than that of a casual labourer he hires. So we adjusted Bookers’ own calculation by deflating

114 Aid and Inequality in Kenya: British Development Assistance to Kenya outgrowers’ surpluses by a fifth to allow for the value of outgrowers’ labour. This significantly reduces the rate of return which could have been expected from Mumias—it then becomes some 8.5 percent. As a development project, as opposed to a commercial venture, therefore, Mumias ex ante did not appear outstanding.7 Given Bookers’ generally conservative treatment of costs and benefits, it was still adequate for the Kenya government, which, after all, hoped to get British loans at 2 per cent interest. Indeed as the carrying out of the project led to Britain extending soft loans to Kenya that might otherwise have remained undisbursed for years, given the current aid-spending problems, the grant equivalent of the British loan should be counted as a benefit of the Mumias project from the Kenya government point of view. From the British point of view, however, it might seem less clear why ODM was so keen. In fact given the conservative assumptions made, the rate of return is up to the average standard for aid projects. The point is that where there is a shortage of projects, donors spend money on what they can. And if the project includes a large factory dominating the landscape and entirely imported from the donor country, so much the better. Whether the project was backed for sound reasons, however, now appears a somewhat cold speculation for in fact subsequent events have vindicated it absolutely. The project turned out to be profitable even before the rise in world sugar prices. Then came the collapse of Uganda as a sugar exporter after the political upheavals in that country, followed by the boom in the world sugar price already noted. Under bilateral deals sugar was imported into Kenya at about K£140 a tonne and the world free-market price soared to K£600 a tonne before subsiding to its level at the time of writing of some K£135 (£165). The Mumias project was very well run and technically its performance exceeded projections. Production in 1973 and 1974 was 163 per cent and 145 per cent respectively of the expected figure. And in 1975 it is expected to be 121 per cent of the original projection. This is owing both to better yields of cane per acre by both nucleus estate and outgrowers than expected and to the greater efficiency of the factory. The ratio of tons cane to tons sugar, for example, is commonly about 10 in a modern sugar factory, but at Mumias in 1973 it was 9.3 and in 1974 8.79. For 1975 it is now expected to be 9.2. Costs have also been lower than expected. For example, the actual cost of harvesting and transporting cane in 1973 and 1974 were 74 per cent and 96 per cent respectively of the projected figures, despite inflation and the greater output. The factory has also been economical in its use of fuel, running for most of the time on combustion of bagasse (the waste vegetable matter from the cane) and using no oil at all. The rise in the c.i.f. import price of sugar has boosted the internal rate of return (IRR) of Mumias considerably. If output is valued at the year’s average world market price in New York, deflated to 1970 prices,8 it has been: 1973 K£1.17m.; 1974 K£11.57m., 1975 K£4.30m.9 The project had paid for itself, therefore, at the end of 1974. However low the sugar price falls in future, the project will have been justified even if it has to be closed down. It is possible, if doubtful, that Kenya could have continued to import sugar more cheaply under trading agreements than the price being charged in the residual free market. Against that, however, is the fact that Kenya since the oil price increase has run into a severe balance of payments problem. The official exchange rate does not reflect the true scarcity value of foreign exchange. Computing a shadow foreign exchange price goes outside the scope of this study, but if, for example, it were 1.2 of the current par value, a

Aid in Action

115

perfectly possible situation, then the value of Mumias output to Kenya in 1974 could have been as high as K£14m. at 1970 prices. Clearly the rate of return on the Mumias project depends on a number of factors that are wide open to different assumptions—what would sugar have cost Kenya over the past three years, what will happen to the world market price over the next 25 years, what is the true or shadow exchange rate? On any reasonable mix of assumptions, however, the Mumias IRR will be very high. For example, suppose that Kenya could have obtained sugar during 1974 and 1975 at the price ruling in June 1975—some K£135—and this price is on average maintained during the life of the project; Kenya remains short of foreign exchange for ten years with a ratio of shadow to actual foreign exchange price of 1.2. These assumptions would give Mumias an internal rate of return of over 75 per cent. Viewed technically and from the point of view of the Kenya balance of payments, therefore, Mumias has been an unqualified success. It has not, however, been immune from criticisms. These have concerned its impact on the development of the Wanga area, and whether the technology in use is appropriate to Western Kenya. The local developmental effects of Mumias will largely come from the outgrowers scheme which involves twice as many heads of households as are employed in the factory or on the nucleus estate, so as a prelude to consideration of Mumias as a development project and the criticisms made of it, we give an account of the outgrowers scheme.

The Outgrowers The 14,000 acres or so now under outgrowers’ cane is made up of individual sugar holdings with an average size of 3.9 acres. This average hides some variation. Some farmers have as little as one acre and one is known to have 58 acres, despite a company policy to try to restrict the size of individual holdings. The average small-holding size in Wanga is some 10 acres, but it is considered socially undesirable for a farmer to have more than half his land under cane. The intention is to leave farmers with enough land to continue to grow their subsistence requirements. The individual holdings are grouped into blocks of at least 15 acres so that cultivation and transport can be viable. This means that a man cannot easily become a registered outgrower unless his neighbours wish to be so too, as contiguous plots have to be pooled to form a block. The company surveys each potential farmer’s plot to determine its suitability and the farmer is charged with clearing an agreed area. It is ploughed, harrowed and furrowed mechanically by the company cultivation unit. The company then provides appropriate disease-free seed cane and fertiliser, and the farmer himself is responsible for his own planting, fertilising and weeding. The company does, however, employ casual labourers and if a farmer cannot cope with his own weeding a gang can be sent along. Only about 10 per cent of outgrower weeding is handled in this way; most farmers make their own arrangements. Harvesting is undertaken by the company. The 1,000 people on the canecutter register work around the outgrowing areas. Cutting and stacking is done by hand and the cutters are paid by the task—some 5s a tonne in 1974. The idea is that people should turn up from the area where cutting is taking place, with outgrowers taking precedence, and about 600 cutters will be working at any time. There is considerable turnover of workers and no excess supply, so that someone could get work every day if he turned up. Once

116 Aid and Inequality in Kenya: British Development Assistance to Kenya stacked the cane is loaded and transported mechanically. All the outgrowers’ operations are checked by the company. There are three superintendents and a fourth responsible purely for harvesting. These posts have been Africanised. Under the superintendents there are 17 supervisors who have certificates in agriculture, and below them some 48 headmen who have basic literacy and have been given on-the-job training. So the company operates a lively extension service with regular inspection of outgrowers’ plots providing advice and guidance. All these goods and services are provided on credit to the farmer and the cost of plot development, planting material, fertiliser, weeding if necessary, harvesting, transport and maintenance of access roads is eventually docked from the money he gets for his cane. The company, at the Kenya government’s request, charges the farmers interest of 8 per cent per crop for these services. At present each farmer has a contract with the company which specifies the conditions on which he becomes an outgrower and the costs of services. The contract obliges the Mumias Sugar Company to buy all the cane grown according to contract and to provide the specified services. The farmer is obliged to follow approved cultivation practices, to plant at the time specified by the company, to grow approved varieties of cane, to maintain standards of husbandry, and to sell his cane to the company. At the end of 1974 the price of cane was 62s a tonne, transport and harvesting costs were 22s a ton. After other costs had been deducted the net payment was in the range of 15s to 22s a tonne. A yield of 55 tonnes per acre, which was average for the plant crop,10 would give an income of over 950s an acre for the 22 months of the crop. The yield achieved varied from 28 to 90 tonnes an acre and the income varied accordingly—a few outgrowers were receiving over 2,000s net an acre a crop. Both MSC and the Kenya government are anxious to regularise the outgrowers’ position and provide a corporate body that could negotiate terms with the company and deal with it on behalf of the farmers. The outgrowers themselves were hostile to a co-operative and the intention has been for a long time to form an outgrowers’ company, which would take over inter alia the financing of services for outgrowers from the Mumias Sugar Company. At the time of writing, however, the problems of setting up this organisation had not been solved, although CDC had agreed to provide finance.

Mumias and Rural Development The picture that emerges from the above account is of a group of small holders whose activities are very closely regulated. In effect they rent a bit of their land and some of their labour time to the company for an assured payment. As a company employee put it: ‘They have an assured market, no risk and no decisions to take. We take all the decisions.’ Most of the farmers in Wanga regard this as an enviable position to be in, and the competition to be outgrowers is strong. In the eight-mile radius from the factory, only one small holder in five is an outgrower and most of the others would like to be. In one area the company was proposing to plough about 150 acres of land for sugar. So many farmers had come into the field office offering plots averaging four of five acres, however, that there were over 400 acres on offer and registered as available. Some farmers clear their land in advance of agreement, hoping to induce the company to plough. It is too early yet to assess the developmental impact on the neighbourhood because production only started in 1973, and no farmer has yet received payment for more than

Aid in Action

117

one crop. In principle the scheme should boost agricultural output in the area generally. Financial constraints are being removed; after 22 months farmers are receiving lump sum payments, often of more money than they have seen in their lives. The agricultural extension service provided by Bookers for cane has also influenced agricultural techniques generally. In particular, farmers are aware now of the gains from planting in rows and using fertiliser. In fact the sugar company finds that some small holders have been embezzling fertiliser issues—half-dosing the cane and putting the rest on their maize crop. A black market in fertiliser has appeared following its rise in price, with people coming from other neighbourhoods to try to buy the small holders’ fertiliser issues. The wage employment in the area should also provide something of a local market for food crops. There are a few signs, however, that development might not proceed so smoothly. Most of the first cash from sugar was certainly not invested in agricultural inputs. Tony Barclay, an American sociologist resident in the area, has found that, on a sample of farms he studied, off-farm investment was preferred.11 Sugar revenues were spent on children’s education or acquiring a small shop as well as on durable consumption goods like wives12 and tin roofs. There seem to be two reasons, one general to the Wanga district and one that may be peculiar to the subdistrict he studied. The general reason is a local labour shortage in smallscale farming. The Abaluhya social structure is not conducive to certain kinds of co-operative effort. People regard it as demeaning to work for an outgrower if he is doing well financially but does not enjoy a higher traditional status than his fellows. People would rather work for the company, and if they are employed on a neighbour’s plot are not above a little sabotage, for example putting dirt over weeds instead of uprooting them. Faced with a labour constraint therefore, farmers concentrate their labour (or that of their wives) on sugar and perhaps spend less time on the subsistence crops they previously grew. The less general reason is that the quality of soil is very low in that part of Wanga where Barclay was resident. Careful husbandry yields only 10 bags an acre of maize, and 2 bags an acre is not uncommon. Farmers do not regard it as a worthwhile investment to plough their earnings back into such soil. This observation is confirmed by Mumias Sugar Company employees. They notice that the farmers work much harder at their husbandry where the prospects of high yields are better. Apart from the issue of whether the benefits from sugar will help to generate much local development (and we repeat that judgement is premature), there is the vexed question of the distribution of the benefits. After disastrous experiences with government-sponsored cash crops like sisal (the market collapsed), cotton (the area was too wet and pests proliferated), and coffee,13 farmers were sceptical about sugar. The first to become outgrowers were therefore those who agreed to try it, either the most enterprising or the møst desperate. Now things have changed; everyone wants to be an outgrower and the company can afford to pick and choose, usually on the basis of soil suitability and ease of access. Farmers whose plots happen to enjoy these advantages can become outgrowers and receive a relatively large cash income without the exercise of any enterprise at all. This has led to enormous pressures on the employees of MSC, especially the African staff. Kinship ties are called on, bribes are offered. Relatively well-to-do people from other areas, sometimes civil servants, attempt to acquire land in the district and then put pressure on to get the land ploughed. Even though MSC employees act with considerable probity it

118 Aid and Inequality in Kenya: British Development Assistance to Kenya is difficult to get this widely believed in the district.14 Refusal to plough someone’s ground because, for example, it is too stony, is usually rationalised by the disappointed farmer as owing to nepotism, tribalism and bribery. Furthermore a process of cumulative causation may be under way. When outgrowers receive their money, although they have not used it much on their own plots so far, one of the assets they try to acquire is more land. This is because of the emotional significance of land but also because they hope to get the new plot ploughed too. The company has no hard-and-fast upper limit although it is not supposed to take more land off a single person than he can cultivate adequately. But of course it is very difficult to establish when it is dealing with a front man, either for an absentee owner or for someone who is already an outgrower (or in extreme cases both). Some indication of these trends is given by the rising price of land in the Wanga area. Variations in land prices are considerable, but broadly an acre of land changed hands in 1969 for 400–500$. In 1973 it cost 600–700s. Now over 1000s an acre is not unusual. It is possible for people in Britain to be over-sensitive about inequality in a poor country. Almpst by definition, development entails making someone better off than he was before and therefore better off than some other people still in the boat he was in. You have to start somewhere. However, there are obvious disadvantages to the situation in Mumias just outlined. People may be getting better off not because of any quality that means they are more likely to make productive use of their opportunity than their neighbours, but because of the accident that their plot has convenient characteristics for the Mumias Sugar Company. Or worse, because they can wield some other sort of influence. There is little recourse for the enterprising man without these advantages, although a rising demand for food crops owing to higher cash incomes should help him somewhat. Because the Mumias factory turned out to have more capacity than expected some non-contract cane has been bought from freelance growers. About 13,000 tons has been bought in this way—about 4 per cent of the factory’s requirements. However, this is being phased out. From this year on, the company has said it will buy no more cane from non-contract farmers, because planning of outgrowers’ output now takes the factory capacity fully into account. We are not idolators of the market mechanism who believe that it is bound to allocate resources to best advantage. In this situation, however, residual market purchases of cane would have one obvious advantage. They would give everyone some sort of chance to make a living from sugar even if non-contract growers were at a disadvantage. From that point of view the ending of non-contract cane purchases is a pity. It is clear that no blame at all attaches to Bookers for these difficulties. Having sugar produced by outgrowers was itself a considerable experiment. If the factory was not to suffer cane shortages leading to the kind of low capacity working that has blighted other plants in Kenya, it made perfect sense from a management point of view to take what steps were necessary to see the cane supply was smooth and ample. Indeed that was the first priority. However, we believe that the patterns coming to light should be taken into account in future sugar developments. A higher priority should be given at an earlier stage to setting up an outgrowers’ organisation. This could perhaps make supporting services available more widely if less intensively than at present, leaving something to farmers’ initiative. To compensate perhaps for the greater uncertainty about production a measure of overplanting could be encouraged. This might entail some subsidisation or at least easy financing for the outgrowers’ company, especially in the early years. This would be an

Aid in Action

119

appropriate place for government support, perhaps with foreign aid. The organisational difficulties are formidable. Any arrangements must include provision for control of cane varieties and disease prevention. And it would be wrong to penalise farmers by refusing to buy their output in the event of overproduction, Perhaps the cane price should be set in advance each season on the basis of experience of supply response. This would require relaxation by the Kenya government of its strict price control of cane. At any rate the present arrangements could prevent sugar boosting development as much as it might by concentrating land and money in the hands of a proportion of the local population who are not particularly deserving, and also in the hands of absentee landholders. It has led, and will lead, to an exacerbation of social pressures and enmities in an area already noted for its somewhat fractious politics and it places a heavy burden on company employees in charge of outgrowing who have to tread through this political minefield. However, any expansion of MSC should to some extent automatically mitigate these difficulties by increasing the proportion of local farmers who can participate in sugar production. And by increasing the amount of land under cane in the area it should raise food prices to the advantage of those still growing food crops. The Mumias Sugar Company and Bookers are of course aware that their operations are likely to have profound effects on the district and they have offered between them to pay three-quarters of the cost of a sociological study of the area but at the time of writing finance for the remainder had not been forthcoming. That is unfortunate. Such a study would be an appropriate use of aid funds. A study of the situation in the neighbourhood before the effects of Mumias really appear is particularly urgent as it will provide the base line for later assessment of those effects. The developments in the Mumias area will have implications much beyond the case of sugar. Many of the supposed obstacles to rural development—shortages of liquidity for investment, of knowledge of correct agricultural techniques, of access roads, and of a market for cash crops—are all being removed for many farmers. The money and the roads and the knowledge, although arising from sugar, are there to be applied for other crops. What happens in the next few years must be instructive about rural development, probably more so than developments in the Special Rural Development Areas which have attracted the attention of aid donors in Kenya. It is important therefore that events in the Mumias area should be monitored.

Mumias and ‘Appropriate Technology’ The traditional alternatives to sugar in Kenya were honey and jaggery, a dark substance which is basically sugar-cane juice, partially dehydrated. There are many small jaggery factories in Kenya but they are not encouraged by the government because jaggery is frequently used in making changa’a, an illegal spirit. As a simple sweetener jaggery is not popular, and a small survey has shown it is used for this purpose by only 2 per cent of Kenyans, while 98 per cent use sugar.15 Although it contains trace elements removed from refined sugar, jaggery is said to have strong laxative properties so it does not today constitute much of an alternative to sugar. It has been argued, however,16 that palatable sugar can be produced in small factories on the scale of existing jaggery factories and with only limited refinements to jaggery techniques. The sugar, which is off-white in colour and more or less crystalline, is produced

120 Aid and Inequality in Kenya: British Development Assistance to Kenya by open-pan sulphitation techniques and requires additional power and settling tanks to a jaggery mill. It is widely used in India where nearly 1,000 small plants have been set up in recent years. The Kenya government has asked IBRD to review the sugar industry in Kenya, and part of this review—an assessment of the feasibility of cottage sugar plants making khandsari17 sugar—has been hived off UNIDO and FAO who are investigating the possibility of a trial conversion of a jaggery factory. The advantages claimed for cottage sugar plants are that they are more labour-intensive and require less capital per unit of output, and are hence more suitable for a poor country like Kenya, abundant in underemployed labour; that they are small-scale so that, as well as wages accounting for a higher proportion of costs, profits will be less concentrated and more equal income distribution will result; that by scattering sugar production instead of concentrating it in one area they will even up development in rural areas and reduce cane and sugar transportation and distribution costs. We are not competent to assess fully all these claims, but the evidence suggests there is a prima facie case. Using figures supplied by a Ghanaian consultancy firm proposing the installation of cottage sugar plants in that country,18 it appears that a khandsari sugar plant in Kenya would cost about K£90,000 to set up with a cane-crushing capacity of 60 tonnes a day (Mumias crushes some 2,000 tonnes daily). The output of the factory in a ten-month season, working two shifts (16 hours) a day, would be just under 1,500 tonnes of sugar (Mumias: 50,000 tonnes from three shifts). The factory would employ about 143 people (of whom 23 would be supervisory), skilled or semiskilled. The figures are based on Indian experience. Taking the cost of the Mumias factory complex proper as K£5m., it appears that the capital investment required for a tonne of output per year there is K£100 and for the khandsari plant K£60. These figures assume depreciation is the same on both plants. Taking the employment in the Mumias factory complex at 2,500 (an overestimate as some of these people are employed on the nucleus estate), the capital required per job created is: Mumias K£2,000; khandsari plant K£630. In other words for the capital investment represented by Mumias it seems you could have more than 50 khandsari factories which between them would produce over half as much sugar again as Mumias and employ nearly three times as many people. This capital-saving excludes the saving to the Kenya government of not providing housing for large clusters of workers. In addition, the khandsari plants are not more prodigal in their use of skilled workers, shortage of whom is in fact a much greater constraint on output in Kenya than capital. The same output as Mumias (requiring say 33 khandsari plants) would require 759 skilled or semi-skilled workers. Mumias employs 175 management personnel, 275 skilled and 450 semi-skilled people—900 altogether. In some other ways the khandsari plants would be less efficient. The number of tonnes of cane used to tonnes of sugar produced would be in the region 13–14, compared with Mumias’ 8–9. For the same output of sugar, therefore, assuming similar cane-yields per acre (and these are likely to fall without the close supervision arranged at Mumias), Kenya would need to put some 50 per cent more land under cane—perhaps an additional 10,000 acres, for the production of another Mumias. This is a considerable drawback in parts of Western Kenya where population densities are high on good arable land and population growth is rapid, so that land for food production could become scarce in the near future. The khandsari plants would also be less economical in their use of fuel than Mumias, as

Aid in Action

121

they would not operate 24 hours a day and would use fuel to restart production even if running afterwards on combustion of bagasse. There are other considerations: it would be more costly to collect and levy an excise duty on the produce of a large number of smaller factories compared to doing it with a handful of large ones. The taxing difficulty, plus the fact that Mumias is largely governmentowned, might take some of the substance out of the distributional argument for small-scale factories, though clearly not all. The present technology does have favourable implications for public finance, as well as unfavourable ones for concentration of economic power. In addition the types of skills required in the two sorts of factory are not entirely similar. Some of the workers in a khandsari sugar plant, in particular sugar boilers, would require ‘craft’ skills requiring ‘touch’ and not easily systemised. These skills are less generally usable than those of the mechanic and accountant, and in fact are specific to sugar production, while many of those acquired at Mumias are common to much modern industry. Whereas training at both sorts of factory should yield some external benefits to the economy as a whole by adding to ‘human capital’ these would be less therefore in the case of khandsari plants.19 The sugar produced by khandsari plants would be less white than that from a larger factory and while marginally more nutritious would probably meet with consumer resistance. To maintain consumer welfare it would have to retail at a lower price. In India the market price has been 10–25 per cent below white milled sugar. However, by introducing more flexibility into production, as running at below full capacity would be less expensive in the case of the less capital-intensive techniques, khandsari plants would obviate the need for such close regulation of outgrowers, and could be supplied perhaps by local farmers individually or in co-operatives. So there may be a case for establishing a number of cottage sugar plants near large-scale factories to cope with variations in cane supply that might result from a desirable reduction in the control of outgrowers’ activities. In conclusion, there are clearly a large number of factors affecting the relative social profitability of large-scale and khandsari-type sugar production and they do not by any means all point in the same direction. The greatest question mark over khandsari-type production is simply that if it is so simple and profitable, why has it not been widely undertaken or indeed undertaken at all by private enterprise? As we have seen, existing large-scale sugar plants do not supply all of domestic demand.20 There should have been room for small-scale enterprise, especially given the presence in earlier years of so many Asian entrepreneurs themselves coming from a khandsari area. The Kenya government and aid donors should, however, be aware that much more large-scale sugar development is likely to pre-empt the market and close the gate to what might be a profitable avenue of development, given recent advances in small-scale techniques. Now that the sugar price is lower, therefore, it is important that alternative techniques should be assessed in advance of further investment. If this is not done, aid donors will almost certainly bear some of the responsibility. Offers of cheap capital for sugar development make it harder for the Kenya government to assess objectively costs and benefits of the different methods, The motives of donors are obvious; large-scale technology is the type that their countries’ firms can supply—Bookers, for example, finding that the methods they knew were profitable, did not consider many

122 Aid and Inequality in Kenya: British Development Assistance to Kenya other techniques. Aid in such circumstances is simply an export credit on very soft terms. There is nothing wrong with that unless it hinders development by seeming to undercut more genuinely profitable alternatives. In the case of sugar, there is just a possibility that it might, and donors genuinely interested in development rather than exports and monuments will investigate the possibility before acting. Unfortunately we suspect that most donors will submit to competition among themselves and rural Kenya may be one of the losers.

Other Considerations on Mumias as an Aid Project Cost-benefit analysis of sugar projects in Kenya throws up several well-known problems associated with this sort of exercise. The first is that of uncertainty. If the output is to be valued at world prices on the grounds that these best represent the value of output to the economy, as the prices that would have to be paid in the absence of the investment, one is faced with the difficulty of guessing the average price that will rule over perhaps 25 years in a commodity market that is notoriously unstable. Alternatively one would have to guess the price at which Kenya could obtain sugar under bilateral deals. Bookers surmounted this problem in the case of Mumias by taking a highly conservative view of the value of output which made the project look unexciting in prospect. Sugar prices have varied over the life of Mumias so far to an extent that renders insignificant possible variations in assumptions about other variables such as technical coefficients. Future assessments will have to take more explicit account of possible world market trends. A more difficult question concerns whether the output should be valued at world prices at all. If there were a boom in sugar prices which were passed on to the Kenyan consumer, the consumption of sugar would probably fall. In other words the world price of sugar during a boom almost certainly exceeds the marginal utility of sugar in domestic Kenyan consumption. Some would say that does not matter because (a) the government must value sugar at the world price to import it in such a situation and subsidise its consumption, or (b) if it does not, and behaves ‘rationally’, it would not subsidise consumption but would export the surplus sugar, so realising the world price anyway. However, if the Kenya government were not to stop subsidising sugar or to export it, we would be in the familiar dilemma of wondering whether to recommend ‘efficient’ production in the sense of that which would yield best results if the government acted optimally, or whether to accept government policies as a constraint and go for a ‘third best’ policy.21 To value sugar at the world price instead of at its marginal utility in domestic consumption in this situation is to underwrite the Kenya government’s pricing policy; to say that as, for example, an aid donor, one considers that the political benefits to the Kenya government of maintaining a low sugar price are a worth-while return for one’s tax-payers’ money. That is a perfectly possible judgement but it ought to be recognised for what it is. It might be a way out to argue that the Kenya government would persist in this policy anyway, so that if a sugar project has a positive return one is reducing the cost to the economy of the government’s policy. This gain has to balance against the effect of the project in helping the government to maintain a wasteful policy. To call it such does not conflict with the notion of consumer sovereignty which entails letting consumers buy as much as they like of something at a price that covers the opportunity cost of production. Consumer sovereignty does not entail subsidisation. Our recommendation to a donor

Aid in Action

123

would be to look askance at any sugar project in Kenya whose social cost-benefit analysis did not have a positive estimated net present value, valuing output at the price the Kenya government proposes to pay for it. If such a rule encourages the government to let the price reflect possible world scarcity, so much the better. At present the Kenyan domestic price is much closer to the world price. Mumias took place with practically no private commercial investment at all. Commercial companies were involved in putting up equity, but Bookers did so to obtain the management contract and the quid pro quo of Kenya Commercial Bank investment was that it obtained the company’s banking business. The reason for commercial reticence was the controlled price regime that operates in Kenya with respect to sugar. The government lays down both the price paid for cane and the price paid to the factory for sugar. It can vitally affect profitability, therefore, at the stroke of a pen. Given that situation, aid funds played a role in getting a worth-while project under way. Now that Mumias has shown its profitability, however, paying a dividend in its first year of operation, there seems a less clear-cut case for aid money for further similar projects. It could be argued that certainly no more should be given than is necessary to launch a worth-while project; there seems little point in displacing commercial money. The only effect of that would be to facilitate a transfer of assets—from the potential investor to the Kenya government. Again that has a political point but it is reasonable for ODM to ask whether it is a political objective worthy of British tax-payers’ money. Perhaps this is the sort of quasi-commercial enterprise for which in some ways CDC is better fitted for a role than ODM. If a project is likely to be profitable and the profits will largely accrue to equity-holders rather than to the economy in a general way, to allow them to go to a recipient government is to admit it can probably put the money to better developmental use than HMG. That may well be so, but then why project-tie aid and evaluate projects at all? For consistency, therefore, this sort of quasi-commercial investment might be left to CDC which can put in equity and re-cycle dividends in the form of fresh aid to Kenya or some other country, whereas ODM cannot vary loan terms for particular projects. There seems to be a conflict of evidence on the question of whether CDC could have handled Mumias alone. But it seems unlikely in fact that it could have found sufficient funds quickly enough. In the case of attractive projects like sugar factories such considerations tend to get swamped. Donors go ahead, as ODM did with Mumias, on the grounds that if they don’t someone else will. The money spent does not finance a sugar factory at all therefore— that would have happened anyway—but the expenditures into which Kenyan, or more particularly, other donor funds were displaced.22 This is satisfactory to the sugar-factory donor on export grounds if the other expenditures are less import-intensive.

Notes 1. Kenya Government, Development Plan 1974–78, p. 240. 2. An account of the Mumias project is given by J.A.Haynes, Director of Booker McConnell Ltd., in Industry, Employment and the Developing World, Overseas Development Institute, 1975. 3. CDC was asked to provide some £750,000 of loan. The request was later withdrawn but has since been reinstated, following the proposal to develop Mumias’s capacity to 75,000 tons.

124 Aid and Inequality in Kenya: British Development Assistance to Kenya 4. The controlled prices were changed when the factory began operation. At the end of 1974 they were K£61.5 a tonne of sugar and K£3.09 a tonne of cane. During 1975 they were increased again, to K£92 a tonne ex factory (after excise duty) and K£4.60 a tonne of cane. 5. 20s=K£1=(approx.) £1.25 Sterling, at that time. 6. An interesting rationale for the Bookers’ pricing is that traditionally the growing of subsistence crops was regarded as the work of women. There was no such stigma attached to growing the cash crop of cane, which therefore mobilised previously unused man-hours. Nevertheless, we doubt if this justifies a zero shadow price. 7. This ignores possible external benefits arising from improvements to the small holders’ general agricultural practices. These were not considered in the report. 8. Prices for 1973 and 1974 from FAO Commodity Review 1973–74, pp. 111, 114; for 1975, newspaper reports. Deflated to 1970 prices: implicit price deflator index for Kenya, UN Statistical and Economic Information Bulletin for Africa No. 6, p. 40. 9. Registered output. 10. Sugar is a perennial crop which can be harvested after a variable growing period depending on conditions. The plant crop is the first harvest taken at Mumias about 22 months after planting. Subsequent crops—ratoons—are taken at 18-month intervals from the same planting. Yields of sugar per acre fall with each subsequent ratoon until eventually, after perhaps 5 years, it becomes worthwhile to replant 11. A.H.Barclay, ‘Aspects of Social and Economic Change relating to the Mumias Sugar Project’, IDS, Nairobi, mimeo., 1974. 12. Wives could be regarded as an agricultural investment good given the labour shortage. As they are in inelastic supply, however, their acquisition for bride money is an asset transfer and the final destination of the money is unclear. As marriages are frequently accompanied by parties, some people have suggested that East African Breweries are the eventual beneficiaries. 13. This failure was, of course, purely local. Elsewhere in Kenya coffee has been the miracle crop to small holders that sugar is in Mumias. 14. These pressures were foreseen and were one reason why MSC wanted an outgrowers company, to take over the selection of outgrowers, subject to technical considerations specified by MSC. 15. Malcolm Harper, ‘Sugar and Maize Meal—Cases in Inappropriate Technology’, IDS, Nairobi, Working Paper No. 170, p. 3. 16. Ibid., p. 12; C.G.Baron, Sugar Processing Techniques in India, Geneva, ILO; Industry, Employment and the Developing World, op. cit. 17. Khandsari was the name of the product of traditional cottage plants; it was powdery in texture. We also use it for the more crystalline product of contemporary small-scale plants. 18. Results vary with location and these figures are approximate. Baron, op. cit., gives slightly different ones. 19. See Deepak Lal, Appraising Foreign Investment in Developing Countries, London, Heinemann, 1975, pp. 61–6. 20. This fact seems to be a complete answer to those who argue that Mumias was a poor investment on grounds of its inappropriate technology. Speed in execution was important, as it happened, and this was achieved by using more tried techniques. 21. This problem is lucidly discussed in I.M.D.Little and J.A.Mirrlees, Project Appraisal and Planning for Developing Countries, London, Heinemann, 1974, Chapter XIX. Also see A.K.Sen, ‘Control Areas and Accounting Prices: An Approach to Economic Evaluation’, Economic Journal, Vol. 82, March 1972. 22. An instance of the aid fungibility problem discussed in Chapter Six.

Aid in Action

125

Part Three: The Special Rural Development Programme Promoting rural development is seen by many concerned people as the greatest task facing the Kenya government. Virtually all the important aid donors have come around to this view, and all of them pay lip service at least to the need to direct aid to help in that area. The objective reasons are clear enough and were set out both in the ILO report and in successive World Bank reports on the Kenyan economy. The overwhelming majority of Kenyans, nearly 90 per cent, live in rural areas and they are very poor; the higher incomes earned by some in urban areas have led to migration into towns but the rate of growth of productive employment in urban areas has failed to keep pace. The ILO report1 estimated that 20 per cent of the men in Nairobi and an even higher proportion of the women were without work or in very low-paid casual occupations. This situation is breeding slums, crime and social unrest. Given an overall population growth rate of over 3 per cent and the much faster increase of urban populations, it is difficult to see how industrial employment can ever hope to keep pace. The policy of encouraging import-substituting manufacturing industries has increased manufacturing output but, starting from a low base, employment has grown more slowly, owing to improvements in labour productivity.2 Rural development is regarded as essential, therefore, because it will affect areas where the mass of poor people live; it will reduce the imbalance of urban-rural incomes, directly alleviating some social problems, and by concentrating economic growth on agricultural and other processes that are labour-intensive it should expand employment along with output, That in turn will have a beneficial effect on income distribution. An investment in agriculture and related agro-industry might be less import- and capital-intensive than that in concentrated urban industry, so faster growth might be consistent with balance of payments equilibrium. Rural development is seen therefore as a key to long-run economic growth as well as to greater equity by the reduction of wealth inequalities within as well as between areas. The Kenya government In its Development Plan for 1974–78 stresses the primacy of rural development: The 1970–1974 Development Plan stated that the key strategy for this Plan is to direct an increasing share of the total resources available to the nation towards rural areas. That strategy will be continued and indeed reinforced during this Plan.’3 This is the situation that has lent importance to the Special Rural Development Programme (SRDP) which has been supported by five bilateral donor agencies including Britain as well as the FAO. Quantitatively, it is a relatively minor British aid project—some £250,000 in capital aid and a couple of technical assistance appointments over a number of years. Yet in view of ODM’s commitment to rural development the lessons of SRDP have an obvious importance. The idea for SRDP was born at a conference at Kericho in 1966 on education, employment and rural development.4 Donor representatives at the conference expressed interest in considering Kenyan government proposals5 in the field of rural development and this may have been the reason for continued activity on the subject within the then Ministry of Planning.6 It was decided to select a number of areas for pilot experiments in rural development, and the Institute for Development Studies (IDS) at the University of Nairobi was called in to help formulate a programme. The major proposals of its report7 were the following:

126 Aid and Inequality in Kenya: British Development Assistance to Kenya (i) (ii) (iii) (iv)

The carrying out of a base-line survey of existing rural conditions; that pilot projects must be designed for replicability; that a central co-ordinating national body be set up; that donors be utilised but made to understand that this was an extraordinary programme demanding flexibility on their part; (v) that the programme be handled through the existing machinery of government, and that one of the prime goals of the entire exercise be to enhance Kenyan capabilities in every aspect of rural development planning and implementation. ODM offered a central co-ordinator who arrived in Kenya in March 1968. Ajoint governmentIDS working party was formed and selected fourteen areas of the country which were then surveyed for information on actual conditions and existing government programmes. At the beginning of 1969 a National Rural Development Committee was formed to co-ordinate the programme and negotiate external assistance for it. Its members were the Permanent Secretaries of the most important ministries concerned with the programme. Six areas were chosen for the first phase of the programme8 and the preparation of detailed costed plans for these areas, with outline five-year programmes and the annual budget for the first year’s operations, took up to the middle of 1970, The aims of the programme were established as: ‘The primary objective is to increase rural incomes and employment opportunities. The secondary objective is to establish procedures and techniques for accelerated and self generating rural development which can be repeated in other similar areas and, in particular, to improve the developmental capacity of Kenya Government officials in this field.’9 During 1970, discussions also went on with donors over finance. The Kenyan idea was simply to have a rural development fund to which donors would contribute. This, however, violated the entrenched donor prejudice for the results of their aid to be identifiable. Some donors were also loath to break rules against supporting local and recurrent costs. This situation led eventually to the identification of each SRDP area with a single donor who reimbursed the Kenyans for SRDP expenditure in ‘his’ district. Even that took some time to accomplish and the programme was held up for six months while the Treasury refused to allocate funds to SRDP in the 1970/71 budget despite its inclusion in the 1970–74 Development Plan. The reasons for this reluctance are unclear but it has been suggested that it was symptomatic of a general lack of enthusiasm for SRDP on the part of some officials.10 As it was to prove so critical, the coolness of influential people in the Kenya government towards SRDP stands analysis. It was perhaps to the disadvantage of SRDP that its National Rural Development Co-ordinating Committee was established in the Ministry of Planning.11 Some officials in operational ministries feared that their autonomy was endangered by the planners whom they characterised as paper-pushers. This interministry jealousy was aggravated when the University of Nairobi became involved in the planning of SRDP. There has at various times been tension between senior civil servants and senior academics in Kenya. The former do not always regard the criticisms of the latter as being offered in a constructive spirit. The conjunction of the Ministry of Planning and the University was enough to alienate some senior bureaucrats in operational ministries and the alienation was increased when an energetic expatriate took over SRDP co-ordination at the national level.

Aid in Action

127

The operational ministries also regarded SRDP projects as an additional administrative burden on top of their routine sectoral programmes, which were, of course, nearer to their collective hearts. The Treasury took a sceptical view because as far as it could see SRDP expenditures were recurrent in nature, and could be classified as consumption rather than investment. It was not entirely happy about authorising them even if they were covered by promises of aid, partly perhaps because of a reluctance to finance consumption out of repayable loans. The programme received a fillip when the Norwegians announced in August 1970 that they would fund all expenditure in the Mbere division. Norwegian aid takes the form of grants. Other donors gradually followed suit.12 The Norwegian decision had an important effect in strengthening Kenya-based representatives of donor agencies in appeals to their national headquarters to fund local and recurrent costs on an unprecedented scale. ODM was the last donor agency to provide capital funds for SRDP but was relatively innocent of attempts to change its nature. Of the 1970 UK loan commitment, some £250,000 had been earmarked for rural development and when asked if this could be put into a common SRDP pool ODM said it would not rule that out if other donors agreed. After other donors had begun to support individual areas it was the Kenyans who in February 1971 proposed to spend the £250,000 on SRDP and suggested that the UK government might wish to have this contribution identified with a particular project area. After considering Tetu, ODM finally decided to support the Kwale area on the disinterested developmental grounds that it was much the more backward of the two. The other areas were already associated with foreign donors. Donor preferences had other effects, however, beyond changing and complicating financial procedures. Once donors had ‘their’ areas many began to want to put in their own personnel as full-time technical assistance. This had obvious dangers, because the whole point of the programme was to try out management procedures as well as physical projects relying as far as possible on Kenyan resources and the Kenyan administration so that any successes in promoting development could be replicated elsewhere in the country. The Kenyans were unable to resist13 the pressure for local-level technical assistance involvement. The pro-SRDP group in the Kenya government, located in the Ministry of Planning, was not all-powerful, and donor interest and leverage had been essential to it in getting things moving at all. ODM, starting with its project a year later than other donors, was aware that both the USAID approach in Vihiga and that of the Swedes and FAO in Migori, of providing a number of advisers and carrying out extensive planning, had come in for criticism. So it opted for providing just one ‘project adviser’ who would work in Kwale attached to the Ministry of Finance. It seems, however, that this was seen as a point of diplomacy rather than a point of substance. Other donors were regarded as having been tactless rather than wrong-headed. The British project adviser would in fact be a project director, but he would be unobtrusive and lead from behind. He could be supplemented by other experts or consultants on a short-term basis. Perhaps attitudes can be summarised as follows. The pro-SRDP group in the Kenya government needed donor backing to make any headway against high-level opponents of SRDP, particularly in the Treasury and Ministry of Agriculture; this made it vulnerable to donor pressures. Many Kenyan administrators regarded SRDP as an ‘expatriate show’ which they would not allow to interfere with their own departmental policies; its usefulness to them was as a device for getting aid donors to

128 Aid and Inequality in Kenya: British Development Assistance to Kenya commit untied funds. Rural development experts in donor agencies wanted an involvement in SRDP because of the lessons it might yield them, as well as the Kenyans, about rural development, and no doubt they found it hard to resist the expert’s suspicion that he is indispensable. Donor administrators had already had the bracing experience of committing untied funds to no clearly defined project and were keen to insert technical assistance as a means of regaining some control over the expenditure that was being incurred in their name. Although more aware of Kenyan sensitivities than other donor agencies, ODM by and large shared these attitudes.14 The need for technical assistance in rural development, what form it should take and what its role should be, we regard as very open questions. A lot depends on the details of the situation and the personalities involved. The constellation of attitudes described above, however, indicates that cases were unlikely to be treated on their merits. There was a systematic donor bias in favour of employing expatriate technical assistance on a fulltime basis. The programme began on the basis of plans that had been largely drawn up for each area at the centre. The plan for Kwale was almost entirely an exercise in planning from Nairobi by the Ministry of Finance and Planning. IDS monitored progress in SRDP areas during the first year of the programme’s life and in 1973 published an overall evaluation of SRDP.15 At that time very little had happened at Kwale. The programme there had started later as it was not clear which donor, if any, would provide support for the area during most of the Kenyan financial year 1971/72. The IDS evaluation was rather critical of SRDPin general. It re-emphasised the experimental nature of the exercise and the requirement that experiments should be aimed at achieving replicative processes. In fact, it argued, a lot of the planning had been concerned with putting in more resources to obtain rapid results with the probable outcome that benefits would end with the particular projects, and long-standing bottlenecks to development had not been overcome.16 Apart from excessive haste another factor making for failure was lack of success in involving local people and tapping local abilities and knowledge. District Development Committees, for example, are largely instruments of the government. The evaluation also noted the lack of authority which SRDP enjoyed at the centre and recommended a top Kenyan of at least Deputy Secretary level to be in charge of it. External planning without local involvement led to one or two well-publicised failures, notably an attempt to get farmers in Mbere to grow cotton because it appeared technically suitable. It transpired, however, taking risks into account, that cotton was not an economic crop in the area and the farmers’ resistance was entirely rational. A credit programme in Vihiga to get farmers to concentrate on hybrid maize production was also unpopular with local farmers for substantial reasons.17 That is not to say that SRDP had been a total failure. A number at least of the 123 sub-programmes under way in the six areas represented interesting experiments in the areas of agricultural extension and training, credit provision, and supply of other services. Supporters of SRDP claim that perhaps its greatest achievement then and later was in ‘beginning to get to grips with the fundamental organisational bottlenecks which so often hinder effective rural development. The decision taken at the outset to work through existing government machinery had important beneficial results.’18 Experience led to the development of prototype management systems. The most significant, Programming

Aid in Action

129

Implementation Management (PIM) for the government machine, was built around the appointment of ‘area co-ordinators’ who were established in the Office of the President and had the task of co-ordinating the development work of other ministries. PIM provided procedures for monitoring the work of field staff and focusing attention on bottlenecks within the administration in implementing programmes.19 In 1975, IDS prepared another report on the progress of SRDP, We shall be concerned in the remainder of this chapter with the experience of the SRDP in Kwale, and that of ODM as an aid donor, merely fortifying conclusions where possible by drawing parallels with experience in other SRDP areas. Here, therefore, we give a list of the main IDS conclusions classed into favourable (A) and unfavourable (B) as background to our own discussion: (A) SRDP has demonstrated area-based plans can be prepared and implemented by local staff. Infrastructure, especially of roads and water supply, has been developed in SRDP areas. Management and reporting systems have been developed with varying degrees of success. Some useful lessons have been learned about specific problems, notably about operating credit schemes for small holders and using labour-intensive methods in road construction. (B) The major weakness has been, despite the emphasis on experiment, that very few new strategies have been developed. Money has been spent on the obvious social services and infrastructure needs so that SRDP has emerged as another form of area development on a weak agricultural base. Although the areas are now more fortunate than average there is little sign of self-sustaining indigenous development. There has been disproportionately heavy expenditure on recurrent costs and social services. In the first two and a half years, 75 per cent of spending was on salaries, housing and vehicles and only 25 per cent on capital formation.20 The principle of replicability has not been adhered to. One demonstration is that when SRDP concludes, as it is now scheduled to do after five years (January 1976), the Kenya government will be liable to recurrent costs of £K1m. generated by the programme. The total cost of SRDP programmes was under £K4m.

SRDP in Kwale Kwale district is on the Kenya coast south of Mombasa. The SRDP area, some 417 square miles, did not for a long time include the whole district but only the southern part of it, taking in coast and hinterland. It was extended in 1974. The area is not densely populated. The 1969 census recorded 53,000 people, one to five acres. However, population is increasing rapidly not only through net reproduction proceeding at least as quickly as the high Kenyan average, but also because of migration into the area. The indigenous tribes, the Digo and the Duruma, are Moslem-animist. They are not intensive agriculturalists; the Digo rely largely on tree-cropping (principally coconuts and cashew) and the Duruma on animal husbandry. The attitude of the elders of these societies to change appears to be more than usually conservative. The migrants are mainly Kamba, who are generally Christian and more progressive farmers. Many of them hold land as a result of transactions with the Digo, but in a not unparalleled fashion, the nature of the transaction, whether for freehold

130 Aid and Inequality in Kenya: British Development Assistance to Kenya or leasehold, is disputed. This tribal interaction, accompanied by a good deal of bad feeling, is one of the outstanding facts of social and political life in the district.21 The area falls into three basic regions. The coastal croplands are a narrow strip some 25 miles deep, with high rainfall (over 50 inches yearly) and good soils. Population here is relatively dense and tree crops and cassava are the main crops. The marginal lands are further inland where rainfall falls off sharply and is less reliable. The ranching lands are further inland still, where the rainfall is inadequate for any other enterprise. The district is not untypical in climate, physical features, and economic activity of many other areas in Coast Province and that Province, in general, is not among Kenya’s more developed areas. Experience in Kwale therefore could have relevance to pressing problems of development. The plan laid down for Kwale SRDP, issued in January 1971, was written by Ministry of Planning personnel in Nairobi and was largely an exercise in planning from above. Existing projects in the area were supposed to be integrated with SRDP. These included a rural water project, financed by the Swedish aid agency, SIDA, and plans supported by IDA and HMG funds for a Development Centre incorporating a demonstration farm and facilities for courses in agriculture and adult education. The constraints to further development were identified as lack of water, roads and transport, and an overstretched agricultural extension service. A number of directly productive projects were planned. These included establishing sugar outgrowers to supply an Asian-owned factory running below capacity; setting up tree-crop and cassava nurseries; building cattle dips as part of an animal disease control programme and instituting a commercial ranch on the basis of co-operative ownership. There were also more general supporting programmes. Land adjudication was already proceeding in the area, supported by British funds; a programme of grading tracks to make roads providing access to the interior was started and various community development campaigns, including women’s courses and functional literacy courses, were planned. The latter were necessary as census returns indicated 79 per cent of the population had received no formal education. The implementation of the parts of this programme, in Kwale as in other SRDP areas, was in the hands of the operational ministries concerned.22 Co-ordination was in the hands of the District Development Committee. There was also a District Development Advisory Committee to include people not in the administration, like MPs and local self-help leaders. SRDP was expected to revitalise these committees. Early experience in other SRDPs also led to the appointment of an area coordinator, who was a sort of district planning officer charged with co-ordinating the work of the different ministries. This had to be done by cajolery as he had no authority over them and no access to funds. The area co-ordinator arrived in post in Kwale in January 1972. Each ministry in Nairobi had ‘linkmen’ who were supposed to liaise over SRDP at head-quarters level. The funds for SRDP activities were voted to the ministries within whose normal area of responsibility the activities fell. The area coordinator had to apply to the appropriate operational ministry for authority to incur expenditure falling within its ambit. In the case of projects involving more than one ministry, each had to be approached for authorisation and approval. Given even ordinary bureaucratic delays this could cause problems in the timing of projects but given the indifference or hostility to SRDP within some ministries, severe problems were encountered which the National Rural Development Committee

Aid in Action

131

failed to overcome. Even after authorisation to incur expenditure had been granted, no cheques could be signed by officers of the administration below provincial level23 which led to further delays in spending voted funds. Authorisation delays led to some funds which had been voted being lost and returned to the Treasury at the end of financial years. Many of the projects have been dogged by failure from the start. The Matuga Development Centre was to be multidepartmental with a farmer training centre incorporating a demonstration farm, and provision for public health and literacy courses. Some of the funds were voted to the Ministry of Agriculture and some to the Ministry of Education. Disputes arose over who was responsible for financing and running the centre. It was delayed owing to the inability of the Ministry of Works to provide adequate plans. It has only just been built and may not be in use before the end of SRDP in 1976. Seed nurseries were quickly set up but half of a 26-acre sugar cane nursery was destroyed by arson ascribed to local jealousies; only 15 acres of a planned 100-acre nursery could be planted with seed-cane owing to shortages of it. Some 15,000 coconut seedlings were planted in another nursery but arson occurred again and 13,000 were destroyed. All did not go smoothly with the ranching plans. There was considerable opposition to the idea from the Duruma despite barazas (public meetings) and a seminar at which an attempt was made to sell it to them. Even when their approval in principle had been won, it was difficult to reconcile the interests of everyone in the ranching area as some people were cattle-minders, not cattle-owners. It was decided to set up a central herd, but drought and opposition meant that by the end of 1974 only 100 animals had been pledged to it. A number of dams were planned for the ranching area but there was a delay over receiving plans for them from the Ministry of Agriculture, despite repeated requests. When plans arrived there was a further delay in authorising expenditure. Authorisation was finally received so late that building had to be rushed through, so expenditure could be incurred before the end of the financial year. A local contractor was employed and not adequately supervised. The dams were badly sited and poorly built and further expenditure will be necessary to make them serviceable. Cattle dips were planned but authorisation was received for only two. These were built partly using self-help labour but one was caught by the end of a financial year and left halffinished. The local people who had promised to provide finance for dipping fluid failed to do so and the dips are unused. A health centre was nearing completion at the end of 1974. It has, however, come in for criticism as being too elaborate—it is capable of dealing with 400 outpatients in a single day. On the other hand it is not clear whether there will be enough trained staff to match the building. There are not enough to go into outlying areas effectively and in any case they have no transport to do so. Functional literacy courses and women’s clubs have been started although the former are behind schedule. In addition social difficulties have been encountered. The Digo and Duruma are more reticent than the Kamba about attending—and this applies strongly to their womenfolk where Moslem mores have impeded participation. On the other hand the provision of some other infrastructure has gone well. The SIDA water project, after early difficulties, is working. And the road programme was carried out quickly and at very low cost by the Ministry of Works which simply graded the most used tracks rather than planning new roads ab initio. Some 94km has been laid.

132 Aid and Inequality in Kenya: British Development Assistance to Kenya A number of common threads can be traced through many of the failures. One is that co-ordination between ministries has not been adequate, despite the efforts of the area coordinator and the National Rural Development Committee. One example is the Matuga Development Centre, which is not strictly part of SRDP, but it points a general pattern. The failures of co-ordination are a symptom of the second, deeper problem, that many ministries are indifferent to SRDP and give its expenditures no priority. An example of this is related below, but the fact, combined with the absence of financial delegation, led to the failure of the dips and dams. This indifference of the ministries is confined to the Provincial and Nairobi level. Local staff are usually glad enough of the opportunities and money that SRDP (eventually) brings. A third problem is the failure to involve the local people adequately in SRDP project planning and implementation. The evidence of this is scattered but widespread: the arson, the slow progress with ranching, the dips not being supported, the low attendance at courses. A contrary sign is that farmers are said to be responding to the stepped-up agricultural extension effort in the area. There are a number of reasons for this lack of local involvement. One may be a faulty government strategy to start with. The criticism that locals were not sufficiently involved in the planning of SRDP was noted and countered with the proposition that their support would come quickly enough when they saw a demonstration that the government was making a real input—roads and water—and then it could be hoped that in time the people would ‘become involved in the developmental processes, especially the entrepreneurs, the better farmers and the younger generation…’. The involvement was thus seen as responding to governmental initiatives. One such initiative was to persuade herdsmen in the hinterland to stop milk production and switch to ranching. Milk production was to be concentrated nearer the coast. This radical change was ‘planned’ without consultation. More recently it has been suggested that tsetse fly control will not be as easy as once thought and might hamper ranch developments. On the other hand, milk might, after all, be a good cash product if more roads were built so that its producers could get it to market. In general it is a bad idea to attempt to implement radical and risky plans without taking the ‘plannees’ along with each step of the planning process. Another problem is that government officers are usually from cultures that are quite alien to the local tribes, as the Kenya government tends to post officers away from their home areas. Many of the administrative and agricultural officers we met in Kwale were Luo whose homeland is about as far as it is possible to get from Kwale without leaving the country. Being Christian they have more in common with the immigrant Kamba than the Digo or Duruma, which is itself enough to create suspicion. There has been a tendency for government officers and extension workers to try to introduce agricultural techniques, for example, that are suitable for other parts of Kenya, but not for the coast, sometimes in response to national ministerial directives. An area like Kwale has a humid and debilitating climate and enervating diseases such as bilharzia are endemic among the population. On the other hand the land and sea have provided people with subsistence without their having to work very hard. Increasing population pressure will make this more difficult, but the local response is more likely to be hostility to immigrants than a desire to start working harder. Development can be seen in such circumstances as something that makes invasion of one’s land possible rather than something to benefit one’s family. Therefore it was particularly

Aid in Action

133

desirable to relate projects as far as possible to the local people’s perception of their needs rather than trying out the bright ideas of outsiders. A further problem is that not only do government staff represent outsiders but they are seldom in place long enough for personal relations to build up. Rapid turnover of locallevel staff is a ubiquitous problem in Kenya, and Kwale is no exception. Practically no post is held by the same man now as it was when SRDP started and that is true of the area coordinator and the District Commissioner as well as less important officers. The lack of continuity increases the risk that government staff will be unaware of local constraints to the policies they are trying to implement. Quite apart from the problems of Kwale SRDP as an area development project, the IDS report on SRDP in 1973 had pointed out that there was nothing ‘special’ or experimental in many of the developmental programmes under way. In Kwale the extra finance made available through SRDP was used to employ more staff by the ministries, so that more livestock and health officers, for example, were in the district but they were pursuing standard departmental programmes using standard techniques. SRDP crop programmes using small agricultural demonstration trials were normal extension activities pursued on a piecemeal basis without reference to any crop rotation or system. The British project adviser who arrived in post in September 1972 therefore pressed for SRDP money to be used for an intensive extension experiment. By his energy and persuasiveness he got this SRDP demonstration farm concept accepted by the local officers. The strategy was based on the belief that farmers needed to be convinced that they could reproduce the same results as those obtained on Department of Agriculture trial and demonstration plots and that the return justified the expenditure. This could only be achieved and farming practices modernised, it was thought, by physically demonstrating entire farming systems on the farmers’ own land. The idea was for field staff of the Department of Agriculture to plan the farm layout and then work alongside the farmer every day during an initial on-farm training period. A number of small and large willing farmers, adjacent to one another, would be chosen, at first in areas where soil maps were available, adjudication complete, piped water on, access roads opened up, a farmers’ co-operative established, and crops proven by trial. At a later stage a number of farmers in the marginal lands would get the same treatment when a system had been developed for such districts. All SRDPAgricultural Assistants, Animal Husbandry Assistants, and sub-staff were to be assigned full-time to the chosen blocks. Once modern farming was established, the extension staff would move on to another area. Some data were available as a basis for farm planning; there was a semidetailed soil survey covering part of the district, rainfall figures were available, and some crop trials had been conducted by FAO. Requests were forwarded to the Planning Unit in the Ministry of Agriculture for profitability studies to be made of possible crop rotations and for an extension of the coverage of the soil survey. Given the Planning Unit’s inability to cope with its existing work, leading, as noted, to long delays in preparing major project submissions to aid donors, there was no chance of these studies being done. And given the Ministry’s indifference to SRDP it was predictable that the requests would not be used to obtain technical assistance for the job. After waiting for eight months the project adviser began to do the planning himself for non-tree crops, coming up with a proposed rotation based on maize, sim sim, groundnuts, pigeon peas, chillies, and sunflower. He was an

134 Aid and Inequality in Kenya: British Development Assistance to Kenya experienced man and the work was thorough: the crops were largely familiar to local farmers and the project adviser took account of the labour requirements of each and their seasonal distribution; he included the cost of labour and took the profitability aspect of farm planning into account. When it was proposed to get the local resources of the Ministry of Agriculture devoted to the programme, as well as channelling credit to selected farmers, the project stalled again. The Ministry of Agriculture had recently decided on a policy of district-wide extension and refused to authorise its staff to go against that directive. Eventually the deadlock was broken but not until the second half of 1974—three and a half years after the Kwale plan was issued, and two years after the arrival of the project adviser. Thirteen farmers were then selected, although the Ministry had still not issued any instructions to the Agricultural Finance Corporation about credit. Clearly, the experiment is most unlikely to have yielded any useful results before the end of the SRDP period. The Ministry’s actions were not simple obstructiveness or inefficiency. As a general policy the project adviser’s approach was subject to a number of objections. The spread effects of which he was confident have often failed to materialise when intensive extension has been applied elsewhere.24 In other SRDPs, experiments were being conducted to widen extension coverage with given resources, on the reasonable grounds that if some farmers are given an advantage over their fellows there is no reason to expect them to work hard at throwing it away by educating the others.25 In the absence of spread effects, if the farm plans were successful the effect would be to increase inequality especially as the farmers chosen would be the enterprising ones with access to inputs like water and roads. It may well be politically unacceptable to concentrate government resources to achieve such an effect, especially in Kwale. The sociological background in Kwale did not, a priori, look too good for the concentrated approach. Initial success could simply lead to more arson. In addition the approach could not be said to be economical in its use of skilled man-hours. We do not know what sort of farm plans local staff in Kwale could have drawn up but we do know that the Ministry of Agriculture does not have the capacity to prepare worth-while detailed farm plans on demand for the different districts of Kenya. Doubtless any plan adopted would require modifications in the light of experience, and who would make them, as junior field staff are often timorous about departing from plans they have been told to implement? These questions put in doubt the scheme’s replicability without a continued technical assistance input. The Ministry of Agriculture no doubt feared this was a case of ‘over-sophisticated techniques which generate shock-waves of demands for sophisticated information that only more planners can provide.’26 In short the drearily familiar phenomenon of technical assistance generating the need for more technical assistance. The justification for the project adviser’s approach was that it was experimental. By using rural growth centres it sought not simply to effect marginal improvements in agricultural practices but to ‘rationalise’ farming in the area, revolutionising the outlook of a few farmers at a time, and preparing their entry into the monetary economy. It was clearly ambitious, perhaps excessively so, but it is a pity it did not get off the ground earlier. Although it was risky, risk expenditure, even if classed as recurrent, is what SRDP was supposed to be about. Experiment can only consist of trying out various ideas, and failures can be as instructive as successes.

Aid in Action

135

Yet it must be said that the episode illustrates the dangers of technical assistance proposing plans that contradict a departmental policy. So much time is spent in in-fighting that none is left for the implementation of any policy. In the final analysis plans have to be evaluated not in isolation but in the political and administrative environment for which they are proposed. It is a pity both that the Ministry was not more flexible and that the project adviser did not plan taking into account the fact that it was not.

The Lessons of SRDP for Aid Donors It is our conclusion that technical assistance is unlikely to be productive in area planning at the local level in Kenya. In the first place, in areas as undeveloped as Kwale we suspect that extensive resource studies leading to a master plan for development are a waste of time. In our view experience of SRDP, including Kwale, bears out some remarks of Chambers: The resource use strategy approach…much espoused and advocated by perfectionist planners, has a powerful appeal. While the forms it may take vary, it may be designed to include resource inventory and appraisal, determination of objectives and criteria for choice between alternatives, a search for and formulation of alternatives, and then choice between them and strategy design, with various forms of iteration between these. Ideally, of course, it should lead through to the working up of projects, budgets and action programmes and then to implementation. In practice the bitter and repeated experience has been that the process is slow, ponderous, given to premature elephantiasis, and exceedingly difficult to push through to implementation. The most common outcome is a large mimeographed document which presents much data about an area and sometimes suggestions of development strategy, but no detailed realistic of costed action proposals.27

Chambers concludes that such an approach leads to excessive data collection,28 delays in processing, an absence of action proposals, domination by expatriate advisers, and low involvement of local-level staff.29 He might have added low involvement of local population too. If this approach is no good in a pilot rural development scheme, what are the alternatives? We believe a more gradualistic approach is called for, accepting by and large the choice of crops of the more progressive farmers at least and testing extension methods to get the techniques of all farmers up to the best local practice level, combining this with experiment in minimising the cost of providing services like roads, water, and agricultural inputs. The great spread in level of technique found among farmers in rural Kenya implies considerable gains in output if the average is merely raised to the level of the current best.30 More ambitious project proposals could grow out of such an approach if improved twoway communication could be developed between government staff and the local people, and between field-level and headquarters-level staff. This entails appointing local staff with local knowledge and/or leaving them in post longer and improving management techniques within ministries. Even then we are not starry-eyed about local staff generating experimental project ideas. One of the great supporters of SRDP concluded: ‘It was found…that field officials could not even after two years produce good project plans let alone comprehensive area plans without a great deal of direction and guidance from widerlooking, more experienced, centrally based officers.’31 It is difficult to get the balance right; nevertheless, in common with many observers, we believe that the balance of initiative and power lies too much with the centre at present.

136 Aid and Inequality in Kenya: British Development Assistance to Kenya Our conclusion is that any technical assistance is best posted at Provincial level or higher where it can be called in by District staff to give specific technical advice as and when required, or better, help in the training of planning officers. Technical assistance at a lower level is often able to give salutary advice, but ultimately the role of an adviser not involved in implementation is difficult to sustain. When planning, without the responsibility to see to implementation, there is always the temptation to assume that an efficient administration is waiting to receive and act on one’s quasi-optimal plan which will thus be implemented irrespective of whose political or administrative nose may be put out of joint. That is not true of any administration in the world. Yet it is only people who have to carry plans out who are likely to take such tedious constraints into account and whose plans are likely to have more than pedagogical value.32 That is why the experience of technical assistance in SRDP was almost uniformly unhappy. It either moved into implementation, as in Vihiga, or planned without implementation, as in Kwale. In Migori, things only began to happen when the original SIDA/FAO team moved out.33 ODM was aware of the dangers that a large-scale technical assistance input would, through the internal logic of the situation, tend to take over implementation and destroy replicability, so why did it put in technical assistance at all at local level? In fact it does not seem to have considered not doing so. One function which a project adviser could have—and has had—is that of helping local officers to train themselves, but this is not the whole story. British aid is in general underadministered. There are too few FCO or ODM officials on the ground to monitor projects or keep adequately in touch with British technical assistance personnel. Technical assistance can thus be given in an attempt to control projects by proxy in the belief that yet another expatriate will keep his eye on things and make sure ‘British’ money is being well spent. As Kwale cannot be described as a success, either as a development project or as an experiment, it is clear that this technique did not work. It was not the project adviser’s fault. There can be no doubt that the greatest bottlenecks to SRDP were not at Kwale but in Nairobi. A project adviser attached to the Ministry of Finance and Planning and stationed in Kwale was not in a position to do anything about them. Yet at the Nairobi level HMG entirely failed to use any leverage at all to support the proSRDP group in the Kenya government. Its negative reaction to a US initiative to tackle the Kenya government about its lack of support for SRDP has been referred to.34 This incident occurred at about the time that the National Rural Development Committee co-ordinator, supplied by ODM, finished his contract, having been told, to his surprise, that the Kenyans were not applying for a further term of his services. Although the reason given was that the job could be done by a Kenyan, no appointment was made for five months and then a very junior officer was appointed. The co-ordinator was quite clear after leaving that SRDP had failed to realise its potential because of lack of institutional support at the centre. His conclusions are now accepted by some British officials with local experience and were indeed admitted by a very senior official of the Ministry of Finance and Planning at a staff seminar on rural development at the Kenya Institute of Administration in January 1975, who said support in key ministries had been less than he hoped. This danger was clear at the end of 1972.

Aid in Action

137

Rather than putting in full-time technical assistance at Kwale, which was always likely to damage replicability and did not in fact much speed up the project, ODM would have been well advised to employ another adviser in the East African Development Division, Nairobi. Meetings with SRDP field staff and the SRDP central co-ordinator would have persuaded an adviser of the importance of backing the US initiative. The Kenyans could have been flatly told in 1973 that unless there was more co-operation and faster release of funds (if not financial delegation) by Nairobi ministries, no more money would be forthcoming for SRDP. Some ministries might have been annoyed but elements in the Ministry of Planning would have been grateful to ODM for this drive. The Kenya government is not monolithic, and in any case there are pleasant as well as unpleasant ways of telling someone you are going to pull the rug out unless he stands up. Leverage could also usefully have been exerted, or at least suggestions made, at Nairobi level on details instead of leaving a project adviser to tell field staff who, without influence, then had to transmit the message up blocked ministerial channels. In this way representations could have been made about the rate of staff turnover, for example. Such pressure might have resulted in requests for technical assistance to do specific short-term tasks and relieve specific bottlenecks—perhaps carry out dam survey and designs. Requests from field staff carried no weight. It was up to British aid advisers and administrators. We are not entirely confident leverage would have been successful. Perhaps the political basis of SRDP was after all so friable that it would have fallen apart, so to speak, around the lever. At worst ODM would then have saved some money.

SRDP and District Development The Kenya government has decided after 1976 to phase SRDP expenditure into the general budget. It now proposes District Planning instead. District Development Committees consisting of the former bodies of that name and the old District Development Advisory Committees will be charged with preparing plans. A District Development Officer, trained in development planning and project preparation, will be posted to each district to assist. Provincial Development Committees will assist with professional guidance. Plans are to be forwarded to Nairobi and any projects not fitting into ministry programmes may be financed by District Development Grants. An average of £K25,000 a year for each district on this basis was set in the current Development Plan,35 now increased to £K50,000; Forty District Development officers are to be posted by June 1976. The justification offered for passing from SRDP to general District Development Planning has been that substantial political pressure is being exerted for a more general programme, and great concentration in SRDP provided a politically vulnerable target. It is also argued that SRDP has already demonstrated the potential for area-based planning and further experimentation is unnecessary. However this, in our view, is by no means conclusive. Certainly, to be successful District Development Planning would require adequate co-ordinatory units in Nairobi with full-time responsibility for district planning, headed by adequate influential and senior staffs, able to make swift decisions. Absence of this has led to many of the failures of SRDP. Lack of high-level support in both policybacking and day-to-day administration must be overcome if District Development Planning is to work. The adequacy of the existing administrative machinery is open to doubt in view

138 Aid and Inequality in Kenya: British Development Assistance to Kenya of the pace of implementation of projects. Several government departments are heavily committed to ongoing sectoral programmes, and staff and facilities are barely adequate to meet present needs. It could prove difficult to reconcile specific district requirements with sectoral priorities. Most detailed write-ups for SRDP projects were done at provincial level or in Nairobi, and the same is true to date of district plans. Few spontaneous proposals from field staffs are received and there is very limited feedback even when they are asked to implement proposals that are inappropriate. Acceptance of the decisions of senior staff is strongly ingrained in the system and leads to a tacit acceptance of instructions that had just been passed down to the field.36 Where grass-roots proposals are made, they are usually requests for infrastructure or social service spending rather than pointing opportunities for directly productive investment. The Kenyans wish to get donors to subscribe to a Rural Development Fund. So far the Dutch, Swedes and Danes have accepted the idea in principle and agreed to contribute. The UK, using the excuse of the 100 per cent local-cost content required for district development funds, has demurred. The ODM feeling appears to be once-bitten at Kwale twice-shy over District Planning, and it is better to stick to discrete projects, although technical assistance might be provided.37 That seems a pity. Any technical assistance given other than for training is likely to be a waste of money. On the other hand the usefulness of District Development Planning in generating proposals for development projects will depend on management procedures, including the arrangements for financial devolution and the willingness of the centre to give up some authority and some able personnel. This is a worth-while area for ODM to hold discussions with the Kenyans, in conjunction with other donors. Needless to say, capital aid should be given only if the details of the operation of District Development Planning and its financial control are satisfactorily worked out, the political adherence of the centre is established, and adequate District Development Officers can be made available. ODM experts may well have an important contribution to make to such discussions, and an entrée could be bought by the offer of a few hundred thousand pounds. The returns are likely to be greater than dropping some hapless technical assistance officers into a local situation. Some devolution is necessary in Kenya if local people are to be involved in development and it may eventually be a way of breaking the current bottleneck in project identification, if not preparation, District Planning is therefore an important initiative. What is in doubt is the extent to which the Kenya government takes it seriously. Some £K8m. is all that has been allocated—not enough to make the Treasury wince. SRDP, inter alia, was a device to get donors to commit untied funds, and no doubt District Planning is the same. In our view the correct response is not to shun it but to play back at the Kenyans by taking them at their word. Offer money on the condition that District Planning is made a reality. Such leverage involves taking a position in Kenyan politics, supporting the devolvers against the centralisers, but if ODM does not like such an involvement it had better just hand over money and have done with it, as the involvement is implicit in any attempt to influence the use of aid. Of course supporting District Planning means British projects will not be identifiable enough to have plaques attached, the aid will not help British exporters much in the short run, and much of the expenditure may be recurrent. These should be unimportant considerations anyway. There is a deep ambivalence within ODM about such things. For example, ODM is understood to

Aid in Action

139

regard the health centre being built at Kwale as rather elaborate and to consider that more staff and transport might have been a better use of funds. On the other hand they appear somewhat alarmed by the fact that 75 per cent of SRDP expenditure was recurrent. You cannot have it both ways.

Notes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

15. 16.

17. 18. 19. 20.

21.

ILO, op. cit., p. 10. See Chapter Two. Development Plan for 1974–78, Nairobi, Government Printer, 1974, p. 109. See James R.Sheffield (ed.), Education, Employment and Rural Development: Report of the Kericho Conference, Nairobi, East African Publishing House, 1967. Rasmus Rasmusson, Kenyan Rural Development and Aid, Uppsala, Nordiska Afrika Institutet, 1972, p. 10. J.W.Leach, ‘The Kenyan Special Rural Development Programme’, Journal of Administration Overseas, Vol. XIII, No. 2, April 1974, p. 358. IDS, ‘Pilot Projects in Rural Areas’, Nairobi. Subsequently it was decided to confine the programme to the six first-phase areas. Leach, op. cit., p. 359. Ibid., p. 360. The Ministry was later joined with the Treasury and both became distinct departments of a single Ministry of Finance and Planning. The final line-up was: Migori (Nyanza Province), SIDA/FAO; Vihiga (Western Province), USA; Mbere (Eastern Province), Norway; Kapenguria (Rift Valley Province), Netherlands; Kwale (Coast Province), UK; Tetu (Central Province), Kenya government. ILO, op. cit., pp. 576–7. A polar case was that of USAID. Having once bought the project they threw themselves into it with typical energy. In Washington, Vihiga SRDP came to be regarded as a showpiece—a pilot project of general importance for future USAID policy and, given the competitive element introduced by each donor adopting an area, a chance for USAID to show what it could do compared to other agencies. USAID’s Kenya representative had to reconcile this intense interest in Washington with a good deal of indifference shown in Nairobi. IDS, ‘An Overall Evaluation of SRDP’, Occasional Paper No. 8, Nairobi, 1973. This was to some extent a political necessity. The pro-SRDP group had to have something to show high-level Kenya government sceptics quickly; aid donors had to have ‘results’ to justify untied aid and get good marks in the interdonor competition implicit in the arrangements; the local people whose cooperation had been elicited with promises of development also had to be kept agreeable. IDS, op. cit., pp. 27, 920. Leach, op. cit., p. 361. For a description of the management techniques and how they came to be adopted following IDS research, largely in Mbere, see Robert Chambers, Managing Rural Development, Uppsala, Scandinavian Institute of African Studies, 1974. We do not accept this point entirely. In our view, recurrent expenditure and capital formation can easily be the same thing. The fact that much money went to increase the routine extension activities of various ministries simply illustrates that these were undercapitalised for the task of reaching the rural masses. It may be true, however, that too much money went into the social services instead of more directly productive activities. ‘For some discussion of the tribes’ social structures and the impact of immigration, see G.Mutiso in IDS, op. cit, pp. B7-B11, B23-B28.

140 Aid and Inequality in Kenya: British Development Assistance to Kenya 22. In some areas donor representatives took an active role in project implementation. This was not so in Kwale. 23. As in the British colonial administration District Officers are subordinate to a District Commissioner who is subordinate to the Provincial Commissioner. All these officers are attached to the Office of the President. 24. In SRDP Migori for example, where a ‘Master Farmer’ scheme was proposed, of concentrating inputs on the ‘best’ farmers to make them into models for the rest. An expatriate team there made implementation possible. 25. See N.Roling, F.Chege, and J.Ashcroft, ‘Rapid Development for Kenya’s Small Farms’, IDS Discussion Paper 173, Nairobi, 1973, for a description of techniques employed in Tetu, and Schonherr and Mbugua, ‘New Extension Methods to Speed up Diffusion of Agricultural Innovations’, IDS Discussion Paper 200, which analyses extension techniques in terms of innovation theory. 26. Chambers, op. cit., p. 117. 27. Ibid., p. 142. 28. In Kwale the data collection still failed to keep pace with requests for more data. While the project adviser was requesting extension of the soil survey, none of the local agricultural officers had seen the original, and it was not mentioned in the Kwale plan. 29. Chambers, op. cit., p. 144. 30. See R.M.Mbithi, ‘Innovation in Rural Development’, IDS Discussion Paper No. 158, Nairobi, 1972. 31. Leach, op. cit., p. 361. 32. The dangers are greater in the case of expatriates who often do not know the political constraints. And as their advantage over locals is likely to be in technical expertise they are likely to employ it even when inappropriate. 33. IDS, op. cit., p. D37. 34. See Chapter Four, p. 94. 35. Development Plan 1974–1978, pp. 111–13. 36. For discussion of these points see C.Trapman, Change in Administrative Structures, ODI, London, 1974. 37. Another, perhaps better justification for continuing this policy is that there is some doubt whether the Kenyans will spend all the funds now committed.

6 THE INFLUENCE OF AID

In the case studies of the last chapter we have tried to illuminate some of British aid’s micro effects. We approach a tracing of the effects of aid in this chapter as follows: first we shall discuss the implications of fungibility and project-tying; then we look at the macro-level effects of aid on the Kenyan political economy. It will be difficult to single out the peculiar effects of British aid in this, except where particular British aid flows have characteristics which enable them to be distinguished. As official aid is given between governments, the recipient government is the interface between the donors and Kenyan society. In assessing the influence of aid, therefore, we focus on the Kenya government and move from the particular to the general. We shall assess aid’s impact firstly on public finance in a narrow sense, then on other government policies, then on the style and pattern of government administration. Then we shall have to say something about aid’s putative effects on the political nature of the Kenya government. In these discussions we shall distinguish capital aid and technical assistance on an ad hoc basis, where the distinction is relevant. Finally we shall give our conclusions based on this and the previous chapter about the influence of aid on income levels and income equality in Kenya.

Fungibility Fungibility has sometimes been seen as providing recipients with the opportunity to use aid to finance projects that do not appeal to donors. So funds negotiated to build a hospital might release domestic resources for the construction of a presidential palace. The shunting of funds between discrete projects like that, however, is only one instance of a more general problem. After aid a recipient may be able to increase the salaries of civil servants or to cut taxes while maintaining an expenditure programme. Fungibility embraces consumption as well as investment, private as well as public spending. Where there is more than one aid donor the possibilities are multiplied. British aid for a project in Kenya might displace the funds of another donor whose disbursements to Kenya may fall in consequence. So, for example, HMG loans to Mumias may haye led to there being different West German aid projects elsewhere in Africa, or may perhaps simply have led to a swelling of the West German foreign exchange reserves. Faced with such vertiginous prospects it is no wonder that many donors avert their gaze. They continue to tie aid, taking it as an article of faith that ‘their’ aid finances ‘their’ projects and, as project appraisal is a large part of the business of an aid agency, they continue to evaluate the projects. Such evaluation can have useful effects even given fungibility. It might improve the micro-level effects of aid, it might have value in demonstrating rational methods of planning and appraisal and inducing the recipient to adopt them, and it might serve a cosmetic purpose with public opinion in the donor country. We shall return to the question of project-tying.

142 Aid and Inequality in Kenya: British Development Assistance to Kenya Apart from being tied to end-use, aid can be tied to purchase of goods from specified sources, usually the donor country. Fungibility, in this context sometimes known as ‘switching’, then means that the aid can be spent without increasing imports from the donor country by the full amount of the tied aid. If such imports are increased by the amount of the tied aid, unless the marginal propensity to import from the donor is one, the presumption is that it was not fully fungible and was consequently worth less than its nominal equivalent in free foreign exchange.1 Here we are concerned with fungibility’s effects on the composition of total spending rather than on the source of imports, hence with ‘internal’ rather than ‘external’ fungibility. We should, however, clarify the relation between the two. If aid is extremely fungible, although procurementtied it will lead to additional imports from the donor only to the extent of the recipient’s normal marginal propensity to import from that source. For it to be externally fungible it must be internally fungible too, but the reverse is not the case. Suppose aid is procurement-tied and tied to a project that in its absence would have taken place anyway, using other resources. The project, if unaided, might or might not have used imports from the donor country. If not, the aid increases the donor’s exports by substituting them for those of a competitor, However, it releases to other uses the foreign exchange that would have been used to import for the project. It releases resources equal to the cost of the goods it has replaced, rather than to its own nominal value. As these goods were originally preferred, the aid may well be worth less than its nominal value but, for what it is worth, it has swelled general resources. It was therefore internally but not externally fungible. If the aid is tied to a project that would have used donor imports anyway, it is both externally and internally fungible, If, however, aid is tied to a project that would not have taken place in the absence of aid, the project is additional and clearly imports for it are additional too. You cannot have external without internal fungibility. Therefore internal fungibility is necessary but not sufficient for external fungibility. From now on we consider the former. There is a temptation to say that, in the presence of fungibility, an aid disbursement does not finance what we might call the target project, but a marginal expenditure—that which would not occur without the aid. This formulation can however lead to paradox.2 In a sense every pound of anyone’s income always finances the marginal expenditure, because (tautologically) it is that expenditure which would not be undertaken if any pound were withdrawn. To put the responsibility for the Kenya government’s marginal expenditure (or revenue-releasing) decisions in each financial year at the door of British aid would be unreasonable. That presupposes the total of all other aid is fixed and invariant to the British aid decision, and can be known in advance. The essence of fungibility is that where it applies it does not always make much sense to say any particular aid tranche financed any particular project.3 Those who would use a marginal expenditure formulation are in any case lost as soon as account is taken of time. The reaction of the administrator and accountant is to divide up time into financial years. But this year’s marginal expenditure may be promoted to the middle of the order next year so someone ‘really’ financing it this year by courtesy of shunting, is ‘really’ financing not it but next year’s marginal project. And so on ad infinitum. Can, for example, British finance for the Nairobi-Mombasa road really be said to have financed any one of every project of lower priority since, and any one of all those lower-priority projects still to come?

The Influence of Aid

143

An aid disbursement causes a change in the time-stream of receipts and expenditures of the recipient government. The pure no-fungibility case is the one where the expenditure streams with and without aid would differ only by the addition to the aided stream of the ‘target’ project, The simple-shunting case is the one where another project,4 not the target one, is added to the aided stream compared to the hypothetical unaided stream. Both are unlikely special cases. A number of projects may appear and disappear as a result of aid because of the repercussions on government spending and resource mobilisation and the actions of other donors. Or there may be no change at all in the inventory of projects appearing in the time-stream; the same projects may occur simply displaced in time. Policy makers cannot be concerned, however, with the infinite ramifications of their decisions and have to confine themselves to proximate effects. The practical question which emerges is under what conditions will project-tying of aid make its effects more predictable and controllable? We are also concerned with whether past project-tying has made British aid’s effects distinguishable from those of aid in general. A necessary condition for project-tying to be effective is that expenditure on the project would not have been undertaken anyway by the recipient government. This is not however a sufficient condition because of the presence of other donors who may have supported the project. If the target project were favoured by other donors, the donor could not even be sure he was adding it to the expenditure stream. And then additions to the expenditure stream would depend not simply on the preferences of the recipient government but also on those of other donors who had some say in what, if anything, would be done with their ‘displaced’ funds. That sort of fungibility would not matter so much if donors cooperated and had compatible priorities with each other. Even without other donors there may be problems not usually considered under the heading of fungibility. Even if project-tying secures the genuine addition (or a considerable advancement in time) of the target project to the expenditure-stream, it could thereby give rise to other effects which might be neither predictable nor desirable. There is no reason to suppose that the priorities in projects of a recipient government remain unchanged whichever projects are undertaken, so to say, out of order. Projects can compete by sector or by geographical area. Push through a project in a deserving sector in a poor area and you may demote another project in the same sector or the same area. There really is no way the donor can be confident that there will be no ramifications of this kind.5 If the project would have been undertaken anyway and aid swelled the recipient government’s revenue for other purposes, that is one type of failure which we call shunting; if the project would not have been done, but with its execution sets up other changes in government policy of which the donor disapproves, it is a second type of failure which we call shuffling.6 Now we can review the conditions in which project-tying is said to be effective. Fungibility is said to be reduced or eliminated if the recipient is not aware of it. Unawareness does not eliminate shunting. It means the recipient Treasury will be surprised at finding ‘free’ finance was greater than it expected and will consequently be slower to put through its own unaided projects. That eliminates crude, conscious shunting between projects but means either more private sector activity is permitted or the recipient’s foreign exchange reserves are higher, at least until the Treasury wakes up. As domestic borrowing or credit creation by government tends to be residual it will probably be reduced.

144 Aid and Inequality in Kenya: British Development Assistance to Kenya Unawareness does not eliminate shuffling either. Governments do not have to be aware of fungibility in order to, for example, change their mind about preferences when faced with new constraints. Lack of preference-ordering of projects is said to reduce or eliminate fungibility. As long as there was a good probability of the donor’s target project being carried out by whatever random selection method would have been used, this does not eliminate shunting, although project-tying can then be seen as a sort of insurance which may be useful in replacing the probability that a project will take place with a certainty. This situation might seem to rule out shuffling; if the recipient does not mind much what is done, he won’t change his lack of mind because something has been done. If, however, there is a need to balance competing claims for resources in different ministries and in different regions, while there was no corporate set of priorities initially (conditions which seem to be true in Kenya), the execution of certain projects may set up political pressures for some other projects to be carried out before their alternatives. Therefore shuffling is still possible. Fungibility, it is argued, can be exorcised by binding resource constraints on the recipient. So, for example, if the recipient has no domestic funds at all, or all of those he has are pre-empted as counterpart funds for projects promoted by donors, he cannot shunt or shuffle finances to advantage. That is true. This position is however entirely theoretical as far as Kenya is concerned, and if it applied anywhere, did so to only very small territories for a period immediately after independence. A more interesting variation of this argument, however, is to point to possible administrative and skill constraints on the part of the recipient. Even if he has foreign exchange reserves, it may be that his limited number of planners and administrators are fully occupied once he accepts aid and cannot find time to work up the other projects which donors would not approve, Although that obviates project-shunting, the result is likely to be increased government consumption spending, or reduced resource mobilisation by government. Another possibility, observable in Kenya, is for funds to be on-lent to parastatals or for equity to be taken in private enterprise so that central government expertise is not strained. This certainly imposes some limits on fungibility in that it prevents the use of additional resources in administration-intensive activities. If the project priorities of the donor and recipient are very different, the donor gets his way to a certain extent in that he influences the recipient government investment programmes. He does not, however, limit recipient government consumption or private sector spending, which he may or may not prefer to the public investment he has squeezed out. If he does not prefer it, he might improve matters by giving technical assistance. It is only sensible for the recipient government to accept aid, when constrained in this way, if it prefers a deformed investment programme with more consumption to its original investment programme with less consumption. In the real world, however, and certainly in Kenya, there is often no one to take that sort of decision and aid offers are taken up willy-nilly. If the donor has tied aid to projects that would have been carried out anyway, even with an administrative constraint, this is a classic shunting case. There is increased consumption or private sector investment. Provision of technical assistance would not obviate the shunting. It would merely enable the recipient to shunt resources into public sector investment as well as, or instead of, consumption or private investment. This analysis has tended to the conclusion that project-tying of aid can only influence the macro-economic allocation of resources if the aid is tied to a project that would not

The Influence of Aid

145

have been carried out in the absence of the aid. Neither indifference nor ignorance nor shortage of skilled personnel on the part of the recipient necessarily makes tying effective. It possibly means that more consumption, not different investment takes place. If the recipient’s indifference is considerable, project-tying can be a form of insurance. Even then, however, compensating resource shifts, what we have called shuffling, are possible. The only qualification to this conclusion is based on the vagueness of the notion of a project, which also means that our division of aid’s effects into the macro and the micro is a convenient classification, not a complete description of reality. In addition, the vagueness of ‘project’ means aid can be partly fungible rather than simply fungible or otherwise. A road project, for example, which a recipient government was going to undertake in the absence of aid might have been for a single lane of gravel. The donor may use leverage to insist on a tarmac dual carriageway. In that case only part of the aid was fungible, enough to finance the gravel; the rest was not fungible. Hence micro-effects can become so large as to qualify as macroeffects. In this example (not appropriate to Kenya now, although it might have been ten years ago) the donor has increased the proportion of the recipient’s public expenditure going to road-building. It could, of course, work in reverse; the donor by imposing economies on projects to which his aid is tied could reduce spending in that area and release more resources than the nominal amount of his aid. Project-tying, therefore, is not always pointless. The donor may want to ensure that the recipient undertakes certain projects which he suspects may be marginal in the eyes of the recipient. And he may make the judgement that the chances, and magnitude, of other effects arising from shuffling are relatively small. Perhaps most important, he may think microeffects reaching the status of macro-effects are possible. It is true, nonetheless, that these reasons are probably inadequate to explain the great popularity of project-tying among donors. Apart from the other reasons for it already noted,7 which have some developmental legitimacy, there are three other reasons less creditable. One is what has been called the ‘strange vanity’ of aid agencies that there should be monuments to their efforts (although this cannot be entirely separated from the donor public opinion argument); a second is the organisational interests of the aid agency itself—Parkinsonian growth is facilitated by the requirements of manpower and expertise that project evaluations entail; a third is that project-tying greatly enhances the effectiveness of procurement-tying so that there can be strong commercial reasons for it. In case this should be thought too cynical, one observation is apposite. Whatever the scope for the shunting of aid funds as a whole by the recipient—and in the case of Kenya we shall argue there is not great fungibility between Kenya funds and aid en bloc—the aid of any single donor is much more fungible. We should say that in the case of Kenya the views of a single donor in influencing Kenyan resource allocation are not worth much unless shared by other donors, when they can become very powerful. If the point of project-tying were seriously to influence resource allocation, this fact would put a very high premium on donor co-operation. Such co-operation is emerging in some areas and is certainly more conspicuous than it used to be. But donor competition is still common and it reflects badly on the developmental pretensions of project-tying.

146 Aid and Inequality in Kenya: British Development Assistance to Kenya

Public Finance The Kenya government’s domestic sources of finance are substitutes for the aid it receives from abroad. For any given level of expenditure the government can obtain the necessary resources by various combinations of aid, foreign and domestic borrowing, taxation, and charges for the services of the state. The revenue-raising and expenditure decisions are not of course independent. Expenditure plans partially determine what efforts will be made to raise revenue, and the constraints on revenueraising influence expenditure plans. Attempts to formalise this interaction have been made by specifying a ‘utility function’ for governments with the usual diminishing marginal utilities for each policy objective. The government can then be thought of as a rational agent seeking to maximise its own utility subject to the constraints of its budget. As a literal description of what happens in Kenya (and indeed in most countries) that is not accurate. Expenditure decisions are in fact the outcome of a bargaining process between individual ministries and the Treasury and are partly determined by successes and failures in Treasury control and by factors outside the control of the bureaucracy altogether. Agents who can rock the boat include the President and perhaps a few other political heavyweights in a position to insist that certain things are done, the Kenya public where its desires are so strong and general as to threaten political stability if not met (the proliferation of schools is an example), and aid donors. Total expenditure is nothing but the sum of all the individual decisions made under these pressures. How far the utility function approach can serve as a useful metaphor or model we shall not go into here, but as long as there is some interaction between government objectives in revenue-raising and expenditure it is reasonable to suppose that aid tends to increase spending, while acting as a partial substitute for domestic resources. We cannot go from there with any confidence to give a quantitative estimate of what proportion of each aid tranche to Kenya has brought about extra government spending and what proportion has gone in reduced domestic resource mobilisation. An interesting attempt at such measurement has been made by Dr Peter Heller, who computed that for a sample of Anglophone African countries, bilateral aid flows at the margin increased expenditure by only 18 per cent of the aid; 55 per cent went in reduced domestic borrowing and 26 per cent in reduced taxation.8 Heller’s model, which relies on the utility function approach, is very interesting, but not much confidence can be put in the actual figures it yields because of the large standard errors of the estimates.9 The figures, anyway, would be averages for all the countries in the sample,10 and would not necessarily apply to Kenya alone. However the financial sector in Kenya is at least as developed as that in most of the other countries, so a priori one would not expect domestic borrowing to be less important in Kenya than for Heller’s sample as a whole. Table 17 shows the importance of domestic borrowing in financing Kenya government expenditure. The other significant domestic source of finance, taxation, is more important. In the period 1969–74, for example, taxation accounted for an average of 92.7 per cent of’ recurrent revenue’. In the early post-independence years, except for 1965/66, the government easily covered the borrowing requirement with long-term borrowing. In fact in 1966/67 and 1967/68 long-term borrowing exceeded the deficit and government cash balances rose.11 Short-term borrowing has been used in subsequent years, mainly through the issue of Treasury Bills.

The Influence of Aid

147

In 1969/70 and 1970/71, the government borrowed directly from the central bank. Since then, while short-term borrowing has remained a feature of the government’s finance, longterm borrowing has kept pace with the total borrowing requirement. It was 62 per cent of the requirement in 1971/72, 80 per cent in 1972/73, and 79 percent in 1973/74.12 It seems clear that there was scope for more domestic borrowing by the government than occurred in the ten years after independence. That is not because the Kenyan banks by-andlarge had ample liquid balances so increased borrowing would not have upset financial markets. That is true but not the point. We have to consider what the effect would have been on real markets and there we believe both that the government could have obtained more resources from the private sector by long-term borrowing and that it could have afforded to borrow more short-term and create credit. Furthermore we believe that with less capital aid but the same technical assistance it would have done so. The effect of a larger government deficit partly financed by short-term borrowing and credit creation, even if it mobilised some resources that would otherwise have been unemployed, would eventually have been reflected in the balance of payments. In most years since independence, Kenya has had a modest surplus on the balance of payments, and foreign exchange reserves have risen from £K42m. in 1966 to £K82m. at the end of 1973. The increase in reserves represented an average annual growth rate of 10 per cent, comfortably faster than that of imports at current prices. Import price inflation began to exceed 10 per cent only in 1972. In 1970 the reserves were enough for four and a half months’ imports compared with three months’ in 1966. For most of the post-independence period, therefore, the government could have run the economy at nearer full throttle than it did, and reductions in aid would not have needed to be matched pound-for-pound by expenditure cuts. This was in fact accepted in the 1974–78 Development Plan—written before the oil crisis developed which pointed out that the government had drawn little on the banking system for short-term credit but stated it intended to do so more in future.13 There are, of course, dangers to such a policy. Fine tuning of the level of demand is no more possible in a country like Kenya than in a country like Britain, and we cannot say how much extra government expenditure (or unrequited aid reductions) would have been consistent with a tolerable net drain on the balance of payments of say £K25–30m. over seven years, without accelerating inflation. Clearly, however, some slack existed. The government could also have obtained more resources without inflation, either by long-term borrowing or by the combination of shortterm borrowing with quantitative credit controls on the banks, a policy it adopted in 1971 in response to a temporary balance of payments deterioration. Either course represents the channelling of resources from the private sector with no necessary increase in demand. Long-term borrowing does so through the voluntary saving and lending of the public. Borrowing from the banks and imposing credit controls does it by a form of implicit taxation on the banks, forcing them to lend to the government rather than other customers. Either policy could be disadvantageous to the banks; the former might well reduce their deposits, as bank deposits are one of the few substitute assets for government securities in a country like Kenya with rudimentary capital markets, and the latter would reduce their profitability directly.

−6912

8059

Recurrent surplus

Net development expenditurea

14525

6122

Total external financec

Domestic borrowingd 7167

11313

3474 6893

9733

1902

7831

−16626

7356

−9270

11545

8089

974

7115

−19634

10177

−9457

13645

+4188

80515

84703

1968/69

10992

12215

1440

10774

−23137

11030

−12107

18898

+6791

91136

97927

K£’000 1969/70

20680

11765

824

10941

−32445

15757

−16688

29354

+12666

111317

123983

1970/71

25186

13283

1794

11489

−38469

12087

−26383

39341

+12958

128670

141628

1971/72

26623

25178

521

24657

−51801

17053

−34748

44169

+9421

139578

148999

1972/73

24073

24357

4357

20000

−48430

−48430

69470

+21040

166203

187243

1973/74

Source: Kenya government, Economic Survey 1974, p. 165.

b

a

Net of development project earnings and other receipts (almost negligible). Largely purchase of equity, and loans to private and public corporations. Loans to public corporations were the most important in each year except 1970/71 when equity purchases were higher. These loans average 74 per cent of total investment expenditure. c Almost entirely supplied on concessionary terms. Grants have been for both recurrent and development account with the latter becoming more important. Loans have been mostly for development account; recurrent account loans have been nil since 1970/71 and were negligible after 1967/68. d Long- and short-term borrowing and changes in government’s cash balances. This is the difference between the total deficit and total external finance.

5098

External grants to development

7839

−18480

−14971

9427

6614

−11866

11356

+2086

−2533 9333

74991

77077

1967/68

68529

65996

1966/67

5676

External loans

Financing of the deficit

Deficit

Investment expenditureb

−20647

63267

Recurrent expenditure

Total deficit

56355

Recurrent revenue

1965/66

Table 17: Kenya government expenditure and finance

The Influence of Aid

149

There does not seem to be any reason why the Kenya government should have been concerned with moderate falls in banks’ profits, as they were not likely to quit the country, so all it had to consider was from whom it would have diverted resources by increasing its own borrowing. In 1967 Africans and African farm enterprises received only 2.6 per cent of total commercial bank loans and advances.14 Recent Kenya central bank figures give the following breakdown by province of commercial bank advances, loans and discounts: Table 18: Distribution of commercial bank lending Nyanza 5.0 Eastern Province 1.9

(percentages) Western Province Rift Valley 0.6 Coast 14.8

7.1 North-eastern Province 0.03

Central Province 5.7 Nairobi 65.3

Not all large-scale business or expatriate business is in Nairobi, of course, but there is no doubt that the great preponderance of lending in Nairobi, followed a long way behind by the Coast and Rift Valley, reflects a continued bias in favour of established modern sector business. There is nothing sinister in this and the commercial common-sense reasons are plain enough. It is still an issue of considerable political sensitivity in Kenya where politicians and district administrators have claimed that while the geographical distribution of bank deposits is heavily skewed it is not quite so uneven as that for advances, so the banks are in effect channelling funds from rural to urban areas. African businessmen also claim that British banks still discriminate against them.15 Of course, because the average bank borrower in the private sector is a businessman, often urban and often expatriate, it does not follow that the marginal borrower is. A public borrowing squeeze on bank lending could hit the small upcountry borrower rather than more favoured customers. It is doubtful, however, how many such there are. We are probably safe in concluding that any big expansion of public borrowing would have largely hit modern sector business. Exchange control regulations now limit local bank overdrafts to a proportion of a company’s foreign long-term liability. Where more than 60 per cent of the equity is foreign-owned, overdrafts are limited to 20 per cent of that liability; for foreign holdings of 51–60 per cent, a sum equal to 40 per cent of the foreign equity can be borrowed locally; and if 50 per cent or less is held by foreigners 60 per cent of their holding can be borrowed. Historically, foreign private investment in Kenya has made great use of local savings. B.Herman, in an examination of foreign investment, computed the ‘degree of use’ of local savings as private plus public local debt divided by private foreign equity.16 All the enterprises he examined—81 set up over 11 years—had a ratio of local funds to foreign equity of more than one.17 For British firms the ratio was a startling 4.4, much higher than that for firms of other countries of origin. The figures are somewhat misleading in not taking account of inflation or, more importantly, the companies’ retained earnings as part of the foreign equity. Considering retained earnings would reduce the apparent dependence of British firms on local capital both absolutely and relatively, because as they had on average been in existence longer than the firms of other foreign investors, their total retained earnings would be greater. Nevertheless, Herman’s hypotheses that foreign investors extensively use local savings and that British investors find it particularly easy to do so because the banks are British, probably have a good deal of truth.18

150 Aid and Inequality in Kenya: British Development Assistance to Kenya We are not arguing that the Kenyan government would or should have followed a policy of forced borrowing and reckless credit expansion. It seems clear, however, that while building up foreign exchange reserves may have been a good policy, given the aid flows into Kenya, if those flows had been significantly lower it is doubtful whether the build-up would have been preferred to various forms of government expenditure. More government borrowing would therefore have resulted if aid had been lower. If the government attempted to borrow without inflation the burden would have fallen on Kenya’s banks and/or modern sector business, significantly involving British interests. One effect of aid has therefore been to maintain liquidity in the economy and so help those interests. That could have been good for private investment, but it could also mean that aid has been a substitute for private capital flows. Even if it did increase private investment, by artificially cheapening capital for larger firms, aid could have distorted the investment in a labour-saving direction, although it is doubtful whether the size of such an effect was significant, even if it occurred. An examination of taxation leads to similar conclusions. Kenyan tax revenues have not been particularly buoyant and without periodic changes to tax rates, allowances, and bases, revenue would have grown more slowly than GDP since independence.19 Owing to the general rise in money incomes, the proportion of direct tax receipts in total revenue from taxation rose from 36 per cent in 1963/64 to a peak of 47 per cent in 1972/73 and then declined to 37 per cent following the introduction of a sales tax on both imports and local manufactured goods. In its first full year of collection the sales tax netted about a third of indirect tax receipts and a fifth of all tax receipts and was responsible for the recovery in the government’s recurrent surplus (see Table 17). If the government had been faced with the necessity of raising more revenue earlier it seems probable that it would have been done by the same technique, namely an increase in indirect taxation. There is scope for a tightening up of assessment and collection of income tax but that is expensive in skilled manpower (although we have little doubt it is worth doing in Kenya).20 In any case the Kenyan direct tax structure is not very progressive, although reforms in 1973 made it more so, and it may be that this accords with the government philosophy towards development. It is noticeable that the current Development Plan, for example, defines progressivity as a situation in which the rich pay more tax than the poor!21 It seems likely, therefore, that an increase in indirect taxation would have resulted from aid reductions, probably involving, as subsequently happened, a sales tax on luxury goods,22 and perhaps increased excise duties. The burden of a sales tax would fall on the consumers and producers of locally manufactured goods, again the modern, largely expatriate, business sector. Beer, cigarettes and sugar yield most excise tax revenue. The first two are produced by monopolistic firms, one a wholly-owned subsidiary of a British company, the other a Kenyan company in which there are substantial British equity interests. As their products are price-inelastic, the (mostly African) consumers would have paid most of any duty increases. The commodity yielding most import duty since independence is fuel, which is an important part of the costs of any modern firm. Altogether we can say that the provision of aid has alleviated a competition for domestic Kenyan resources of capital, and perhaps manpower, that would have been carried on between the government, faced with political demands, and the modern sector of the economy, in which expatriate business has a dominant interest. It is very difficult to know

The Influence of Aid

151

what the extent of competition would have been and what its effects would have been on tax rates, interest rates, the ‘diffused difficulty of borrowing’, and the level of foreign investment. There is, however, no reason to suppose that aid has permitted an increase in the average propensity to consume of the Kenyan economy. Resources for which the government might have competed have been released to the private sector but they could well have gone into private investment. And while we have argued that aid must have reduced government resource mobilisation we have not said what its effect has been on the balance of public sector consumption and investment within total government expenditure. To government expenditure we now turn. The Kenya government divides its expenditure into recurrent and development. This distinction is not exactly the same as that between consumption and investment. The division is based on the following principles: Recurrent Estimates include estimates of personnel costs, other recurrent expenditures, small items of equipment, including vehicles, maintenance and cash transfers; Development Estimates include estimates of expenditure for major equipment items, buildings, construction works and land; investments, grants and loans to other sectors and purchases of equity. All equipment items relating to new projects and financed partly by external grants or loans are included in the Development Estimates. Similarly, any expenditures on personnel or other items, normally regarded as Recurrent, which are financed from external sources are included in the Development Estimates.24

The classification depends partly, therefore, on whether a particular project is new or not. Salaries and vehicles for a team to carry out a tsetse fly control exercise would be development if the team were new and financed by aid, recurrent if it were already in existence. Many recurrent expenditures are not consumption, but capital replacement or maintenance, or working capital for existing projects. Some development expenditures are not investment but ‘placement’—the acquisition by government of claims on existing capital assets. Even if development equalled investment and recurrent equalled consumption, however, the fact would be largely without normative significance. Without getting involved in semantics we can note that much expenditure that would be classified as consumption in a developed country can raise future output more than expenditure that would be classed as investment. Education and health expenditures among an illiterate population suffering from endemic diseases could be more productive than a four-lane highway or a large factory doomed to run at 20 per cent of capacity. It is necessary to make these familiar points because in Kenya both the Treasury and donors have tended to venerate ‘investment’ above ‘consumption’, and by illicit extension ‘development’ above ‘recurrent’ spending. No doubt, partly to blame are the emotive names applied to these categories which the accountants find convenient. If rational men wish to retain the categories it might be better to rename them the lumpy budget and the smooth budget. At any rate it is not the case in Kenya that a chief effect of aid has been to enable the government to expand its recurrent expenditure or consumption. Table 17 shows recurrent expenditure at current prices rose by less than three times between 1965 and 1974, an average annual growth rate of 13 per cent. Development expenditures, excluding investment

152 Aid and Inequality in Kenya: British Development Assistance to Kenya expenditure, went up over eight and a half times in the same period, a growth rate of almost 31 per cent a year. In Chapter Two we noted that while public sector consumption remained a constant proportion of monetary GDP between 1964 and 1973 (about 20 per cent of resources available), public investment greatly increased its share. All gross investment rose from 16 per cent of GDP in 1964 to 29 per cent in 1973 at current prices, and the public sector’s share in fixed capital formation went up from 28 per cent to 46 per cent. It is possible, logically, that the figures might have shown even more dramatic increases without aid but that is implausible. The reverse is probably true: aid increased government expenditure absolutely, although by less—perhaps much less—than the amount of the aid, and it raised the proportion of government expenditure that could be classified as either development or investment. This latter effect may have been so strong in fact that aid actually squeezed government consumption and tended to reduce it absolutely. That was certainly the finding of Heller’s econometric analysis.25 Although we do not regard that result as self-sufficient, for the reasons given,26 there are ready explanations for it in the case of Kenya. First is the tendency for donors to provide loan finance for part but not all of a project. The Kenya Treasury adopts a conservative approach to life. That was the view of foreign donor personnel and Kenyan civil servants in operational ministries, and as noted in Chapter Four, was one reason why the Swedes give advance disbursements.27 We have noted two cases of the Treasury’s reluctance to authorise recurrent expenditure even though loans had been promised to cover it.28 In other words the Treasury does not regard loan funds as fully fungible and is reluctant to use them to finance consumption. Their acceptance means that an attempt is made to increase investment and where this is impossible, resource mobilisation is reduced. Loans tend to be matched to investment projects but they rarely fully finance them, so that Kenyan counterpart funds are drawn in too. Second, donors exert a good deal of pressure on the Kenya government to show it is doing its best in resource mobilisation and the main concrete evidence of this is taken to be the surplus on recurrent account. So the Kenyans believe, with reason, that they can generate more aid by having a larger recurrent surplus to provide part of the funding for more development projects and then appealing to donors to finance the balance. Aid has to be obtained that way because the donors frequently agree to support only part of discrete development projects. This fact fundamentally alters the return the Kenyan money earns in marginal investment projects because the concessionary element of aid becomes one of the benefits of such projects. It is no wonder then if Kenyan funds are drawn into such projects by the existence of aid. Of course they do not have to be drawn from government consumption but could come from other projects or increased resource mobilisation. It is our judgement that the funds were diverted from government consumption in the sense that this would certainly have been a higher proportion of government expenditure and might conceivably have been absolutely greater without aid. If this is so, one thing follows immediately: donors, as a bloc, were financing the marginal projects and there was only limited fungibility of aid. If the donors had only been providing aid conditional on the execution of projects that were top priority for the Kenyans (believing they were financing them), Kenyan money would have been released and the Kenyans would have had an unconstrained choice between marginal investment

The Influence of Aid

153

projects and consumption. There would thus have been no tendency for money to be drawn from consumption into investment by aid. If, however, the Kenyans carried out their own top priority projects and donors offered to partly fund the more marginal ones, that would change the attractiveness of the marginal projects relative to other ways of using resources, and pressure on government consumption would result. We believe that was the case and that aid as a whole was not much shunted between investment projects. But the aid of any single donor, from the mid-1960s on as the number of donors increased, was in a sense fungible, as most assisted projects would have been undertaken by other donors. There is a third and even more compelling reason for believing that aid increased saving and investment propensities, as conventionally defined, in the Kenyan economy. Through most of the post-independence period the Kenya administration has not had the capacity to plan, execute and run an investment programme of large projects on the scale that the supply of capital would have allowed. That is the view of all aid donors to whom we spoke, and evidence of it is Kenya’s consistent underspending of project-tied aid commitments. The public sector project list which the Kenya government presented to donors at their Consultative Group meeting in Paris in 1974 showed that the Kenyans themselves thought technical assistance was ‘required’ for a high proportion of the projects they proposed. When road projects, which the Ministry of Works can handle perfectly well, were excluded along with routine extensions to institutions like hospitals, the proportion of projects for which technical assistance was required became very high. Technical assistance has failed to close the gap between the administratively-constrained level of investment activity and what finance alone would have made possible, given that financial aid was also available. Yet expatriates have played a significant, often crucial, role in practically every major investment project in Kenya. Two, if not all three, of the projects we examined in Chapter Five were conceived by expatriates and two were largely executed by expatriates in their early stages. They are not untypical. Furthermore, not only are many large projects, especially the bigger ones, planned, prepared and largely carried out by expatriates, but as pointed out in Chapter Four29 the Kenya Treasury surrenders the selection of all but a few projects to donors. It lacks the capacity and the political authority to vet and rank projects so the sine qua non of its approval becomes the promise of aid. If no technical assistance had been forthcoming, could the Kenyan administration have arrived at decisions and invested the 50–60 per cent of the Development Budget available from Kenyan resources? We believe that some of those resources would have gone to recurrent expenditure, or other expenditures requiring less planning than large projects. There is evidence that not only has aid increased government investment relative to government consumption, but that it has done so excessively. Many donor civil servants take the view that the provisions for recurrent expenditure by some ministries in the Development Plan for 1974–78 are inadequate to service the capital in existence, let alone that implied by the Plan’s level of development expenditures. The main work in this area has again been done by Peter Heller, who, from an analysis of projects, calculated the percentage of development expenditure in each sector which would be required as an annual recurrent allocation, if projects were to run at the planned level of efficiency.30 He found that on any reasonable assumptions about capital-output ratios, aid flows, etc., the ‘fiscally consistent’ level of investment the Kenya government could undertake

154 Aid and Inequality in Kenya: British Development Assistance to Kenya if its recurrent expenditures were to be adequate to service capital was below the target investment level of the 1970–74 Development Plan. Whereas the target for government development expenditure was about 6.1 per cent of GDP, the recurrent financial capacity of the government was sufficient at the planned growth rate to finance investments of the composition laid down in the Plan at about two-thirds of that level. With planned aid inflows of 50 per cent of total investment the government could afford to invest only 3.8 per cent of GDP. Higher investment levels, unless made possible by higher-thanexpected growth rates of government revenues, entail falls in the recurrent financing of projects. The experience of SRDP31 showed that when it came to rural development the felt needs were for more recurrent expenditure. Kenya is by no means Nkrumah’s Ghana but it has some expensive aided projects with a distinct whiff of the white elephant about them—the Russian-built hospital in Kisumu, the Cyprus bins for grain storage supplied by Britain— while many an agricultural or health extension worker would do better with a bicycle, a larger petrol allowance, or rather more training. It is difficult to substantiate that statement beyond saying that we met many such in rural Kenya when researching for Chapter Five. This behaviour of boosting investment may, of course, be quite rational for the Kenya government which has to consider whether the loss of efficiency and output from any project owing to a lack of recurrent finance (e.g. the school with fewer teachers or less qualified ones) exceeds the gains from a new project subsidised by donors. Historic costs should not influence its decision. It is hard to see, however, that the action of donors is rational in insisting on new projects. And as a matter of fact we do not think the Kenya government pays enough attention to the problem. While donors, including Britain, are guilty of distorting government expenditures, it is only fair to say that they often seem to be pushing an open door and the Kenyan government has its own appetite for the grandiose. The fine modern buildings in Nairobi housing government departments and parastatals, the KANU conference centre, extensions to Nairobi airport, to Mombasa airport (financed by a Japanese loan) and plans to extend Malindi airport (so far refused assistance by at least two donors) may all be examples of this. Once involved with a project, ODM tends, in fact, to be a restraining influence on the Kenyans, who may also be trying to spend aid promised to them by inflating the expenditure on individual projects. The case of Kenyatta Hospital has been noted and there are others involving, for example, educational establishments. Unfortunately it is all rather late. The emphasis of donors on ‘projects’ and ‘investment’ offering between them more money for these things than can usefully be spent has fortified the illusion in Kenya that capital is cheap for such purposes and has fed the appetite for grand symbols of modernity which is frequently present among the elite of a new state. Having helped to establish such an atmosphere the more responsible donors, who include ODM, have to wrestle with its manifestations in individual projects. Others, including notoriously the IBRD, have sometimes accepted ludicrously high specifications for roads etc., and financed them regardless. The resistance of even the most responsible donor can be sapped by other considerations. It is easier to resist a prestige building or road requiring local labour and materials than airport or television installations involving substantial export orders.

The Influence of Aid

155

Other Government Policies Clearly one area where donors might be thought to have influenced the Kenya government is via its investment programme, affecting both the sectoral and geographical distribution. This could have occurred in two ways, by the provision of experts who gave influential advice to administrators and by leverage in the giving of capital aid, even if only the negative leverage of refusing to finance what one doesn’t like. It is difficult to discern any consistent tendencies in the investment policies of the government which can be attributed to the influence of technical assistance personnel, who vary widely in competence and political leanings. We shall say more about technical assistance in discussing the impact of aid on the style of Kenyan administration and concentrate here on capital aid. We have argued above that donors appeared explicitly to finance the marginal projects in the government’s plans, which suggests fungibility may be limited. The 1974/75 Development Estimates reveal that the development projects in which donors have no involvement are the minority. Some ministries got no external aid by virtue of the kind of activities they carry out. These include the Judicial Ministry, Information and Broadcasting, Labour, Defence, and Vice President’s Office and the Office of the President.32 Together these accounted for 6.1 per cent of proposed development spending of £K88.3m. in 1974/75. Housing is an unfashionable ministry with donors and also received no aid, but accounted for another 6.8 per cent of development spending. Housing may be an exception, but the expenditure of the other unaided ministries can probably be regarded as high priority which would go ahead whether the Kenyans received aid or not. Of the expenditure of the remaining ministries, aid accounted for some 38 per cent. That is lower than in an average year, which fact will merely fortify our points. In Agriculture, aid was expected to account for 43 per cent of total expenditure. Projects with no capital aid input at all accounted for 30 per cent of the total. Capital aid was therefore involved in 70 per cent by value of the Ministry’s expenditures—the 43 per cent of finance that was aid tied up another 27 per cent of the total. In the Ministry of Works, estimates were for projects totalling £K20m., but only £K5m.—25 per cent—seemed to be wholly Kenyan-financed. Here £K6.9m. of aid tied up local funds of £K8m. In those ministries where a higher proportion of the funds were wholly disposed of by the Kenyans their use is revealing. In the Ministry of Finance and Planning, for example, some £K4.35m., out of a vote of £K4.95m., was on wholly Kenyan enterprises, but £K3.54m. was capital subscriptions, loans to public and private enterprise and other share purchases. Similarly the Ministry of Commerce and Industry received little aid towards its development budget of £K2.92m., but £K2.27m. of that was for loans or grants to other bodies, including funds for ICDC to purchase equities. These figures might seem to reinforce the view that Kenyan resources of finance, and more particularly skilled manpower and administrative cohesion, are inadequate to allow much shunting of aid funds as a whole between government investment projects. It does not quite follow from this, however, that donor preferences have dominated or determined the pattern of public sector investment in Kenya, simply because there are many donors and the Kenyans can play them off against each other. It is in the organisational interest of the highest-minded aid agency to spend money in a country once a certain sum has been allocated to it. And representatives of donor governments are conscious that they are inter alia involved in a competition with each other for commercial and political preferment by

156 Aid and Inequality in Kenya: British Development Assistance to Kenya the recipient. Donors therefore sometimes finance projects pressed on them by recipients even though they are against the better judgement of development-minded technicians in aid agencies. In the case of ODM we have already mentioned Mombasa television. Another example may have been the Naivasha-Suswa pipeline to take water from Lake Naivasha to range areas in the Rift Valley at a final cost to HMG of £465,000 (the initial estimate was £150,000). The project seemed a very marginal one, on the basis of cost-benefit analysis. Its social benefits also depended critically on the Ministry of Agriculture controlling range management strictly and inducing Masai herdsmen to prevent environmental damage by moving to rotational grazing—an outcome in which no one could have much confidence. However as the Masai were very poor, ODM authorised the project. One of the features of the scheme, however, which may or may not have been widely known in HMG, was that it benefited several large ranches owned by important people and people with important connections. No doubt the Masai would be flattered to know this was one project the Kenya Treasury submitted for tender before aid funds had been secured—the tender notices were published in the newspapers long before ODM’s authorisation date. Ironically, however, the project’s completion is now held up because ODM wants higher charges for the water than the Kenyans are willing to impose. Nevertheless, donors as a group do have considerable influence over public investment in Kenya because there are not all that many aid projects about which the Kenyans feel so strongly. For the most part donors select projects from a list that is not ranked for priority by the Kenya government. Not having their own ranking, the Kenyans allow donors to choose in order to maximise aid. The question of how to maximise aid is an interesting one, and it does not follow that the recipient necessarily succeeds in doing so by offering donors projects they like. That is likely to be true when dealing with donors like USAID and the Germans, who negotiate the amount of aid they will give at the same time as they discuss which projects they will support. In the case of donors like ODM and Sweden, however, who agree a fixed sum in advance of project discussions, a different technique may be best. If the donor has already committed, say, two-thirds of an agreed aid allocation to projects he likes, he may cavil at marginal-to-donor projects then offered. If, however, he has managed to commit nothing in the first year or two of an aid agreement he may in desperation take those marginal projects, After he has aided these he can be offered the glamorous one. The Kenyans do not play this game in a very sophisticated way, but they certainly play it.33 It seems, however, that donors as a whole have had an important influence over that part of the public investment programme which requires detailed planning. They have had little influence over politically important projects on which the Kenyans were determined, as the Kenyans had enough money to finance these and, up to a point, could obtain technical expertise at commercial terms. Donor competition is, in any case, usually sufficient for the Kenyans to obtain aid finance for such projects once their political importance becomes known. These projects are, however, a minority. Donor influence does not seem to extend over government recurrent spending, beyond a tendency to restrict it by swelling the development budget. This is indicated by the record of the Ministry of Agriculture. Half of its development expenditure is financed by aid, more than the average for Kenyan ministries, and 20 per cent of all capital aid to Kenya has gone to it. Yet agriculture’s share of total public sector spending, including recurrent

The Influence of Aid

157

spending, has fallen from 21.2 per cent in 1963/64 to 8.4 per cent in 1970/71. In the 1974– 78 Development Plan the share of recurrent expenditure going to agriculture is projected to fall, while that for education is projected to rise; tendencies that fly in the face of much that is written in the Plan. There is, of course, a strong flywheel effect operating, and much of the educational expenditure is forced by past spending. But the fall in agriculture’s share cannot be explained, as some observers have attempted to do, by the difficulty of preparing projects in the agricultural sector, a problem which would restrict development rather than recurrent spending. The motivation for these allocations is largely a matter of domestic Kenyan politics. In other words the donors get their way to some extent in the sectoral distribution of development spending, but their writ does not run so far where the recurrent budget is concerned. It is clear, however, that the influence of an individual donor, even Britain, the largest one, is negligible unless its views are shared by the other donors. If a project is attractive to other donors, or they can at least be induced to do it, it does not matter whether one donor aids it or not. There is, however, some tendency for donors to be cagey of projects which they know other donors have turned down. Quite a good way to kill a project therefore may be to say one is considering it. If one then rejects the project, the others may be shy of it. The Kenyan Treasury suspects the British of adopting this technique, but ODM denies it. We are inclined to take ODM’s word for that and to put delays down to normal bureaucratic difficulties. In any case, the technique does not always work; other donors can step in and take a project over, as the West Germans did with the Mombasa Polytechnic which ODM thought it would be aiding. (It is fair to point out that there has been as much delay in prosecuting the project since the Germans took it on as there was when it was a British ‘possible’.) We conclude, therefore, that there has been considerable, but by no means total, donor influence over the sectoral pattern of public investment in Kenya, owing largely to a lack of overall policy by the Kenyans and their readiness to accept aid offers which tie up their own resources. The influence is moderated by competition between donors and is generally insufficient to kill those few projects for which there is considerable Kenyan political pressure. As the Kenyans lack expertise in project preparation, the influence of donors has been considerable over the details of projects. There is, however, little or no evidence of donor influence over the pattern of Kenya government recurrent expenditures although we believe aid has tended to keep these down—probably too much so, if one believes they would have gone where they were needed. We could find no evidence that donors’influence had any marked effect on the geographical distribution of public investment. It is almost universally believed in Kenya that economic development has proceeded very unevenly with Nairobi and Central Province growing much faster than other areas, particularly Nyanza and Western Province. ODM appears to share this belief and to look kindly on projects in areas that are thought to lag. It chose Kwale not Tetu for its SRDP area, and its currently favoured projects of rural access roads and livestock marketing facilities are essentially rural enterprises focused on areas outside Central Province. While rapid urbanisation is a serious problem and Nairobi is growing very quickly, there is less evidence that Central Province as a whole is developing faster than many other areas of Kenya since the spurt some years ago owing to the dissemination of small-holder coffee. We approached the sales managers of a number of firms producing

158 Aid and Inequality in Kenya: British Development Assistance to Kenya mass consumption goods for a regional breakdown of their sales figures. Each subscribed to the general view that development continued to be lopsided, but no set of figures when produced bore out the belief. In nearly every case there was an enormous disparity in levels with a very high proportion of sales being in Nairobi and Central Province (which is reason enough for public policy to favour other areas) but the proportions appeared to have been roughly constant for several years.34 This in no way constituted a scientific survey, however, and we had no information on, for example, the income elasticity of demand for the goods in question. In one case British aid can be said unequivocally to have helped Western Kenya relative to other areas. The National Agricultural Research Station in Kitale, which has received British and other technical assistance on a large scale, has developed the hybrid strains of maize that have undoubtedly had a considerable effect on Kenyan maize production.35 The hybrid maize development also received CD&W funds during the period 1955–65. Being sited in Kitale, the station developed first of all hybrids suitable for Western Kenya and the Rift Valley and so increased the relative importance of those areas in maize production. The Katumani, or dry-land, maize which helps Eastern Province areas has had rather less dramatic success. As more strains are developed and introduced, it may be that other areas will catch up. Quite apart from its geographical impact, the maize breeding programme has been one of aid’s biggest successes. Some 86 per cent of hybrid maize in 1973 was on small-scale holdings, and the social returns to the expense involved (less than £1m.) are enormous. On the wider range of government policies, economic and social, we could find little evidence of aid having any effect at all. Most donors seemed well-disposed to the East African Community, for example, and both HMG and IBRD have aided it substantially. The Kenyans, on the other hand, have seemed to regard it as very much an ‘optional extra’, and if donors have been concerned to promote its well-being, they have failed. As noted, over Mumias ODM felt unable to resist Kenya’s desire for its own sugar factory despite the implications for Uganda. There has tended to be a donor consensus on the correct economic policy for Kenya, which follows the lines of successive IBRD analyses. These in turn tend to reflect the prevailing orthodoxy in development economics. So in the early years of Kenyan independence, industrialisation via import substitution seemed a reasonable policy. Now that is out of style and the accent is on rural development, encouraging manufacturing industries to export, reducing and rationalising the structure of protective tariffs, and devaluing as and when necessary. All of these things have found their way into the current Development Plan. However, as we have noted, the Plan is largely written by expatriates and, the cynical would say, for expatriates. Just as the distribution of recurrent expenditure does not reflect that of development expenditure, neither reflects the statements in the Plan all that well. The expenditure on rural development hardly matches the importance the Plan says that it has. And we have heard of negotiations to afford trade protection to firms in flat contradiction to the policy of the Plan for reducing protection.36 Foreign investment, even where it yields negligible social benefits, affords substantial pecuniary opportunities for well-placed Kenyans in a position to render it services. That makes it difficult, in the absence of a strong and coherent government policy, for foreign investment to be controlled. It is an error, however, to assume that aid agencies are merely handmaidens of every private firm

The Influence of Aid

159

from their country that wants to set up in Kenya. There is precious little contact between the British High Commission in Nairobi and most British businessmen in Kenya and practically none at all between them and the East African Development Division. In so far as British aid can be seen at all as a means to support private enterprise in Kenya, it does so in a much longer-term way than trying to influence Kenya government policies to the shortterm advantage of British firms. In fact if the current fashions in development economics, which ODM professes, were really taken to heart by the Kenya government, many British enterprises would lose protection and have a harder time of it. Sometimes donors have attempted directly to influence policy in certain sectors. One example is the IBRD and USAID support for birth control. An IDA credit of about £5.5m. in 1972/74 was given for a national family planning clinic, a health education centre, five nursetraining schools, and 27 rural health training centres. The Kenyans accept the family planning element as part of a larger health scheme, but for the most part they cannot be persuaded to give birth control the sort of priority that the North Americans think it deserves. Even references to population in the Development Plan,37 where donor opinions tend to be courted, are muted. The British have attempted to improve Kenyan educational policy. Supported by an enormous groundswell of popular feeling, education expenditures in Kenya, especially on secondary education, appear to be out of control. The demand for school places, leading, it is hoped, to paper qualifications and a lucrative modern sector job, is so intense that the government seems powerless to resist. Where it does not build secondary schools itself, villagers construct their own Harambee schools and it becomes the test of the local MP or other leader that he should obtain government finance for the schools. A secondary school with a fifth form or higher is a great status symbol, so there is also a tendency for the smallest schools to attempt to provide training right up to sixth-form level. Thus resources of teachers and other inputs are increasingly stretched and the result is declining examination pass rates and an increasing output of arts students, as the smaller schools rarely have the trained teachers or equipment to teach science adequately. Many of the classes thus created cannot be filled with pupils with reasonable qualifications. One outcome is the production of more arts specialists than can be accommodated in higher education or than can be employed. There is now an excess of Kenyan non-graduate teachers in all arts subjects except English and French. As a very important supplier of secondary school teachers under the OSAS agreement, Britain might be thought to have some influence over this. It has attempted to use that influence in a number of ways. It has deliberately run down the number of teachers supplied to Kenya, phasing out all arts teachers except those teaching English; it has restricted its teachers to approved educational establishments, keeping them out of the smaller schools.38 Its manpower planning reviews with the Kenyans have occasionally become heated as the British insist on reducing the number of teachers faster than the Kenyans would prefer. (To some extent, the battle has been beside the point as ODM has been unable to recruit the planned numbers anyway, owing to inadequate salary supplementation.) ODM also insists on projections of the number of students and the employment requirements for them when considering capital aid for higher education establishments. It did this, for example, when proposing to aid the Kenyatta University College expecting it would train science teachers which was a priority, especially as OSAS was running down. The result has been four years of delay in getting the project moving as

160 Aid and Inequality in Kenya: British Development Assistance to Kenya there is some opinion in the Kenya government that the college should be a second separate university and this opinion cannot be justified with figures satisfactory to ODM. Technical assistance was also given for the Planning Unit of the Ministry of Education in order to improve the Ministry’s planning. This pressure has been largely, although not entirely, in vain. Secondary schools have proliferated in spite of it all. A withdrawal of British teachers would not halt the process which goes on in response to political imperatives. The political nature of the expansion is shown by the fact that where schools have gone is correlated with political influence, with Central Province being specially favoured. One of the difficulties is that the consumers of education in rural Kenya do not have any well-formed expectations about the quality of the education they are getting. They are fatalistic about the fact that some schools are better than others. An argument that a school should be closed and merged with one a mile away, or even that top classes should be merged, falls on deaf ears even if it would undoubtedly be educationally better for the children. As one civil servant despairingly put it: ‘closing a school or taking away a fifth form is taking away “development”, and is always bad, and that is that’.39 The scope for British leverage then is strictly limited. In collaboration with other donors ODM can stall the construction of over-elaborate university buildings. It may have had something to do with the Development Plan commitment to manpower planning in education and the concentration of upper secondary teaching into fifth- and sixth-form colleges. These remain paper plans, however, and ODM has been unable to do anything about the mushrooming secondary schools or recurrent expenditure. The problem will probably only be ‘solved’ by an increase in school graduate unemployment leading the public to re-evaluate formal education.40 This may be beginning to happen. The growth of Harambee schools has peaked and is down since 1970. HMG has also attempted to influence the Kenyans towards a consistent grain policy. The Kenya government decided that a strategic grain reserve was necessary following the harvest failures of 1965. ODM made a loan and Cyprus bins were installed at two points in the country in 1966 at a cost of over £150,000, but they were made redundant by normal weather conditions and the development of hybrid maize varieties which greatly increased production. The bins were large, hermetically sealed structures, in groups of about thirty, designed for a strategic reserve and not easily adapted for rapid handling as part of the country’s normal storage facilities. They were also badly installed and began to develop cracks and other faults. Faced with apparent over-supply of cereals, the Kenyans decided on a policy of being self-sufficient in wheat and exporting maize, and requested aid for storage facilities. IBRD was interested in financing a large grain silo at Mombasa, but would only do so if the Kenyans would guarantee a certain volume of exports. Bad weather in the early 1970s and rising domestic demand, especially for wheat, then put the export policy at risk again and, in response to political pressure from MPs, a new policy of siting grain stores up-country in consumer areas was proposed. HMG was keen to aid grain storage for a number of reasons. One, it was a worthwhile expenditure given the losses to pests which occur when a grain harvest is badly stored; two, the equipment would be British, beating off an Australian challenge to British companies for the supply of grain storage and handling equipment; three, the British were embarrassed about the Cyprus bins and were anxious to renovate them and install handling equipment so they could be usefully integrated into the country’s storage capacity. The aid remained undisbursed, however, because the Kenyans

The Influence of Aid

161

did not develop a consistent policy towards grain production and storage that would enable planning of the siting of the storage facilities to be carried out. Another minor difficulty was that they did not want to spend any more money on the Cyprus bins, which they regarded as pouring good money after bad. The British, on the other hand, did not give the Kenyans credit for realising that historic costs should not influence current decisions, but were determined to relieve their embarrassment by improving the Cyprus bins and getting them into the Kenyan national storage policy—a clear case of a donor wanting its monuments to be seen to be successful. It is difficult to know why over nearly five years the Kenyans have not been able to settle on a grain policy that would enable them to make a detailed submission for aid. One problem was the volatility of national policy in response to climatic fluctuations changing the size of harvests. In addition, cereal prices, particularly the maize price, can be something of a political football with the maize farmers importuning the government, especially the President, for higher prices, but with the government anxious to keep food prices down for the urban consumer. Another problem was the rapid staff turnover within the Ministry of Agriculture. This tended to mean that a grain storage policy lasted as long as the incumbent of an office; when he moved, someone else took over, called for another report on grain storage, and the cycle started again. ODM supplied one team to prepare a report; there has been at least one Ministry of Agriculture report and in 1974 expatriate advisers at the Maize Board prepared yet another, this time proposing a conversion to storing grain in bulk instead of in bags. At the time of writing, however, no policy has crystallised. Meanwhile, while ODM was poised to spend £1m. on grain storage as soon as the Ministry of Agriculture produced a policy and a request, Denmark stepped in and gave £K350,000 worth of grain-drying equipment to be installed on an ad hoc basis at existing storage facilities. British ambitions in this area were modest, and the tactics were gentle. The idea was simply to dangle a large sum of money in front of the Kenyans to be handed over when they came up with a national plan for grain storage that made sense against the background of an agreed grain production policy. Up to date this tactic has not worked. It is not simply that ODM was unable to influence Kenyan policy, but that it was unable even to influence the Kenyans to agree on their own policy, despite technical assistance as well as promises of capital aid. The influence of foreigners on Kenyan government policies appears to be extensive. Foreign investors, aid donors and technical assistance personnel are all important. The influence, however, is largely uncoordinated and in some cases (e.g. over rates of tariff protection) self-cancelling. Aid donors have influence as a group over the composition of the public investment programme and individual donors can influence the detailed specifications of individual projects. Practically all attempts by them to influence sectoral policies have been failures, however: birth control and education are examples. We could find no evidence of donors influencing wider policies in practice, although plenty of evidence of them influencing the contents of the Development Plan, It may be that much the greatest donor influence comes through the activities of technical assistance personnel within the Kenya government machine, but it is extraordinarily difficult to assess this.

162 Aid and Inequality in Kenya: British Development Assistance to Kenya

Administration The existence of aid has profoundly affected the Kenyan administration. In the early years of independence, the presence of OSAS personnel ensured a continuity of government services and of the style in which they were administered. We cannot doubt that this was of inestimable benefit to Kenya. Whatever view one takes of the record of colonialism in Kenya and elsewhere, a precipitate and total withdrawal of the colonialists has usually been disastrous. The events in the Congo and more recently in Angola are examples. Most colonialists have been guilty of considering only very late in the history of colonialism, and then perhaps inadequately, the development of the indigenous people they ruled. But, having committed that crime, simply to clear out when indigenous opposition makes the game unprofitable is to compound the offence, In our view, some form of post-colonial involvement, although it runs the risk of becoming neo-colonialism and the much greater risk of being called neo-colonialism, is a responsibility. At independence, Kenya inherited a relatively efficient civil service. In some ways, it was not adapted to the task of delivering developmental services to the mass of the people. Agricultural marketing boards and some services of the Ministry of Agriculture, for example, largely served the expatriate farmers with extensive landholdings, although the Swynnerton Plan was an exception to the point. The provincial civil service was more concerned with its judicial functions and the maintenance of law and order,41 especially between ethnic groups, than with development per se, although it, too, became concerned with land adjudication and enclosure measures having developmental implications. At all events a relatively efficient and corruption-free machinery existed, and it is rather easier to adapt and reform such a structure than to build one from nothing. That was especially so in a country where few indigenous people were experienced in administration, where loyalty was principally to sub-national ethnic groupings rather than to a geographical entity that was the product of colonialism, and where there was no more political unanimity than in any diverse society, despite rhetoric about development and building the nation. There were shortfalls in government expenditure in the early days of independence owing to constraints of administrative and executive capacity as many experienced expatriate civil servants left anyway.42 OSAS enabled some to stay. Yet the important point is not that many OSAS personnel carried out tasks more efficiently than any Kenyans who could have replaced them, so their absence would have led to an even greater fall in efficiency, rather that their presence preserved traditions, conventions, ways of doing things, and a collective memory that was no doubt highly defective for many of the tasks it had to discharge, but at least worked. Kenyans could come in, inherit this, take it over and change it. It is possible to argue that such gradualism leads to acculturisation and resistance to necessary changes by the new elite and a desire to emulate their predecessors in unfortunate ways. Perhaps the most important of these is the example expatriates set their successors in the matter of standard of living. The continuation of expatriates in government at independence and the gradual movement of Africans into their positions involved the taking over of colonial salary scales. Now, OSAS personnel receiving supplementation and wholly-funded technical assistance personnel being paid by foreign governments can be paid more. This in turn can arouse envy and provide a pattern which Kenyan civil servants strive to emulate by fair means or foul. It is the right of Kenyan civil servants to

The Influence of Aid

163

engage in business and many of them, especially the senior ones, take advantage of it.43 The task of keeping entirely separate one’s functions as a businessman and as a bureaucrat with access to state power is a superhuman one and the people in question are only human. Technical assistance is only one very small element in a welter of foreign and domestic influences making for acquisitiveness in Kenyan society,44 and it would be nonsense to blame the existence of technical assistance for the high living of some Kenyan bureaucrats. Nevertheless, the existence of highly-paid expatriates, many of them understandably eager to enjoy their time in Kenya to the full, cannot help in encouraging the indigenous civil service in the matter of restraint. At any rate, there has been, by all accounts, some increase in corruption in recent years. There is anyway a stronger temptation for people in a privileged position in a poor country to take mercenary advantage than was true of colonial civil servants who are part of a career structure stretching back to London. Corruption is not simply of the venal kind, however. There is also the practice of giving illegitimate preference in jobs or contracts to people of the ‘right’ ethnic or regional background, with or without any pecuniary inducement. Increasing corruption is a fact of life in most African states, no doubt for general sociological reasons, some of which apply in Kenya. By continental standards, the Kenya civil service does not appear to have done badly in maintaining high standards of rectitude. Now that ODM is running OSAS down, it finds, to its surprise, no urgency about Kenyanisation in many ministries. Teaching is the area of greatest disagreement, but other OSAS officers are maintained at skill levels where one would suppose competent Kenyans were available. A few OSAS officers have converted to advisers under SCAAP, but ODM resists this kind of concession in general. It is sometimes argued that donors and technical assistance personnel are responsible for persistent use of expatriates when adequate locals are available. Certainly unsolicited offers of aid, including technical assistance, are made. Apart from budgetary reasons— fully-financed foreigners are cheaper than locals—there are domestic political reasons for the phenomenon too. These have nothing to do with so-called psychological dependency, a supposed relic of colonialism which we failed to observe. Expatriates have the important attribute in Kenya of not owing special loyalty to any of the country’s ethnic groupings. In terms of the balance of posts within a ministry, for example, an expatriate may be regarded as preferable to the commitment of the post to a Luo or Kikuyu or whatever. In addition, expatriates are not suspected of being politically ambitious. This is very important where high officials in a ministry may have risen to their position in a very few years. However able, they must be short of experience and prone to make errors or ask questions which reveal that. To confide in or otherwise reveal this to fellow Kenyans is to make a possibly damaging admission of vulnerability. An experienced expatriate can answer the questions such an official is not supposed to need to ask without risk to the position or authority of the superior. He has no interest in plotting because he has no political ambitions, ethnic or personal.45 Having expatriates in positions is also a sort of saving—their jobs, it has been suggested, represent unused opportunities for the exercise of patronage which help to keep juniors in line. Therefore having political neuters around can be useful and when Kenyanisation reached a certain level, the political pressures for more were balanced with those against it. Many donors at present, quite apart from the British, observe that there appears to be little steam behind the Kenyanisation issue in the civil service.

164 Aid and Inequality in Kenya: British Development Assistance to Kenya Expatriates can cause resentment, especially among middle-level staff who might regard them as a bar to advancement, but they do assist higher-level staff in the ways mentioned. It is difficult to say what the net effect is on morale and efficiency of having expatriates around, especially in view of the obvious difficulty of expatriate researchers in getting frank reactions on the issue. The use by some expatriates of rational problem-oriented management techniques also has some demonstration effect. This is the aspect of their work that most expatriates stress. Each time a report is ignored, or in the case of those in established posts, each time a decison is overturned, on what they regard as improper political or venal grounds, they console themselves with the thought that someone has observed their approach and perhaps some converts have been made. Now this is doubtless true and important; the difficulty is that it provides a convenient justification for any expatriate who realises that little substantive is coming of his work. This might seem to beg the question of whether the techniques and approaches which expatriates use are always appropriate. Clearly they are not always so. Some expatriates do import cultural presumptions or are excessively attracted to planning techniques requiring information that does not exist. It is extraordinarily difficult to generalise here; some do and some don’t. In a diverse country like Kenya, expatriates—at least those with overseas experience—are sometimes more flexible about problems than Kenyan nationals but, of course, so they should be. One does not import people simply to improve the quality of one’s mistakes. Perhaps the errors are more likely to occur where there are a large number of expatriates to get together and insulate each other from the surrounding reality. When that happens it also reduces any educative effect that technical assistance could have on Kenyan civil servants. In Kenya, however, aggregations of expatriates have occurred to the point where they become practically a separate estate of the realm. The UK shows its greater experience over other donors here and is relatively guiltless. Other donors make initial proposals and put in quite large teams to staff planning units, for example, in the Treasury and other ministries. ODM generally waits to be asked. When discussing aid projects in the early stages, donor representatives may talk to these technical assistance personnel and perhaps no Kenyan is present. Another possibility, which has happened, is that a donor holds talks with an operational ministry about a project and the Treasury representative present is an expatriate. If the ministry’s plans are too expensive the donor and the Treasury may concur in opposition, but the confrontation takes on racial overtones and the legitimacy of the Treasury representative is compromised. The most glaring example of technical assistance as a separate estate is in economic planning and particularly in the preparation of the Development Plan. It is no secret that large sections of all Plans since independence, as in many new countries, have been written by expatriates. The Ministry of Finance and Planning is very conscious of the Plan’s function in attracting aid and the Plan often reflects donor opinions. That is not as a result of cynical dissimulation, however. The expatriates who write the Plan sections genuinely share these opinions, and the control of the Ministry of Finance and Planning over their Plan is much better than it is over the actual spending and policies of other ministries. The Plan does not necessarily represent the agreed view of the Kenyan government therefore. The notion of collective responsibility is weak; ministries only fight for their policies to

The Influence of Aid

165

appear as such in the Plan up to a point. Beyond that point they concentrate on ignoring the Plan if it does not suit them. The fact that only certain expenditures are budgeted in the Plan can act as a powerful constraint on a ministry (although within their budget ministries can change policies how they like), but the Plan should be seen therefore not as the outcome of a battle for resources, but merely as a shot in it. Pumping in expatriates to the Planning Units of ‘weak’ ministries can certainly lead to better paper planning (i.e. a Development Plan that coincides better with donor views), but at the risk of making the Plan deviate further from any kind of description of a political consensus. What donors frequently diagnose as a technical failure (‘inability to draw up adequate sectoral plans’) for which they prescribe technical assistance is in fact a political failure. Priorities in development or clashes between ministries over policies or for resources are all too rarely resolved in Kenya by Cabinet decision. Cabinet meetings for that kind of business are rare. A subcommittee of the Cabinet, the Development Committee, used to fulfil the role but it was wound up in 1967 and its replacement, the Committee of Economic Ministers, has failed to fill the gap. Hence some political decisions are taken at administrative level and more are not taken at all. If a powerful Permanent Secretary does not like a policy, he may simply not implement it, unless it is one of those few policies handed down by the President, as for example, when he announced free primary education. The announcement, made without consultation, held up the publication of the Development Plan while it was hastily written into the section on education. The announcement was interpreted so as to water it down and restrict it to the first four years, but it still left the Treasury wondering where the finance was to come from. No Cabinet decision was taken, however, on what expenditure should be cut to allow the extra spending. Instead some things just failed to happen owing to the Treasury failing to find money for projects in an essentially haphazard way. The same applied, to a lesser extent (some attempt at re-planning was made and given up), to the oil crisis. This lack of political direction is why the Kenya Treasury does not rank development projects. That, too, is wrongly viewed by some donors as a technical failure. But no government ever really decided on priorities by establishing parameters for cost-benefit analysis of projects in different sectors and then ranking them all by their net present value at a Treasury-determined accounting interest rate, It is true that the Kenya Treasury does not even have an accounting interest rate, but that is the least of its worries. More importantly, it does not have Cabinet decisions saying this ministry’s baby before that one. Without that, each ministry is a semi-independent fief of its top officials called to render particular service now and then when the President speaks. And that is why donors can determine collectively much of the shape of the government investment programme. Technical assistance cannot solve this basic problem and, in fact, may make it worse. It can strengthen the cost-conscious elements in the government who are in favour of rational forward planning on the basis of minimising costs to achieve given ends. But it can lead to a concentration on just that sort of problem at the expense of concentrating on developing agreement on what the ends are. The resulting plans are then just weapons in the battle to decide what the ends are in the absence of inter-ministerial agreement. The battle is fought within the Kenyan administration with project plans, unilateral policy statements, promises of financial aid from donors, and appeals to public opinion (the last being the particular trump card of the Ministry of Education). Some observers claim, of course, that

166 Aid and Inequality in Kenya: British Development Assistance to Kenya technical assistance is not supposed to help but to make it even easier for donors to control recipient government policies. That is not true of British aid at any rate. As noted, there is no contact between ODM and technical assistance personnel, and while the British are more permissive about supply of personnel under SCAAP than they are about capital aid, they are relatively diffident about thrusting them at Kenya. It is clear that Idcs, including Kenya, have an absorptive capacity for technical assistance that is as limited as that for finance. The fact is not widely realised because, unlike finance, the technical assistance is taken up whether it can be usefully employed or not, if the recipient thinks doing so is a condition for receiving more capital aid. We believe technical assistance is at saturation point in the planning nexus of the Kenya government. It may be so in other areas too. While one can guess that, at the margin, any shift in the supply of technical assistance should be down rather than up, it is almost impossible to generalise validly about the overall impact of technical assistance on the Kenya administration at the present time. We should guess it is positive and to advocate dispensing with all of it would be pushing the baby down the plughole in advance of the water. That was certainly true at independence, though probably much less so now. At any rate, the ODM policy of introducing a manpower review and seeking to scale down OSAS commitments is timely. ODM’s relative forbearance in the matter of proposing technical assistance projects to the Kenyans is also to be applauded. Indeed, there is still a case for offering assistance with specific technical skills to do specific work in the preparation of specific projects on a shortterm basis. There can rarely be a case for offering anything other than that. When donors do offer technical assistance to an operational ministry, there are two different procedures for the offer to be accepted. In the case of wholly-funded personnel—in the British case, advisers supplied under the SCAAP agreement—the request goes to the appropriate division in the Ministry of Finance and Planning for appraisal. In the case of partly-funded personnel who will have an established post in the Kenya civil service—in the British case OSAS personnel—the application goes to the Directorate of Personnel.46 This control, however, is partly annulled by donor practices in extending capital aid. It is easier to generalise about the effects of capital aid on the Kenya administration. On balance, the effects have been bad, but the British again have been less guilty than most other donors. In principle, requests for a budgetary allocation for a new project or any sort of financial request are supposed to be forwarded in the first place from an operational ministry to the Ministry of Finance and Planning. New projects are supposed to go to the Planning Division and routine matters to the Finance Division. Proposals should be considered by the Estimates Working Group which has both Treasury and Planning representatives. The External Aid Division are also represented, as are people from the operational ministry, on the Group. If a project is passed by the Estimates Working Group, it is included in the operational ministry’s budget and then sent to the External Aid Division for a donor to be found. Otherwise, it may be rejected or referred back. When the Plan is being written, Development Plan working groups deal with sectoral plans as the Estimates Working Group does with projects. So much for the theory. In practice, operational ministries after initial talks with a donor, at which the Treasury may or may not be represented, often pass proposals to the External Aid Division for forwarding ‘officially’ to the donor without the Planning Division being told at all. This means there is not even an

The Influence of Aid

167

adequate central projects registry. The project planning unit at the Ministry of Finance and Planning (supported as a technical assistance project by the Canadians) has no procedure even to be informed as to what projects are in prospect. Once the donor has agreed funds it becomes very hard for the Treasury to refuse a budgetary allocation for the project, which is a fait accompli. There is a rule that Treasury officers should always be present for negotiations on revenue and finance between an operational ministry and the donor. Even if this obligation is carried out, there is no ruling that Treasury representatives have to be present for technical talks. More-or-less technical talks may be the important early ones, however. In any case, the technical talks always have enormous implications for finance. This is a situation in which the Treasury finds it extremely difficult to exert any intersectoral control. In order to be able to meet obligations which it may be confronted with, the Treasury is reluctant to make a budgetary allocation for any project unless there is a prospective donor. Donors therefore not only have enormous influence, as noted, on the inter-sectoral balance of Kenya development spending, they also undermine the functions of the Treasury and Planning Division. Operational ministries do not have to go to the Treasury and compete for funds on any rational basis. If they try that, however technically good a project may be, they are unlikely to get a cent. The first step therefore is usually to find a donor. As we noted in Chapter Four, the British scrupulously attempted to follow the procedures of always dealing with the External Aid Division of the Treasury, with the effect that their disbursements were very slow. Now they are increasingly making more informal contacts with operational ministries. That is understandable, and it is difficult to see what else to do if one wants to spend money. Some donors confine themselves to making approaches on the basis of the Development Plan, but others, such as the USAID, do not even so confine themselves.47 The situation has led to a condition among Kenya bureaucrats satirically referred to as aid-happiness. Sufferers from the condition may be recognised from their habit of approaching their ministerial tasks not with the question, what needs to be done, but with the question, what can we get aid for? The answer tends to be for something big. All donors prefer to support large projects if only because of the saving on administrative overheads that implies. This preference generates the need for more technical assistance.48 It is largely beyond the capacity of the Kenya administration to assemble and present data to justify many very large projects. Consultants or technical assistance teams have to be used, and there is no question of these people having local counterparts—an example is the British technical assistance team largely staffing the Tana River Development Authority. The Kenya government machine can cope reasonably well with small projects. That is clearly shown by the fact that the aid it has such trouble getting through is only 12–15 per cent of total government spending. Some 35 per cent of the projects in the current Development Budget are of less than £50,000 and include most of the wholly-Kenyan ones. Fewer than a quarter cost more than £0.25m. In this area, too, the British used to be more flexible than other donors and were prepared to finance very small projects but that is no longer so. Capital aid has therefore had deleterious effects on Kenyan administration by undermining central control of the development budgets of operational ministries. The accent on large discrete projects has put the staff of these ministries under a burden which they cannot bear and led to a greater felt need for technical assistance. While central control of applications

168 Aid and Inequality in Kenya: British Development Assistance to Kenya for technical assistance is tighter, the need for it, to which the centre accedes, is a result partly of the desire of donors to spend capital aid on large projects, and partly of the desire of the Kenyans to receive foreign exchange. In fairness we must repeat here an observation of Chapter Four that Kenyan civil servants in the Treasury do not seem too concerned about the pattern that has evolved. Donors remove the need for them to make choices in the absence of political leadership, and the procedure facilitates the spending of money. Our judgement, however, is that the situation is regrettable.

Political Economy The way that aid has weakened the Kenya administration and has resulted, in effect, in the organisation of the investment budget by aid donors, has been possible only because there has been a lack of political leadership for development in Kenya. Some donors, including ODM, would be happy to see such leadership, and indeed some ODM practices in the past have been based on the incorrect assumption that it existed. This leads to the question, whatever aid agency professionals believe, does aid have any effect on the political nature of the Kenya government? In weighing the probabilities we shall describe briefly Kenya’s political system. Formally, this is one of parliamentary democracy. There is only one political party, but there are no laws under the Constitution against the formation of others, although they must be registered under the Societies Act which gives considerable discretionary powers to the government. There are few restrictions on citizens joining the existing party, KANU, and standing as parliamentary candidates. The House of Assembly and the British House of Commons work similarly, in theory, in retaining financial control of government expenditure, and the executive is supposed to retain the support of the majority in Parliament to pass legislation and indeed to remain in office, The President is elected by universal adult suffrage directly, but he must also be a member of Parliament to which he must be elected in the usual way. At least one textbook writer has stated49 that this formal political system is in fact subordinate to the bureaucracy in Kenya. Governmental control is exercised and policy executed through the provincial administrations, directly responsible to the Office of the President. The influence of the political party, KANU, and its functionaries is not great and neither is that of the typical back-bench MP. While the system has democratic elements, therefore, it is sometimes represented as being managerial or bureaucratic with the collective interests of the bureaucracy as a class being most influential in the running of the country. Certainly a number of senior and, more especially, middle-level civil servants to whom we spoke declared that the civil service was running the country, that there was a lack of strong political direction and influence. The view was expressed that this may be no bad thing. The bureaucracy was interested in efficient management for development rather than the demagogic considerations that influenced politicians and furthermore it had, if not a monpoly, at least much of the management expertise available in Kenya. We must stress that in our view this picture is inaccurate and does not adequately describe the situation in Kenya. The formal political system may appear to lack power if not animation, but as in many, perhaps most, countries, the loci of political power are not what a study of the Constitution or of textbooks would lead an observer to expect. In fact what might be termed political considerations, as opposed to managerial ones, are

The Influence of Aid

169

dominant in the running of many departments of state and in the allocation of economic resources except, partly, those that are allocated in the development budget. This is so although critical decisions are not made as part of the operation of the formal political system. There is an identifiable power elite in Kenya. It is grouped around the person and authority of the President and members of the Kikuyu ethnic group are prominent in it.50 The President enjoys widespread authority in Kenya both for his symbolic role in the independence struggle, and because subsequently he has come to represent national unity in a society with strong centrifugal tendencies. Nevertheless one important basis of his political power and that of his attendant power elite is ethnic. Kenya is a plural society with considerable regional and tribal loyalties and rivalries. In the past, leaders of clans have been able to maintain their position by appeals to clan solidarity in the face of the rivalry of other groups. The position of such ethnic leaders in the Kenyan power elite is now enormously reinforced by the group’s access to the resources of the state. Political supporters can be rewarded and opponents deprived. The importance of the bureaucracy in distributing resources also means that tribalism cannot be dismissed as false consciousness. It really does help if members of one’s clan are on the selection committee, so it is in one’s interests to keep them there. The elite’s position has depended therefore on a combination of presidential prestige, traditional authority, ethnic solidarity and the power of patronage. Outright coercion and, many believe, occasional assassination have been employed as well as electoral malpractices—certain candidates are prevented from standing or, if allowed to stand, from holding meetings. Overt repression has, however, been rare in Kenya compared to many European and most African states. Obviously it may be that the relative importance of these different pillars is shifting, and that a distinct socio-economic class is emerging under the umbrella of the regime, whose loyalty has been cemented by patronage and who are losing any autonomous tribal base. The general perception, however, remains that Kenya is still a reasonably, if precariously, liberal society, and the Kikuyu as an ethnic group have a dominant importance in the system. The reasons for this eminence have been discussed by other writers.51 In maintaining their position in the polity as a whole, the Kikuyu hierarchy employ similar means to those used in the intra-tribal context. Coalitions are made with acknowledged leaders of other tribal groups. Some of the Kalenjin have recently been in a coalition with the ruling Kikuyu against more disaffected groups such as the Luo. The line-up was different at independence with the Luo and Kikuyu in alliance in KANU against the other, mostly smaller, tribes of KADU.52 The coalitions are dynamic, and there are tensions and shifting allegiances even within the Kikuyu. Tribal allies are cemented to the elite by the power of a patronage that is both personal and collective. Certain government expenditures can be directed to the areas that are the natural constituency of the allied leaders, thus establishing their value as political leaders by demonstrating their own, derived, power of patronage. And opportunities are afforded to them to thrive in business via the direction of both private and public money. While there may well be an elite group which holds general political views or at least loyalties based on common economic interests, it does not seem that this class consciousness has spread to the people as a whole, with the possible exception only of some unionised urban workers.53 Certainly people do not vote for politicians to represent distinct views on

170 Aid and Inequality in Kenya: British Development Assistance to Kenya economics and politics, and the government does not consist necessarily of people with very similar views on matters of economic and social policy. Politics in Kenya is not about policies at all in that sense. At the grass-roots level groups still identify themselves on the basis of tribal parochial affiliation rather than economic or class interest and they vote for the people who will maximise their access to the pork barrel—certain resources at the disposal of the state. This is demonstrated by the nature of the appeal that politicians attempted to exert in the 1974 elections.54 It is probably inevitable for two reasons. One is the ethnically heterogeneous nature of Kenyan society; the other is the uninformed position of the general public about what can be expected of the quality of government services. For example, the electorate does not vote for the man with the most persuasive transport or education policy and hope to enjoy its share of the fruits of a national improvement; it has no views on such matters, It votes for the man most likely to get a road or a school for its particular constituency. It scarcely matters whether the road is the best possible, or what the examination pass rate of the school is. People in general have few expectations on such matters and are unaware of the possibilities for improving the quality of the contents of the pork barrel. They are simply concerned that their man should bring home his share of the bacon. The power of grass-roots aspirations is often considerable, as shown by the growth of the Harambee movement. In some areas, and education is the most obvious, the peasants make themselves felt and the centre has to make provision for their feelings. In general, however, political power in Kenya runs largely downwards, from the President and that group of people who have his ear or who are very important in the intra- and inter-tribal balancing act. To describe the system is to reveal immediately that there are considerable ‘political’ constraints on planning or the rational allocation of economic resources to maximise some objective function such as the growth of GNP per head.55 Actors and agents in the informal (but real) political system, and their political constituencies, have to get their pay-off. Leaders of important marginal tribal groups have to be given positions of access to certain state resources (ministries etc.) if stability is to be preserved. Such claims take precedence and any ‘economic’ plan that ignores them is waste paper. This raises the question of what, apart from maintaining and strengthening this status quo, are the objectives of the power elite. The answer is that they appear to have few apart from the desire to accumulate wealth. The wealth accumulation by prominent people is a matter of near public comment.56 The Ndegwa Report57 merely legitimised an existing situation in which senior civil servants devoted considerable time and energy to the development of their personal business interests. In case this appears to be painting too dark a picture, it should be stressed that the system has the virtues of its defects. There is no overwhelming politicisation of Kenyan life. Kenyan society is relatively liberal and has enjoyed, for example, significant freedom of speech. Indeed, those civil servants who, out of esprit de corps, wish to plan for rapid economic development are free to do so. As we have pointed out, there is not much political dynamism in support of many of their efforts, but as long as these are not inconsistent with the political balancing act and the desire of the elite to accumulate wealth, and as long as they do not require much coordination between ministries, they can, and do, go ahead. The extent to which ‘rational’ development planning is possible in a ministry depends on the political importance of the ministry’s services. Education and land are the most coveted

The Influence of Aid

171

commodities in Kenya, and this is reflected in the course of educational policy and land adjudication, which both have their own dynamic. On the other hand, with the replacement of large-scale expatriate farmers by African small holders, the services of the Ministry of Agriculture have become less valued (though not less valuable). Consequently, the Ministry has declined in both influence and efficiency. Politically less important ministries are much more open to donor influence, which is expressed largely by the pattern of development expenditure. It should not be assumed that the importance of a ministry is adequately indicated by the apparent prestige and power of the Minister. The Kenyan elite spreads across the conventional lines separating the formal politician from the bureaucrat. And the political operator and disinterested administrator may be the same person at different times. The Minister or civil servant who attempts, via a Cabinet or Civil Service Committee, to restrict the expenditure of, say, the Ministry of Works, in the interests of optimal allocation of resources, may be the same man who pulls strings in the informal political system to get a tar road built in his ‘constituency’ or just past his front door.58 This is clearly a system in which inequalities of wealth and of access to the means of wealth—political office, credit, land and education—are not likely to diminish and may well grow. Up to now, however, sufficient material improvement has been diffused through the system to maintain its stability without comprehensive repression, despite plenty of grumbling by out-groups. As long as the elite does not become too acquisitive, but allows some resources to flow to areas where it has no direct interest, this may persist, although many people have expressed fears that acquisitiveness has become too conspicuous and, once the personal charismatic authority of the President is removed by his death, the system will have lost a possibly critical stabilising factor. An increase in obvious acquisitiveness and a crackdown on parliamentary opponents of the regime have occurred recently.59 Some academic opponents of the Kenya regime have, for some time, prophesied increased repression by the Kenya government as a tendency somehow inherent in the system, although there has been less repression in the past than in most surrounding states. Recent developments may still simply reflect the ageing of the President and an attempt by his immediate associates to secure the succession. The attitude of the main Western donor nations to the system in general has been largely predictable. They have had no quarrel with it. It has provided a stable climate in which foreign commercial enterprises may prosper. It is true that the foreign manufacturer may be pressured into appointing a Kikuyu wholesaler rather than a Kenyan of his choice, and he may have to pay to get favourable decisions out of civil servants about regulations that affect his business. Businessmen, however, are used to dealing with the price system and this is just another form of corporation tax. From an elite dedicated to the acquisition of private wealth, talk of nationalisation and expropriation is not to be feared.60 Indeed, foreign enterprise is to be welcomed for the additional resources it brings in and ‘taxes’ it can pay. So far, we do not think many observers would quarrel with this account. Disagreement arises about whether the situation is to be regarded as exceptionally bad, what are the reasons for it, and what is to be done. Marxists will be inclined to see this situation as arising from the machinations of international capitalism and to view the Kenyan elite as its creatures.61 There is in fact little evidence for this. Britain, for example, had few expectations of an

172 Aid and Inequality in Kenya: British Development Assistance to Kenya independent Kenya and foreign investment fell heavily after independence. The KANU government was allowed to take office as the only group that seemed likely to command popular support and maintain some kind of stability. Little more was expected of it. The causal origins of the individualistic acquisitive ethic evinced by the Kikuyu elite today are doubtless complex. We are not competent to say what are the relative importances of colonialism and longer-lived indigenous cultural factors in its development. It may, however, be a form of inverted racialism to assume that every attitude evinced by an African government is the responsibility of white men. Kenya does not occupy at present a militarily strategic position, nor are there scarce resources of vital minerals there. The effort which an optimising exploiter would put into keeping his heel on the country’s neck would be strictly limited. Because ‘international capitalism’ takes advantage of a situation and donor states support it, we cannot conclude that it was created by external agencies. Furthermore, if as seems likely, the system has indigenous roots, it is not clear that aid donors are wrong to support it, even though it is a highly imperfect system. In supporting it, in fact, donors may be able to effect marginal improvements. The widespread support of donors for ILO-type proposals and the Kenya Sessional Paper on Employment suggests now that donors may be aware of possible limits to the system’s stability and be concerned to broaden the base and legitimacy of the present elite by pressing for a more equal distribution of benefits. In so far as the Kenya elite does not keep pace with donor wishes in this regard, it appears that it is unable to transcend the short-run claims on resources described above in the interests of a securer future. In assessing the ‘whiggish’ approach of donors (which no doubt is in some donor country interests too), the crucial question is: what is the most likely alternative to the status quo? It would be nice to think it was an egalitarian commonwealth of free and progressive people (then donors could be condemned for postponing this consummation), but we find it difficult to believe that. If the pork barrel were smaller, the existing system might be less stable. But inspection of the regimes of poor, ethnically divided countries turns up few model democracies. Whatever the most probable alternative to a relatively liberal ethnically-based plutocracy in Kenya is, there seems little reason to assume that it would be an improvement

Conclusion: Incomes and Income Equality Aid has constituted a flow of resources to Kenya, and nothing that we have seen leads us to reject the common-sense conclusion that it has raised incomes in the country. One effect almost certainly was to release resources for the private sector that might otherwise have been obtained from a private capital inflow, but any such substitution was probably small. It is more probable that private investment would have been lower without aid. In some pathological cases of over-protection of industry that might have been a good thing, but, in general, investment must have increased domestic value added. Alternative patterns of industrialisation might be more appropriate but (if sugar is anything to go by) one cannot be confident that they would have evolved in the absence of foreign investment. The activities of government, boosted by aid, have also by and large raised incomes via the provision of infrastructure and productive and social services. In the long run aid has perhaps been less effective in raising income than it might have been, by causing the

The Influence of Aid

173

government to emphasise discrete expenditures at the expense of recurrent ones and large expenditures at the expense of small. Having induced this bias, however, some donors, including ODM, have resisted some of its worse effects by scrutinising applications for projects and attempting to resist grandiosity. ODM has kept up with current thinking in development. Obviously it would be unreasonable to blame an agency for not thinking more about appropriate technology and rural development in the intellectual climate of the early 1960s. Once a view has taken hold, however, ODM attempts to apply it in a fairly disinterested way. The sectoral preferences it has shown have by and large been appropriate therefore, as far as anyone can tell. In so far as they have been shared with other donors, they have had an effect on the pattern of public investment in Kenya. In general the sectoral preferences of the majority of donors now correspond with those of what we regard as the more enlightened and progressive elements in the Kenyan administration, so their influence is not bad in this particular. The techniques of giving capital aid, however, have weakened central control and the need for consensus in the Kenyan administration, and the benefits of technical assistance, while once overwhelmingly important, may no longer be making up for this. Aid as a contribution to resources may make some marginal continuing contribution to political stability, although we doubt it. The provision of large numbers of OSAS personnel in the early years of independence and the Land Transfer Programme did assist political continuity—the Land Programme not because the Kenya government could not have paid for the land without the capital sums (we believe it could, and indeed should, though at somewhat lower prices) but because it might not in fact have done so. That a different Kenya government might have been instrumental in raising its people’s incomes on average by much more (or much less) than has occurred is pure supposition. Comparing Kenya’s flawed and unequal progress to the continental average the best guess is perhaps that another government would have done worse. Aid has probably increased economic inequality among Kenyans. It has done so partly as a simple by-product of raising incomes. In releasing resources for modern sector development, aid probably caused a relatively few highly-paid jobs to be created, so increasing inequality. All three of the projects we looked at in Chapter Five also had such an effect. The Land Transfer Programme financed by Britain transferred large areas of mixed farming land from European to African ownership. The land was sold to the new owners; it was not distributed free to the poorest of the poor, though the Haraka settlement might be thought to fall into this category. The benefits of this programme accrued to a minority (though large in absolute terms), and not a minority composed of the poorest. The benefits of the acquisition of land were multiplied by the large input of agricultural extension and other services into the settlement schemes, which had its origin in the plans of the colonial government to create a class of substantial small farmers. Britain may be thought, therefore, to have some responsibility for the growth in inequality among the African population that has resulted from the Land Transfer Programme. From another point of view, of course, land transfer must substantially have reduced inequality, because the numerically few European farmers were in general far richer than the numerous Africans to whom the land was transferred. It is to be remembered, in this connection, that half of the aid provided for land transfer was for settlement schemes, that is, for the subdivision of farms, and therefore must have financed the redistribution of wealth in an egalitarian direction. In

174 Aid and Inequality in Kenya: British Development Assistance to Kenya addition to this form of land transfer, large farms have been transferred from Europeans to Africans as going concerns. This form of transfer may not greatly have affected the overall distribution of wealth, Africans having replaced Europeans, but it undoubtedly increased the inequality within the African community. Although it could be argued that British aid was not directly involved with most of these transfers, and cannot therefore be held directly responsible for their effects on the distribution of wealth, British aid provided funds for the Agricultural Finance Corporation and the Agricultural Development Corporation, which have been involved in the transfer of large farms. It could, of course, also be argued that the settlement schemes took the steam out of what would otherwise have developed into an irresistible movement for the seizure of land by the landless, and so made possible the transfer of other European farms to a new class of African large land-owners. It is not implausible that without substantial small-scale settlement, the remaining large farms could not have been kept intact, so that the maintenance of large farms in private ownership was dependent on the settlement programme. But although the absence of aid might very likely have meant that the farms were not purchased for settlement, in the way that they were in fact purchased, it would not necessarily have prevented the operation of enough settlement to take the steam out of the demands of the landless, and to enable some large farms to be transferred intact to would-be African large land-owners, so long as it was the policy of the Kenya government to encourage such large-scale farming. We have found no evidence to suggest that this policy is pursued by the Kenya government as a result of leverage exerted by Britain, or that the Kenya government would have discouraged large landholdings by Africans in the absence of aid for land transfer for settlement schemes. The interests of the rich and powerful in large-scale farming are too great to suggest that such a difference in policy was probable. Of course, the maintenance of conditions under which the continuation of a private large-farm sector was possible is seen by some as the major beneficial contribution of aid to Kenya. The maintenance of ‘law and order’ and the absence of the disorderly and violent seizure of land could be considered as the great achievement of the Land Transfer Programme. Our view is somewhat different, because we do not believe that settlement in much the same way as that undertaken would have been impossible without capital aid for land transfer. However, the purchase of the land, instead of a more or less open dispossession of the European farmers, as well as maintaining the productivity of the largefarm sector, is also likely to have had favourable effects on Kenya’s international image and on her attractiveness for foreign business and investment. Nor do we think that a breakdown of law and order, if it had occurred, even though in the process of an egalitarian redistribution of land, would have been to the benefit of the poor and unprivileged. In summary, then, it is our view that aid for land transfer and settlement can be criticised for contributing indirectly to the growth of inequality, by one who attaches little weight to the absence of major civil disorder and who sees no disadvantage in the creation of an economic and political climate which international business finds unfavourable. For those who rate the absence of violence highly, and who believe that it is possible (though not inevitable) for a poor country to benefit from the operations of international business, the indirect influence of British aid on the degree of inequality in Kenya may be seen as the price to be paid for other benefits.

The Influence of Aid

175

The discussion of the Mumias Sugar Scheme in Chapter Five drew attention to the inequalities it has generated within the African population of the area. Those who were in a position, because of the location of their land, to be enrolled as outgrowers gained very greatly in income. The benefits have not so far spilled over very much to those who were unable to enrol, or who did not obtain wage employment with the company or with outgrowers. The gains were in the nature of an economic rent, unrelated to the supply price of the factors of production contributed by the outgrowers, who were protected by the structure of the scheme from the competitive erosion and wider distribution of their rents. It is certain, however, that with the further development of sugar-processing capacity the benefits will be spread more widely, and the inequities in that sense will diminish. It was suggested in Chapter Five that a differently designed scheme might have been practicable, which would have spread the benefits more widely from the start. But it was not the fact of aid that led to the problems of the scheme in the form in which it was introduced, let alone the fact of British aid. Indeed, a purely commercial scheme would have emphasised estate production more. Mumias is another scheme that would have been undertaken by some other donor, or even without aid, if Britain had stood aside. A design which might have led to a wider distribution of the benefits might have been possible if the Kenya government had been clear on what should have been done differently, and had insisted on such changes in the scheme. But it does not appear that the form of the scheme adopted reflected the intention of the Kenya government, the donors, or the commercial interests, deliberately to neglect the adverse distributional effects in their own interest. The scheme was innovational in its use of outgrowers to supply cane. It is not surprising, therefore, that it was desired to limit risks by supervising outgrowers closely and by adopting familiar, welltried methods in other elements of the scheme, which would be easier to implement (and more certain of success) than if these also required new ideas, adaptation and experiment. It is easy enough to see, with hindsight, that a more beneficial scheme might have been implemented. A repetition of Mumias, without detailed examination of alternatives having more widespread benefits, could be more open to criticism, though investigation might well reveal that a Mumias-type scheme was best. An examination of other British-aided projects leads to much the same conclusion about their effect on inequality. The improvement of the Mombasa-Nairobi road, with the replacement of the gravel surface by tarmac, carried out partly with British aid funds, had several disequalising effects within Kenya, as well as knocking a nail into the coffin of East African Railways. It did benefit the rich, who were able in consequence to drive in comfort and speed between Nairobi and the Coast. In fact, the expectation of this benefit was the main reason for the high priority given to the improvement by the Kenya government. The improvement made the transport of petroleum products by road feasible and, with the relaxation of licensing restrictions, opened opportunities for highly profitable investment in bulk-haulage vehicles by a few Kenyans. In these respects, the investment distinctly added to inequalities. But petroleum products were the railway’s most profitable traffic, and the diminution of these profits as a result of road competition probably added further to inequalities by shifting the burden of charges on to goods consumed to a relatively greater extent by the poor. On the other hand, it must not be assumed that only the rich travel on the road; although the benefits of lower operating costs may not have been passed on to the passengers, the bus and taxi services between Nairobi and the Coast, much used by passengers of only modest means, must not be neglected. The improvement of the road did

176 Aid and Inequality in Kenya: British Development Assistance to Kenya not benefit only the rich. And if it did, can aid be blamed? In this case, the availability of aid certainly did not divert resources from other projects which would have had beneficial effects on the distribution of income and wealth. In fact, the fungibility of the aid may have resulted in the financing of some project that could be judged as more socially beneficial than the road which was nominally financed. The improvement of the road was given high priority in high places in Kenya, and it is unlikely that it would have long remained unimproved, whatever the attitude of Britain and other donors. One might speculate on whether refusal of all aid from a large donor such as Britain—or perhaps the World Bank— if the project were carried out, might have deterred the Kenyans. But it was hardly so bad a project as to justify such extreme leverage; and in many ways it looked a very good project to aid donors at a time when more attention was being given to such easily planned and executed projects as trunk roads than is, in theory, now fashionable. It is not only British aid projects which have contributed to inequality. Studies of other aid projects in Kenya have arrived at similar conclusions. An account of a Swedish artificial insemination programme showed that it benefited more progressive farmers, increasing their milk output, to the relative disadvantage of others.62 A study of the industrial estates programme supported by West German aid63 showed that the estates mostly benefited established industrialists, by subsidising their inputs, and failed to induce new entrepreneurs into operation. We do not know what the effect would have been on income distribution if the Kenya government had spent more on maintaining existing projects and less on new ones, or, not knowing aid was correlated with bigness in projects, had gone in for smaller projects. Much depends, of course, on the details of what expenditure would have been incurred and that is unknowable. There must be some presupposition, however, that a large project is worse for income distribution than a number of small ones. An exception might be if the large project is in fact a programme consisting of many linked small projects. It is only recently, however, that donors have moved to such a concept. Whether one considers that aid is a force for inequality, above that which is inherent in growth, probably depends on the date at which aid is considered. It may be that aid was a force for inequality for the first ten years of independence and cemented the Kenya administration into certain practices. Now enlightened donors, by pushing rural development, may be a force for equality. What has aid done for dependence? It depends whether integration into world trade counts as dependence or not. If so, CDC by increasing tea production has increased dependence while by increasing sugar production ODM has reduced it. In both cases there was considerable initial but subsequently declining dependence on outside experts. Clearly the dependence consideration can rapidly degenerate into semantics. For example by providing experts for the Tana River Development Authority, ODM may be bringing nearer a hydro-electric development that will substantially reduce Kenya’s dependence on foreign oil. The authority itself, if the scheme comes to fruition, will undoubtedly have to be run largely by foreigners for the foreseeable future, this perhaps being one of those schemes that are worth undertaking although any conceivable technology for them is ‘inappropriate’. Is this a net increase in dependence or not? Perhaps increasing dependence could be defined as a situation in which a country becomes more dependent on the outside world at any level of income per head. That would appear to be a bad thing. One can, if one chooses, however, call dependence a situation

The Influence of Aid

177

in which a country becomes more dependent on the outside world (through, for example, trade specialisation) but income per head is increasing. That situation, however, which may apply to Kenya, cannot be inferred to be worse without further value judgements. It is a situation in which one can trade off one desideratum against another. Field Marshal Amin has made one choice and the Kenya government, hitherto, has made another.

Notes 1. See Chapter Three for an attempt to arrive at the true value of procurementtied aid. 2. Consider two donors who prefer project A to project B. If donor one finances project A, he preempts donor two and pushes the latter’s funds into project B. If we hold, with some, that donor one has therefore effectively financed project B, what has donor two financed? We are forced to conclude that they both effectively financed B and no one effectively financed A. 3. An exception might be where there is an historical causal connection, so the recipient Treasury was poised to give the go-ahead to some pioject as soon as a donor could be found to finance some other project. The marginal project was just waiting for the marginal donor. However this is unrealistic as the common scenario. 4. ‘Project’ is used loosely here—it encompasses any clearly defined expenditure or reduction in government receipts. 5. Perhaps the only way would be to find a project that required no counterpart funds from the recipient and no administrative input by the recipient, and was regarded by the recipient as so worthless that it did not affect his resource mobilisation or preference-ordering of other projects. That does not sound like a recipe for a successful project. There may be an iron law here: the more dissimilar are the preferences of donor and recipient, the more likely is project-tying to get a genuine addition to the expenditure-stream but the more likely is the recipient to arrange offsetting changes in his priorities and, anyway, the less likely is the project to be made to work well. 6. There is no reason in general to expect shunting to be worse than shuffling. Suppose, for example, the recipient wants to undertake projects 1 to 4 but can finance only three of them. The donor insists on financing project 5, regarded by the recipient as a substitute for 2. Consequently the recipient undertakes projects 1, 3, 4 and 5. We know the donor prefers 5 to 4 but we do not know that he prefers the above line-up to 1, 2, 3, 4. 7. Seep. 185. 8. P.S.Heller, ‘A Model of Public Fiscal Behavior in Developing Countries: Aid, Investment and Taxation’, American Economic Review, Vol. 65, No. 3, June 1975. 9. Specifically, Heller estimates three structural equations of his model by twostage least squares, but estimates of the co-efficients in one equation have to be used to construct variables in the other equations—and the estimates have large standard errors. Furthermore estimations of the three equations give structural parameters of the model, but to find the effects of aid, treated as exogenous, Heller needs to know the value of the co-efficients of the reduced form equations of the model (i.e. the equations expressing the endogenous variables as functions of exogenous variables only). These co-efficients are obviously functions of the structural parameters and are derived algebraically from them, Such derived co-efficients, however, would obviously have larger standard errors than the parameters directly estimated. Although this is much the best attempt to date to estimate the effects of aid econometrically, therefore, the results are suggestive at most. 10. The countries were: Nigeria, Ghana, Kenya, Uganda, Tanzania, Malawi, Liberia and Ethiopia. The data were pooled cross-section and time series for the post-independence period. Whether Kenya differed significantly from the average could be ascertained by omitting it from the sample and conducting a ‘Chow’ test to see whether parameters changed. This was not done in view of the problems of the methods already noted.

178 Aid and Inequality in Kenya: British Development Assistance to Kenya 11. This would tend to reduce the money supply, which, however, rose owing to secondary credit expansion by the banks and the effect of the balance of payments surplus. 12. Kenya government, Economic Survey 1974, p. 166. 13. Development Plan 1974–78, Nairobi, Government Printer, pp. 178 and 182. 14. D.Ndegwa (Governor), Central Bank of Kenya 7th Annual Report (year ending 30 June 1973), p. 63. 15. See ‘Tribulations of an African Tycoon’, Financial Times, London, 10 April 1974. 16. B.Herman, ‘Some Basic Data for Analysing the Political Economy of Foreign Investment in Kenya’, IDS Discussion Paper No. 112, Nairobi, 1971, mimeo. 17. There was a single exception—a short-lived Rhodesian firm. 18. The principal banks are Standard, Barclays, and the Kenya Commercial Bank, owned 60 per cent by the government and 40 per cent by the National Grindlays group. In 1973 these three banks held over 80 per cent of bank deposits in Kenya. 19. V.Diejomoah, ‘Taxation and Government Savings in Kenya’, Eastern Africa Economic Review, Vol. 2, No. 2, December 1970. 20. International Labour Office, Employment, Incomes and Equality: A Strategy for Increasing Productive Employment in Kenya, Geneva, 1971, p. 272. 21. Development Plan 1974–78, op. cit., p. 25. 22. The advantage of taxing luxury goods is not simply equity but having an income elasticity of more than one, the goods should yield a tax revenue growing faster than personal disposable income. 23. British American Tobacco and East African Breweries, respectively. 24. Republic of Kenya Development Estimates 1974/75, p. (i). 25. Heller, op. cit. For a marginal aid tranche, Heller calculated that 18 per cent of it went to increased government expenditure, but that was made up of an increase in investment of 34 per cent of the tranche and reductions in consumption of 16 per cent. 26. See note 9, p. 228. 27. See Chapter Four, pp. 88–9, and note 26, p. 102. 28. Chapter Four, p. 88 and Chapter Five, Part 3, p. 167, 29. See p. 93. 30. P.S.Heller, ‘Public Investment in LDCs with Recurrent Cost Constraint: The Kenyan Case’, Quarterly Journal of Economics, Vol. 88, May 1974. 31. See Chapter Five, Part Three. 32. The Office of the President received some aid for SRDP personnel attached to it. 33. E.g. over Mumias, see Chapter Five, Part Two, pp. 147–8. 34. Almost all the firms asked for their anonymity to be preserved. 35. Hybrids usually outyield local maize by 40 per cent at least. If the 786,000 acres under hybrids in 1973 had been under local maize, production would have been down by about 2.6m. bags. Over the period 1963–73 hybrids have increased maize output by over 11m. bags. Without this development Kenya could have had severe food shortages in some years. 36. In fertiliser manufacture, for example. See Development Plan 1974–78, pp. 26–8. 37. Ibid., p. 6, para. 1.20. 38. Volunteers have, however, taught in such schools. 39. An example is a letter to The Nation headlined ‘A backward step indeed’, published 11 October 1974. 40. ILO, op. cit., Chapter 14 and pp. 517–28. 41. Law and order is a phrase sometimes used euphemistically by those in power when submission to unjust authority is what is really meant The usage is reprehensible but the concept is not. Law and order is not sufficient for development, but it is quite certainly necessary. 42. M.L.O.Faber and D.Seers (eds.), The Crisis in Planning, London, Chatto and Windus, 1972, Part II, p. 132.

The Influence of Aid

179

43. See Chapter Two, p. 36. 44. See Chapter Two, pp. 25–6 and p. 44. 45. Some expatriates do get involved in the hurly burly, take sides on all issues they detect, and proselytise. Our impression is that such people seldom last beyond their first contract period irrespective of the quality of their work. 46. A full description of the system is given by Laxman Bhandari, ‘Technical Assistance Administration in Kenya’ in Technical Assistance Administration in East Africa, Yashpal Tandon (ed.), Stockholm, Almqvist & Wiksell, 1973. 47. USAID attempts to maintain co-ordination by copying all correspondence with an operational ministry to various divisions in the Ministry of Finance and Planning. 48. At the time of writing, the World Bank was attempting to get a team of no fewer than eight expatriates into the Ministry of Agriculture to help them to spend money on large programmes in that sector in accordance with the current IBRD policy. 49. C.Gertzel, The Politics of Independent Kenya, London, Heinemann, 1970, especially p. 173; also J.R.Nellis, ‘Is the Kenyan Bureaucracy Developmental?’, IDS Staff Paper 103, Nairobi. 50. J.R.Nellis, The Ethnic Composition of Leading Kenyan Government Positions’, Scandinavian Institute of African Studies, Research Report No. 24, 1974. 51. Gertzel, op. cit., p. 17. Nellis, IDS, op. cit, pp. 8, 21. 52. Gertzel, op. cit., pp. 9–10, 93–4, 120. 53. At the 1974 election, the General Secretary of the Kenyan TUC stood as a candidate in an urban Nairobi constituency. He could finish no better than third out of six candidates. 54. See ‘The kind of MP voters want’, by Horace Awori in Joe, Election Review 1974. A partial exception is the case of ethnic groups who regard themselves as out-groups, bound to be discriminated against by the government. They can vote for someone who personifies the group consciousness and opposes the government. An example is the support of the Nandi for Mr Seroney, now in detention. 55. See ‘Systems Management and the Plan Implementation Process in Kenya’ by Michael Chege in African Review, November 1973. Chege, in a perceptive article well worth reading, lists a number of presidential decrees obviously made on the basis of ‘political’ considerations. These decrees are unchallengeable, 56. See Sunday Times, London, 17 August 1975. 57. See Chapter Two. 58. Although his ministry had been involved in a struggle with the Ministry of Education to restrict the number of new secondary schools, the Finance Minister, Mr Kibaki, promised his Othaya constituents more such schools when a candidate in the 1974 election—The Nation, 29 September 1974. 59. The two most outspoken MPs, Mr Shikuku and Mr Seroney, were detained as a result of their questioning of the circumstances of the death by shooting of Mr J.M.Kariuki, a Kikuyu MP noted for his populist approach to land distribution. 60. Article 75 of the Constitution of Kenya expressly rules out compulsory acquisition without compensation. It is often joked by Kenyans that this will be the last part of the Constitution to be amended. 61. See, for example, Colin Leys, Underdevelopment in Kenya: The Political Economy of NeoColonialism 1964–1971, London, Heinemann, 1975. 62. M.Radetzki, Aid and Development: A Handbook for Small Donors, New York, Praeger Publishers, 1973, pp. 244–51. 63. Deepak Lal, Appraising Foreign Investment in Developing Countries, London, Heinemann, 1975, p. 254.

7 POLICY IMPLICATIONS FOR BRITISH AID

Our judgements about a number of issues have appeared in preceding chapters. Here we draw some of them together and bring out the implications for British aid policy in Kenya. How far these implications have validity elsewhere we leave to others’ consideration. Our standpoint is that aid should have the purpose of raising incomes in the recipient country with particular emphasis on the incomes of the poorest. Inequality is acceptable in so far as it serves the purpose of raising the absolute level of income of the poorest, but in so far as it fails to do so, it is to be disapproved. Ideally policies should make a state more able to withstand external pressures and less reliant on external support. Sometimes this objective and that of raising material living standards go hand-in-hand but sometimes, we fear, they do not. Then choices have to be made by the legitimate government of the recipient country. In general terms our preference would be to emphasise living standards where people were very poor, but rarely to the point of sacrificing all freedom of action in any particular area of policy. Such general statements are, however, of little use to someone dealing with policy and confronted with specific trade-offs. In offering policy prescriptions from such a standpoint, we are in effect stating that it is not inconsistent with the notion of development which is dominant in shaping ODM policy. In the past that notion was not clearly articulated and there was no equivalent of the Swedish policy declaration in favour of economic independence and equality. More recently, especially under the past Minister of Overseas Development, Mrs. Judith Hart, an attempt has been made to articulate policy, as in the 1975 White Paper.1 This indicates a wish to focus aid both on the poorest countries and on the poor within those countries. More details of the ODM consensus can, however, be inferred from conversations with officials and the revealed project and sector preferences of the ministry. It is undeniable that ODM is faced with other non-developmental considerations owing to organisational interests within it and the interests of other government departments, notably the Treasury. However these do not seem to be critical, although they will doubtless continue to modify some decisions. The attitude of mind into which a professional aid administrator tends to fall is to ask what are the problems of giving aid in any particular case. In Kenya one answer in the past would have been underspending. The next step may then be to say that the great task facing the administrator is to increase disbursement and eliminate underspending. In beginning this study, we too were tempted into that train of thought. On reflection, however, we believe it is a mistake. At the moment, Kenya undoubtedly faces balance of payments difficulties, hence the current ODM programme grant. In the past, however, that was not so and it may not be so in the future. In that case simply spending money is not a priority. More important is to achieve certain specific developmental objectives which entail ensuring that extra money is well spent—if that is possible.

Policy Implications for British Aid

181

It might be objected that a donor has no right to try to ensure aid is ‘well’ spent by his own criteria, implicitly putting these above the preferences of a recipient government. We take the view that the donor has such a right because it is his money, but that the recipient government should be equally at liberty to refuse aid offered on conditions it finds unacceptable. Then the donor does not elevate his notion of development above national sovereignty and self-determination, as he is not trying to subvert the recipient government, only offering it a free choice. Clearly, the offer of aid can influence policies but that is not a bad thing in itself. Unfortunately, things are rarely so clear-cut. Energetic selling of aid, perhaps using known weaknesses of governments as bargaining points, and offering one thing in the guise of something else can look like dealing between two unconstrained parties, but can in fact amount to an infraction of sovereignty. We do not say that is always wrong—some values can take precedence over national sovereignty—but it places a heavy moral responsibility on the aid donor. One which, in general, we think he should be reluctant to bear. No country’s motives are entirely unmixed in giving aid, so it is bad to set a precedent that could be used for non-developmental ends, and in any case, uncertainties and ignorance are so great in the area of ‘development’ that there are narrow limits on the confidence with which opinions should be held and the methods with which they should be advanced. Our impression is that ODM has stayed well on the right side of the line between having legitimate preferences in the use of its aid and putting illegitimate pressure on a sovereign government. The effort to eliminate underspending will push the ministry nearer to that line, into an area already occupied by other donors. Already technical assistance personnel and donor agencies are too influential in shaping the Kenya government’s development budget. Not only is that regrettable in itself, but it also sets up tensions between this aspect of the government’s activities and others over which the Kenyans retain full control. Donors in general are more worried about the latter problem than that of Kenyan autonomy, hence the desire to increase the numbers of expatriate planners and to introduce sectoral loans as well as project loans2 so that the microlevel influence of aid is being promoted from the project to the sector level. There are, of course, several sound reasons for doing this. The extended leverage which donors exert will, on the whole, probably be beneficial to the majority of Kenyans as long as donors continue to favour a moderately egalitarian and rural-oriented growth strategy. It will enable aid to be spent on sectoral programmes which will include many small items of expenditure, getting away from large discrete projects and perhaps increasing the local-cost content of aid disbursements. It may divert Kenyan manpower from detailed appraisal of narrow projects towards broader sectoral planning, which may be a better use of their skills. However, further penetration of the Kenyan administration and an extension of the area it surrenders to donor decision-making will create new tensions and may set up a xenophobic reaction if there is a change of political direction in Kenya. Many of our suggestions can be seen as attempts to minimise the tension between spending aid, and spending it well, and elbowing excessively the people of the recipient country. Although not without dangers, the move towards sectoral programme support is on balance to be welcomed. It requires, however, a certain forbearance by donors if the dangers of excessive intervention are not be be realised. The aim should be to use leverage not to get an optimal plan from the donor viewpoint, but a self-consistent, feasible and (to the donor) tolerable plan. It is true that Kenyan planning capacity is strained by very large

182 Aid and Inequality in Kenya: British Development Assistance to Kenya exercises but there comes a point where learning can only be by doing and a willingness by donors to accept slow disbursement and the risk of error is probably in order. The threat of no aid or slow aid should be a goad to the Kenyans to improve their planning and administrative capacity, using foreigners for training if necessary, not a blackmailing point to get them to surrender functions to expatriates. Unfortunately it is difficult for one donor to follow this advice if others do not. For project aid, if ODM advisers on their travels spot possible projects, this could, of course, be put informally to Kenyan colleagues, but an approach should then be left to them to be routed via the Treasury. As noted, the Rural Development Fund could be viewed favourably because of the possible benefits to project identification from decentralised planning. If approaches for aid coming via the Treasury appear promising but half-baked, it is only realistic for Development Division officials then to have talks with the operational ministry and to assist with a re-submission—keeping all relevant sections of the Treasury informed. An early agreement to help in principle, followed by active involvement, also maximises the micro-level influence of aid over projects. If a great deal of preparatory work is required, consultants could be used or perhaps ODM headquarters staff seconded temporarily, rather than technical assistance personnel being put into the Kenyan ministry. Working with visitors in that way is likely to be at least as instructive for Kenyan officers as having expatriates in to do the work. There may be occasional exceptions to this policy— e.g., for high-technology projects like the Tana River development—but these ought to be extremely rare. In our view, the whole philosophy of technical assistance ought to be reconsidered. The aim should be to make an input that can be ‘amortised’ rapidly, but which will continue to yield benefits thereafter. In other words, to achieve not a once and for all improvement in recipient government efficiency which collapses when the assistance is withdrawn, but permanent expansion in the technical capacity of the recipient. The training function of technical assistance personnel (as most of them stress) is therefore the most important aspect of their work. It has not been the only aspect. If the ha’porth of tar needed to complete an invaluable ship is an expatriate, it makes obvious sense to supply him, but twelve years after independence this ought to be a consideration of declining importance. Many things which, in a developed country, a good social democrat would regard as quite properly the function of the state, lie outside the competence of the state machine in a country like Kenya. This might include the setting up and manning of large public corporations to produce marketable goods and services, the provision of welfare services with complex rules for allocation, or certain kinds of centralised planning. Ideally, bureaucratic ambitions should not outstrip bureaucratic competence, and we do not see that it is necessarily a good thing for poor countries’ administrations to be in a state of constant or even increasing ‘personnel debt’. This can happen if they are encouraged to tackle enterprises by the offer of capital aid and then technical assistance is inserted to try to close the gap between competence and aspiration. Aid should assist the local administration to push at the borders of the currently possible, not provide foreigners to run the border out of sight, Consequently, we applaud the ODM policy of running down OSAS and we doubt whether there should be any compromise on the issue. ODM’s feeling that it is futile to check individual requests for SCAAP personnel is quite understandable. Even here, however, a general policy of scepticism should reign, and the Kenyans might be told that requests for

Policy Implications for British Aid

183

support for training facilities will be looked on more favourably than requests for advisers or operational personnel. In some areas ODM has done this already. One example is an offer to support teacher training facilities while phasing out OSAS teachers. Many professional skills are in short supply in Kenya. Apart from teachers, there are serious shortages of trained accountants, tax and revenue officers, investment analysts, as well as many sorts of technician. Devising and supporting basic professional training schemes for educated but unspecialised local people should be a top priority. An imaginative attempt in this area was a British Council-sponsored plan for teacher training colleges in the UK to second some of their best staff to Kenyan institutions with perhaps some return facility for Kenyans. Such devices would be preferable to full-time employment of contract personnel to work in Kenya. Sometimes excellent people can be obtained, but often they tend to be either relatively young and inexperienced in their profession, or else to be relatively unsuccessful. Manpower planning reviews should be exercises to identify training needs as much as specific personnel requirements. The need for training is shown by the fact that while Kenya has a school-leaver unemployment and even an arts graduate unemployment problem, the Kenya government cannot recruit enough middle-level staff. Turnover rates and the ‘brain drain’ to the private sector are high. Some expatriate advisers believe technical assistance is popular because, net, it is cheaper to the Kenyan government than employing a local man. On this view, if donors untied their technical assistance the Kenya government would recruit more Kenyans, reducing the numbers of foreigners in public sector employment. ODM may be sceptical but then why not try it and see? Just as good, perhaps, would be more training to increase the supply and so perhaps reduce the cost of competent Kenyan administrators and professionals. The past pattern of technical assistance, whereby an expatriate did a job and had a local ‘counterpart’ who picked it up from him as he went along, was an attempt to kill two birds with one stone. It bears a strong resemblance to ‘apprenticeship’ training schemes which have done such incalculable harm to British industry. The competence of the apprentice is postponed by not giving him structured training and, when he has acquired as much competence as he is likely to get without tackling the job himself, doubts about his competence keep him out of it. There is a further parallel: in British industry few want to be apprentices, and the Kenya government does not usually bother to provide counterparts for technical assistance personnel. Both systems thus perpetuate, rather than eliminate, shortages of skills. The invaluable training functions of technical assistance personnel should be discharged more directly at the expense, if need be, of other duties. These other duties are also best confined to the specific and the short-term. As noted, we cannot imagine that Kenya needs more expatriate planners, for example. Some functions of technical assistance personnel would be better carried out by having more officials in ODM Development Divisions. That is so where technical assistance is given, in effect, to supervise the use of funds and proffer good advice rather than do a specific technical task. Such monitoring, observation and advising is more authoritative coming from a donor operative. That would not look good on paper, of course—reducing output of the aid programme in order to increase its overheads. However, we are sure in many cases—as was true of SRDP—it would pay. We have not studied the structure of ODM as a ministry but it may well be that it could discharge its duties even better with

184 Aid and Inequality in Kenya: British Development Assistance to Kenya a different balance between overseas and London-based staff. There is certainly plenty of work to be done in East Africa, perhaps yielding higher returns than work now being done in London. In order to achieve a better geographical distribution of its manpower, perhaps ODM should obtain the authority to post staff abroad instead of relying on their volunteering, as at present, An alternative might be a larger corps of experts directly employed by ODM to be sent to Idcs like Kenya to carry out specific short-term engagements at present done by people while on contract as technical assistance. Such a corps could also conduct ex post evaluations of aid projects. They could be financed by saving some money on longer-term contract technical assistance. This bears some resemblance to suggestions for providing a career service for technical assistance personnel in order to improve the quality of workers in the field. This was suggested at least twelve years ago;3 there were few good reasons against it then, and there are no more now. An argument sometimes presented is that the government cannot guarantee employment for people hired specifically to spend most of their time working in developing countries. It is inconceivable, however, that there will be no need for an aid programme during this century, so the argument has little force. At least ten years after it was advanced there are still well over a thousand publicly-financed British expatriates in Kenya. There would be a number of advantages from ODM employing people and using them, perhaps in groups, on technical assistance projects of months rather than years. The people would have career records and a known competence in particular fields so the risk of sending duffers would be reduced. Reporting and feedback on the operation of projects would be improved. It would be a more attractive job, perhaps, than that of a two-year expert with no job security and even sometimes no quotable references afterwards. Our attitude to aid strategy then is that the aim should be to make Kenya as self-sufficient as possible in personnel as quickly as possible and to continue to exercise as much leverage as necessary over the use of capital aid from the outside, rather than by a ‘fifth column’ of technical assistance. That situation, in which everyone’s interests and responsibilities are clear-cut, is likely to be less invidious in the long run. That brings us on to the awkward matter of ‘style’ in aid administration. It is strange to find the impression abroad in the Kenyan administration of Britain as ‘perfidious Albion’. Yet so it is. Many Kenyan administrators respect ODM for its greater knowledge and experience of local conditions. They acknowledge it is less likely to make mistakes than other donors, yet they find it unpredictable and suspect it of using deliberate delay to kill projects. We must regretfully conclude that, in general, the British are no good at diplomacy, so perhaps they should give it up. When they fondly imagine they are being diplomatic, they are often suspected of being merely sly. Excessive diplomacy is after all a way of being patronising. We are sure the situation is improving as Development Division officers make closer personal contacts with the Kenyans, but mistaken attitudes linger, as evidenced perhaps by an excessive secrecy in ODM’s dealings. Evidence about style must be anecdotal. The unkindest word we heard used about British negotiating methods was ‘squelchy’, by an expatriate member of the Kenyan administration who related how a meeting over a project had dragged on with British officials being pleasant but faintly elusive. Afterwards one had confided to him: ‘Of course, we could never really give a penny for any project in that form.’ That is not an isolated story. Difficult as it may be for

Policy Implications for British Aid

185

the British temperament to understand, the Kenyans, at least, prefer the downright dealing they get from USAID or the World Bank. It is hard to change the habits of a lifetime, but we strongly recommend greater openness in negotiating, in disbursement planning, and in divulging the contents of the CPP both to the Kenyans and to other donors. The latter would also have the advantage of making ODM articulate more of what it is trying to do and submit it to critical scrutiny. There is another advantage to being blunter. It puts a premium on making sure that the things one is going to insist on are clearly sensible. Many of the conditions which ODM puts on its aid are eminently sensible. We can find little to criticise, for example, in its wish to control the end uses of its aid or in its sectoral preferences. Procurement-tying is, of course, a nuisance that is largely pointless when the pound is not overvalued, but our impression is that it is the fault of the British Treasury and ODM would be as glad as anyone else to see an end to it. Its one advantage might be that, as it makes the Crown Agents buying agents instead of leaving that to the Kenyan ministries, it might reduce the scope for corruption over contract tenders. There is no reason, however, why ODM should not appoint buying agents to consider tenders from anywhere, as the Crown Agents can when so authorised. One rule that really is absurd, however, is the one against financing recurrent expenditure. We really think that if ODM had to operate a rule against financing either recurrent or development expenditure, the latter would be less damaging than the former. The Kenyan administration has shown itself capable of running sensible recurrent programmes, and it could administer expansions of them as long as these were on a modest scale. This organic growth of productive expenditures would not pose the administrative problems that the demand for projects and new large integrated programmes does, although it would impose others of control and supervision by donors. Everyone should realise by now that recurrent expenditures can be as productive as investment expenditures. Furthermore, they do not imply an open-ended commitment by a donor. Any investment almost certainly implies continued recurrent expenditure when it is complete. In the case of, say, a bridge, the recurrent expenditure may be quite modest—painting and maintenance—in the case of a school it will be considerable, quickly outstripping the original investment sum. Projects consisting of recurrent expenditure have nothing unique about them at all in that regard. For the donor commitment not to be open-ended, all that is required is for a phased take-over of costs to be arranged at the outset. The Swedes already do this and have encountered no problems. The nature of the accounting conventions covering public expenditure should simply be ignored by ODM when assessing Kenyan proposals. The question to be asked is would this be done if we did not support it? If the answer is probably not, it does not matter how the Kenyans or anyone else classifies it. Another rigidity which ought to go is the classification of the aid programme into separate funds. We have suggested a single sum in aid agreements for disbursement as capital or technical assistance. However, we would go further. If it is clear to both ODM and the Kenyans that a given agreement sum is not going to be disbursed within its terms, why not an arrangement to on-lend the balance to CDC to be invested in Kenya if opportunities exist? No doubt the Kenyan government would prefer to have the money, but a private sector use is better than no use at all from its point of view. Such a transfer could

186 Aid and Inequality in Kenya: British Development Assistance to Kenya scarcely become a regular or perhaps sizeable feature of British aid because CDC might well have no project opportunities appearing in Kenya when ODM had spare cash, and the corporation is not geared to finding uses quickly for odd funds in particular countries. However, an awareness of the possibility of a match could be fruitful and if the money were lent, CDC could repay it to ODM for credit to Kenya after five, ten, or fifteen years. The point is that the British government wants to spend money on productive enterprises in Kenya and has various channels for doing so. It is pointless to have rules or accounting procedures that make that more difficult. Despite the fact of fungibility, we would not advocate the abolition of all end-use-tying of aid to Kenya, Even if donors were less intrusive, Kenyan forming of priorities would probably remain haphazard to some extent, at least under the present regime. Granted that, however, the problem of fungibility has not been given enough consideration by ODM. In considering support for any project, capital or technical assistance, the first question to be asked is, will anyone else do it if ODM turns it down? If the answer is probably not, and the project is substantial, it then makes sense to ask the question of whether it is worth doing by carrying out fullscale appraisals and feasibility studies. If these lead to the conclusion that the project in some form is worth doing, a third question then needs to be asked, namely, given that no one else would have carried out the project, is there enough Kenyan political support and expertise for it to operate satisfactorily after ODM moves out? If the answer to the first question—will anyone else do it?—is probably yes, the second question ODM ought to ask is—why should we do it? There may be good developmental answers to this question. ODM might believe itself to possess some particular expertise which the Kenyans and other donors could not match, or it might think that the micro-level influence it could exert would be different from, and preferable to, that of any other donor in that particular case. It is necessary to realise, however, what that influence would be costing. In order to exercise influence over the form of some project, ODM would be giving fungible resources, and with the scale of Kenya’s importing from Britain, that practically amounts to free foreign exchange. In any particular case it should be simply considered whether the influence is worth it. This requires a slight reorientation of attitude by ODM staff. Aid agreements are not money already down the drain; they are pieces of paper. Subsequent disbursements are the real cost to Britain, and micro-level influence should be weighed against them. It may well be, of course, that ODM comes under pressure to aid a project with fungible resources for reasons that have nothing to do with development. The project may have a higher import content than any alternative project, or it may bring political kudos with the existing regime. It is idle to pretend that such considerations do not have some place in aid administration, but ODM pursues them only at risk to developmental considerations. In carrying out a project that some other donor would execute, it leaves to that other donor the greater macro-level influence over the allocation of resources. He then may operate in the region of the marginal projects, and his influence may actually be critical in seeing that one such project rather than another comes to pass. If his policies are like ODM’s, no one may mind. In general, however, if donors are queueing up to carry out a project, ODM should consider letting them get on with it. If, however, ODM decides there are sound developmental reasons for aiding a project which would have been done anyway, it is clearly quite pointless to insist on evaluations and appraisals before finally agreeing to aid. On the contrary, the whole point then is to

Policy Implications for British Aid

187

agree immediately and tell the Kenyans the money is certainly earmarked. As much help as possible should then be given in the preparation and planning of the project, in order to achieve the micro-level influence which is presumably the whole point of the exercise. If ODM suspects for a moment that there may be any difficulty in exercising such influence (which would be unusual), there can be no developmental point whatever in aiding the project in question. It would be quite unethical in our view to offer to aid a project that would be aided by another donor merely to delay its implementation or to emasculate it with economies.4 There is an asymmetry in that one can offer to aid a project in order to put in more resources than would otherwise have been devoted to it. One cannot however ‘aid’ a project in order to eliminate or greatly reduce it without being guilty of improper interference. One can state to the recipient that a project is misguided or aid it believing that one can cut costs while maintaining ‘output’, but that is a different matter. Assisting sectoral plans with sectoral programme aid means that fungibility is still a consideration but it is likely to be less of one than with project aid for two reasons. One is that a programme is more segmentable than a discrete project; some bits could be done without others. A sectoral grant, therefore, is quite likely to increase the scale of government spending in the area concerned, to bring it forward in time, and to protect it from economy cuts when these are required in government spending. The aid is likely to be only partly fungible. Secondly any microlevel influence which the donor can exert to improve a sectoral plan, in its own terms, is likely to be more important in terms of overall allocation of resources in the economy, than is the case with project aid. There are a few activities which receive aid, which the Kenya government would undertake anyway even if there were not a donor in the world. Land adjudication and registration may well come into this category. It is a continuous exercise for which the Kenyans have all the necessary expertise. There are strong Parkinsonian bureaucratic reasons for it to go rolling on, and there is political push behind it too. It is largely a localcost programme so UK aid for it is simply free foreign exchange. If ODM wants to extend some programme aid to Kenya as a token of good faith because it is difficult to spend all agreed aid on definite projects or programmes, we can see no great objection to doing so, but there is no reason for tying it to land adjudication. It is not even clear that in doing so, ODM is associating itself with a winner, We have made no special study of the programme and its effects, so we rely on the balance of informed opinion among agriculturalists and academics to whom we spoke. The consensus appears to be that the programme was useful in its early years in sorting out claims to land in densely populated areas. Its benefits were probably somewhat over-rated, as it was untrue that no one could get agricultural credit without land title—it was possible with other collateral—and it is untrue that anyone can get credit with land title—other collateral is still regarded as necessary by most lenders. So one of the claimed benefits for adjudication and individual freehold is a myth. Nevertheless, no doubt the programme has yielded other, perhaps psychological, benefits. But these must be declining as it is pushed into less and less densely populated areas where it will not change the methods of semi-nomadic pastoralists. Even in more densely populated areas, it appears the land register becomes rapidly out of date owing to unofficial transfers and subdivisions.

188 Aid and Inequality in Kenya: British Development Assistance to Kenya We conclude that support for the land adjudication programme has become simply a way of disbursing aid so that underspending does not reach too embarrassing a scale. We recommend an end to aid specifically for the programme. A modest amount of general programme aid as part of future aid agreements would be more suitable. That would not necessarily even mean that adjudication would be prejudiced but nor would it have the effect of fortifying the adjudicators, if other elements in the Kenyan administration ever decided to take a long hard look at the programme. Would the Land Transfer Programme also continue without aid? We believe it might, at least under the present regime, although the prices offered for farms might well fall. We cannot see any sound developmental reason for maintaining the land price in Kenya. There may be good arguments for subdivision of holdings, but there are now many large Kenyan holdings which could be subdivided and there is no need to look exclusively to expatriate farms for land to be allocated to the poor. The present tranche of aid for land transfer was supposed to be the last. In 1970 about 200 British citizens were thought to be still farming in Kenya, and 150 of them had landholdings that might qualify as mixed farms. It is in fact unlikely that the present tranche will be enough to buy out all those who want to sell. There are three reasons. First is an inflation of land prices; the tranche was worked out assuming a price in the region of £20 an acre, but £30 an acre is nearer the going rate. Second is the fact that some Kenyan citizens may opt to revert to British citizenship, as they have the right to do, and their farms then become eligible for transfer. No one really knows how many people are likely to do this, but as whites become more exposed in the Kenyan countryside and their postcolonial way of life changes with the departure of European neighbours it could become a frequent occurrence. The third reason hinges on the definition of mixed farm. It is an exceedingly complicated one that is execrated by British and Kenyan civil servants alike. Unfortunately a degree of mutual suspicion has so far prevented successful negotiation of a new definition. The existing one specifies that the farm must have more than 25 inches of rainfall a year, have more than 25 per cent of the land under cash crops, be 50 per cent arable land, and have not more than 50 acres or 25 per cent of the land under plantation crops. The Kenyans point out, with reason, that a lot of basically mixed farms do not fall under this definition, having too much coffee or a slightly higher proportion of pasture or arable land, but that if this definition had been applied from the first some 40 per cent of the Million-Acre Scheme land would not have been eligible! (Kinangop, for example, one of the most politically sensitive areas, was often less than 50 per cent arable.) The British are loath to abandon the definition for fear of incurring an open-ended commitment to buy out all Europeans, including ranches and plantation owners. The Kenyans want the definition replaced with a clause simply stating that the farm is ‘suitable for settlement’. Ideally they would like it left up to the Department of Settlement to decide whether a farm qualified, although it could be done by a joint committee of both sides, a procedure already used in the case of Compassionate Purchases. A difficulty here, however, is that while ‘suitable for settlement’ meant something when settlement implied subdivision, as the land had then to be sufficiently arable to support peasant small holdings, now with Shirika it can mean anything. Even ranches could be taken over and run on a co-operative basis.

Policy Implications for British Aid

189

Ranches, however, cannot usually support more people than are currently employed on them, and though some plantation crops such as tea and coffee have been successfully grown on small holdings, there is limited scope for making production of these crops more labourintensive and increasing yields per acre. With current policies, at least, there does not seem to be much economic or social rationale for an extended programme. ODM has long since surrendered any attempt at micro-level influence in land transfer; it was not consulted over the Shirika schemes, for example. Therefore we can see no developmental reason for the continuation of aid for this programme beyond the current aid agreement. It is simply transferring assets without much changing their economic use or contributing to political stability. That applies a fortiori to any extended programme to cover acquisition of ranches and plantations. We say that, having argued as strongly as anyone, in Chapter Five, that the Land Transfer Programme constituted extremely valuable aid to Kenya in the past. However, there is no longer the danger of a precipitate flight of most of the country’s best farmers. There are, of course, non-developmental reasons for going on with the programme which HMG might think potent. People unable to take advantage of it because of the mixed farm definition, whose farms are mixed, could reasonably feel discriminated against, especially if a future Kenyan government expropriates expatriate landowners. However, those British farmers still in Kenya must surely have written off their original investment by now. Human problems are involved about which we have no wish to be callous, but further money for land transfer would be underwriting the assets of people who, by and large, have not done too badly. Apart from the human problems, there are political problems in stopping this sort of aid and HMG must search its conscience. All we would say is that further money for land transfer is without developmental merit. So there is no reason why any further sums should be a charge on the aid programme. As with adjudication, some of this money might in future be re-allocated to untied programme loans, accounting for some part of the total aid programme to Kenya. There has been a considerable improvement in the area of donor cooperation over the past few years. However, it should be clear that many of the suggestions we make would be more effective if there were closer liaison between donors. Fungibility is largely a problem in Kenya of the substitutability of one donor’s aid for another’s. Some knowledge of each other’s preferences and agreed areas of specialisation therefore greatly reduces the risk of it. There is also more point in ODM winding down its technical assistance commitment if it is confident that other donors will not unthinkingly replace British personnel with their own. Our suggestions for a more open administration of the aid programme apply to relations with other donors as much as to relations with the Kenyans. Greater openness would also reduce Kenyan suspicions of donor collusion. The traditional reasons for donor co-operation—elimination of duplication of effort and wasteful competition, and the opportunity to match complementary skills and inputs—apply with undiminished force. Perhaps the UK could seek to invigorate the East Africa Donor Consultative Committee. Strangely enough, it is traditional for researchers to conclude reports by calling for more research. Without being too predictable one can point out that in Kenya some of aid’s greatest successes have had a strong innovative element. Maize provides an example of successful research but there are also concrete programmes, such as CDC tea, which owe their achievements to the development of new techniques suitable to the local situation and factor endowments. Donors should, therefore, be willing to accept risks in financing applied

190 Aid and Inequality in Kenya: British Development Assistance to Kenya research and innovation. They are often better placed to do this than either commercial concerns, out to minimise risk by following tried techniques, or recipient governments who may be highly risk-averse both because of poverty and the need to avoid providing ‘hostages’ to political opponents.

Notes 1. 2. 3. 4.

The Changing Emphasis in British Aid Policies: More Help for the Poorest, HMSO, Cmnd. 6270, October 1975. The World Bank and USAID have already made sectoral loans, and Sweden and West Germany as well as ODM want to move in this direction. See, for example, I.M.D.Little, Aid to Africa, London, ODI, 1964, p. 69. We do not suggest that ODM has followed this course.

8 CONTROVERSIES OVER AID

The previous chapters analysed the effects of aid on Kenya and gave policy prescriptions. This final chapter presents our conclusions in a different context and examines some general propositions about the purpose and effect of aid in the light of the Kenyan experience. Views about aid range from the over-simple altruistic, that it is a straightforward act of charity, to the over-simple Machiavellian, that it is the pursuit of colonialism by other means. Despite its obvious appeal to the better instincts of the people of the rich countries, economic aid has been much attacked. Aid is unpopular with the extreme of both left and right. Even where there is no ideological viewpoint to generate hostility to aid, its failure to transform the state of the poor countries has caused disillusionment among some who underestimated the difficulties in the way of their economic progress. The naive altruistic view of aid is inadequate because it ignores the mixture of motives in the donor country, assuming a single-minded benevolence, and ignores the possibility that aid can have detrimental effects on the recipient. The Machiavellian view of aid is inadequate because it ignores the mixture of motives in the donor country, assuming a singleminded and narrow self-interest, and excludes the possibility that the beneficial effects of aid will predominate. Curiously, this view also excludes the possibility that aid can benefit the recipient even though it also serves the interests of the donor. In reality, donor policies reflect the balance of diverse views and interests, and the effects of aid on the recipient reflect the balance of beneficial and adverse effects. Both these over-simple views of aid policy are inadequate for another reason: they ignore the effect of ignorance and unintentional error. It is itself a great error to assume that the path to development is known and charted, and that it is simply a matter of the will and the means to travel along it. A donor may, indeed, pursue a particular policy for non-developmental reasons: it may tie aid, for instance, in an attempt to stimulate its exports. A recipient may also adopt a policy for non-developmental reasons: it may, for instance, decide on the location of a particular project because it benefits places of political importance, or even because it benefits influential individuals. But donor and recipient may take a different view about development policy, about the regional and sectoral distribution of investment, for instance, each for perfectly genuine developmental reasons. Either view, and indeed both, may be mistaken, and the policy implemented may be judged on balance to be detrimental. But it would be a mistake to attribute such failures to anti-developmental motives rather than to ignorance and error. Our study of British aid to Kenya does not incline us to an extreme view of either the benefits or the burdens of aid for the recipient country. Nor has the fact that Kenya remains a poor country caused us to conclude that aid has been a failure. Our judgement is that in British aid policy towards independent Kenya the developmental motive has increasingly predominated. When better policies could have been pursued, and

192 Aid and Inequality in Kenya: British Development Assistance to Kenya better projects supported, it is easier with hindsight to see the better way than was possible at the time, and the deficiencies have been generally the result of faulty judgement, rather than of bad intention. On balance, in the effects of aid on Kenya, the benefits have vastly outweighed the costs. One proposition about the allocation of aid is that it is given to compliant governments, those that follow the dictates and support the policies of the donor. Such a government receives aid because it is compliant, and is compliant because it receives aid. A difficulty in fitting the facts of aid to Kenya into the model of aid to a compliant state is that there are so many different donors whose interests and policies are not uniform. Can Kenya be compliant to them all? It is not as if donor cooperation were so close and strong that a single donor policy prevailed. It was suggested in Chapter Six that donor co-operation has been weak in fact, and there are still possibilities for Kenya to play off one against another. The theory of the compliant state does not stand up to investigation in terms of Kenya’s domestic policies. It is even more difficult to fit Kenya’s international policies into the model, and to demonstrate that Kenya is not only a compliant but also a client state. Where the international interests and policies of donors differ as widely as those of, say, Sweden and the United States, it is impossible to give credence to such an idea. Of course, on some large issues donors agree. They might, for instance, unite in opposition to a wholesale expropriation of foreign capital, and the pursuit of such a policy by the Kenya government might reduce the aid allocated to Kenya. As an explanation of the allocation of aid, therefore, to that extent the theory of compliance has some basis. But the other leg of the proposition, that governments are compliant because they receive aid, does not follow. It does not follow that Kenya fails to expropriate foreign capital—to continue the example—because of the aid it receives. For one thing, there are too many private Kenyan interests, to which the government is sensitive, which would be adversely affected by such a move.1 But it is not only a concern for particular Kenyan private interests which could justify the attitude of the Kenya government to private capital. It is perfectly possible that an objective assessment would lead to the conclusion that the benefits of foreign investment in Kenya outweigh the costs. It is a flaw in the compliance theory, when used as a criticism of aid policy, that donor policies are assumed necessarily to be against the social interest of the recipient. The historical roots of British aid policy are far too deep and complex for a simple ‘compliant state’ explanation of aid to Kenya to withstand even the most superficial examination of the facts. It is true that the origins of the aid programme may be traced to the time when the extreme of compliance—colonial status—existed. It is also true that the Land Transfer Programme was instituted at a time when Kenya had not yet achieved independence, and when the acceptance of the programme by the African leaders cannot have been uninfluenced by their desire to ease the negotiation of independence. But Kenya has not sold herself for aid, nor has aid been given in the hope of buying her. Britain’s aid to Kenya is the result of history and of a British interest, real or imagined, in the maintenance of law and order, and in Kenya’s development. It has been argued that: part of the explanation why aid has not led to faster development is that it is not designed for this purposes. That is, the major purpose of aid is to further the interests of the donors rather than those of the recipients.2

Controversies Over Aid 193 That is not an adequate assessment if applied to Britain’s policy in aiding Kenya. It would be even more of a caricature to present Britain’s aid to Kenya as being ‘from a metropole to a satellite’ and as ‘a way of safeguarding relationships with a client state’.3 Such an assessment is sufficiently refuted by the general diffidence of ODM in its relations with the Kenya government, and by its tendency to respond to requests rather than to take the initiative in project proposals, for fear of being thought to push projects in the British interest. It is also refuted by the occasional differences that have arisen between the rather narrowly ‘developmental’ view that has been predominant in ODM and the less austere view of what projects should be supported in the interest of political goodwill to Britain in Kenya. Another proposition about the allocation of aid—that aid is too often given to ‘bad’ governments—is somewhat at odds with the compliance theory. Of course, it is not at odds if it is believed that the donor is also ‘bad’. But from a less demon-ridden point of view it is sometimes suggested that donors do not sufficiently refrain from aiding countries of whose governments they disapprove. There are many tests by which a government may be thought ‘bad’ and unqualified to receive aid. It may be externally aggressive, internally oppressive and lacking in popular support; its policies may frustrate development or foster the growth of inequality and the impoverishment of sections of the population; its administration may be so ineffective or venal that it cannot properly administer the aid it receives or prevent its misappropriation. We do not think that the Kenya government would score anything like zero in all these tests of unworthiness, but nor do we believe that its score of black marks would be notably high, as compared with that of many countries. It may in any case be questioned whether it necessarily follows that a donor should refuse aid to a government of whose policies it disapproves. If it did so the donor could adopt a high moral tone, but it might worsen the lot of the people in the country deprived of aid. Should a donor exert its influence to improve the policies of the recipient government? ‘Leverage’ by donors is commonly condemned; it is often thought of as being exerted where a recipient government is not naturally compliant, but can be levered into compliance. Aid is then a cog in the machinery of domination. The condemnation of leverage could derive from a belief in the virtues of independence, even if it meant going your own way to perdition, but more commonly it derives from the hidden assumption that leverage is always exerted against the interests of the recipient country. But if the donor is ‘good’ and the recipient government is ‘bad’, leverage could be exerted with beneficial effects. In fact, of course, this approach to leverage is entirely too simple-minded. The uncertainties surrounding development policies are such that serious differences of view could exist between a donor and a recipient, when the concern of both was with development and economic improvement. Leverage is often about means rather than ends, and an attempt by a donor to exert leverage does not imply that either the donor or the recipient must necessarily be characterised as ‘bad’. In practice, on the occasions when ODM has tried to exert leverage on Kenya, it has generally been in the interest of development rather than in Britain’s more direct commercial or political interest. Sometimes the leverage has been ineffective, and sometimes unnecessary. The simple stereotype of the reluctant but weak recipient being levered into pursuing policies in the interest of the powerful donor is far from the reality of the aid relationship between Britain and Kenya.

194 Aid and Inequality in Kenya: British Development Assistance to Kenya The policy on land transfer and the inclusion in the Kenya Constitution of rules protecting investors could be seen as the consequence of leverage exerted by the outgoing colonial power. But the official attitude towards private property in independent Kenya suggests that in these respects Britain may have been levering an already open door. Since independence Britain may be detected exerting leverage to resist an application for aid for the development of television in Mombasa, to accelerate the reduction in the numbers of OSAS personnel, to prevent an extension of the Land Transfer Programme through redefinition of ‘mixed farm’, and for a greater emphasis on rural development by indicating preference for projects concerned with this. In none of these cases can the leverage be seen as outstandingly successful. The lever was either too weak, was not pressed hard enough, or was superfluous. Funds were eventually given for Mombasa television; the number of OSAS personnel is falling at a slower rate than the British aid authorities think desirable; it would not be surprising if Britain relented in its determination not to extend the Land Transfer Programme. In the case of a greater emphasis on rural development, the Kenya government would claim that this was its own policy, and that no leverage was needed. In none of these instances—successful or unsuccessful, necessary or not—can the attempt at leverage be seen as being made in an anti-developmental cause. Nor does it necessarily follow from this fact that the policies against which leverage was exerted were themselves anti-developmental or pursued for non-developmental motives. There can be a sufficient diversity of judgement on the effects of a slower rundown in OSAS staff and of an extension of land transfer for opposite policies to be supportable on developmental grounds. Aid policies and motives have been judged too much in terms of absolutes, of good and bad, and too little in terms of the balance of judgements. In any case, these are pretty anaemic examples of leverage. It is only a negative kind of leverage to refuse funds for one kind of project and to express a preference for another kind. The same may be said of several other decisions: the refusal of aid for Phase II of the expansion of Kenyatta National Hospital, because it was judged that other forms of expansion of health services would be more beneficial; the unwillingness to supply funds for grain storage in the absence of an explicit policy on storage. It could be argued that Britain should have exerted more positive leverage to improve Kenya policies—for more egalitarian policies on land tenure and settlement, for example. In fact, there is an instance of leverage being used in this way. The settlement programme included the provision of socalled ‘Z plots’, which would be larger than was usual under the High-Density Scheme and which included the farm house of the old European farm. It was suggested by the settlement planners that the occupants of ‘Z plots’ would act as community leaders and would set an example to their neighbours of good agricultural practice. The judgement of the British aid authorities was that this form of settlement would not yield the suggested benefits, that the plots would often be taken by absentees, and that they would add to the inequality in the distribution of wealth. The provision of further funds for land transfer was made conditional on the abandonment of the ‘Z plots’ scheme, and the settlement authorities did, in fact, abandon them. This was direct leverage to change Kenya policy by exerting the influence of an aid donor. It should not be presumed, however, that such leverage by a single donor would always be effective. The characteristic of this example is that there was no other aid donor which would take on land transfer if Britain refused further funds. The existence of several potential donors for most projects makes land transfer a very special

Controversies Over Aid 195 case. After 1966 Britain used leverage to resist extensive subdivision of transferred land, a policy with disinterested intentions but inegalitarian effects. The rarity of cases like the Z-plots does suggest that an over-nice regard for the autonomy of the Kenyan government in such fields of policy could have been a fault on Britain’s part, but the end of subdivision shows the dangers of an eagerness to lever Kenyan policy in the direction that Britain thinks it should go. Leverage of a kind, of course, may be exerted by the simple decision of a donor to support one project and not another. Unless there is full fungibility, the pattern of investment in the recipient will be affected by the decisions of the donor in giving project aid. Projects will then be undertaken as a result of project aid that would not be undertaken if the recipient were carrying out the same total expenditure on development from its own funds, or with the support of programme aid. Our view is that some British aid, probably most, has been shunted, but some has not. For some projects the funds were fungible. These include the Mumias sugar scheme and the improvement of the Mombasa road, which would almost certainly have been carried out in the absence of British aid, or even of aid funds from any source. Some British aid funds were not fungible. Funds for land transfer were not fungible, because in the absence of British aid the farmers would not have been bought out in the same way and on the same terms. They were not fungible, we suspect, for SRDP, because the absence of British funds would most likely have meant that one area would have been omitted from the scheme. Because of limitations in the capacity of the Kenya administration to administer and execute projects, the acceptance of aid for projects may sometimes have precluded or postponed the execution of some other projects. Aid as a whole, of which British project aid is a part, is therefore open to the comment that: the donors are able to pre-empt domestic resources and alter the entire investment programme, thereby substituting their preferences for those of the recipient government.4

It is not clear that such a change in priorities must inevitably be antidevelopmental. Why should it be assumed that the donor always knows worse? What if the preference of the recipient is for Mercedes for ministers? But in fact there may not be wide differences between the preferences of Britain and Kenya. Britain selects projects from the list offered by Kenya, and it is not certain that Kenya has strong priorities within the list, and there are always other donors. Hence the leverage nominally exerted through project aid may not be effective, and if it is effective our examination of the projects selected for British support from among those offered them by the Kenyans does not suggest that it should be condemned. But leverage can be more subtle and insidious than that exerted as overt pressure by the donor on the policies of the recipient or by the selection of projects under project aid. It is possible to consider as leverage the introduction of the various distortions in the economy and society of the recipient which have been alleged to be the consequence of the acceptance of aid on a large scale. ‘External doles’, it has been claimed: ‘tend to bias the development process in directions based on external prototypes which are often inappropriate and therefore damaging.’5 A similar thought is contained in the suggestion that ‘aid may have retarded development…by distorting the composition of investment’.6 These effects of aid might be thought of as ‘implicit leverage’.

196 Aid and Inequality in Kenya: British Development Assistance to Kenya One alleged effect of implicit leverage is on the centralisation of power. It has been said that ‘foreign aid augments the resources of the recipient governments compared to those of the private sector, there by promoting concentration of power in the recipient countries’.7 There has certainly been some increase in the relative size of the public sector of the Kenya economy since independence.8 Aid is a direct addition to the resources of the government, so that it is to be expected that aid should increase the size of the public sector and the economic role of the government. However, if the inflow of aid and the facilities provided by aid encourage an inflow of private capital and an increase in domestically financed private investment, it is not inevitable that the relative size of the public sector should increase. Nationalisation of existing enterprises would increase the relative size of the public sector, and this has had some part to play in Kenya. Although the Kenya government has adopted a pragmatic and not an ideological attitude to nationalisation, it has taken some enterprises into public ownership—the Power and Lighting Company and the oil refinery, for example—and there is no evidence from Kenya for the belief that ‘aid agencies…tend to discourage government ownership’.9 But neither is it true that there is a simple connection between aid and the concentration of power. If the centralisation of the economy and the concentration of power are to be deplored— if, so to say, small is beautiful—it would be an error to attribute the fault solely to the increase in government economic activity fostered by aid. However objectionable ‘the insolence of office’, the concentration of power in private corporations should not be forgotten. The argument for decentralisation against centralisation, for small against large, is not a matter of private versus public enterprise, but concerns the nature of economic development, whether the agents of development are public authorities or private enterprises. Aid must take its share of the blame for the fact that Kenyan development has not emphasised the small and decentralised. It does not follow that, if there had been no aid, the pattern of development would have been better; it is more likely that there would have been just less development. A development process of a different character would have required a firm policy decision to go for the small and decentralised and the will to enforce it. It would also have required a knowledge—not to be presumed to exist—of how to develop in this different way. Far from inhibiting the possibilities of such development, aid for investigation of alternative patterns of development and alternative techniques could facilitate it. Donors could also in their selection of projects favour the small against the large. Although the change in British practice away from a willingness to pick up the ‘odds and ends’ left by other donors towards a liking for larger projects is, on these grounds, to be regretted, the policy of favouring projects which benefit agriculture and the poorer sections of the population does go in the right direction. There is, in practice, no visible connection between aid and the concentration of power. The danger of distortions being introduced by aid has particular application to technical assistance. The supply of personnel can be a powerful instrument for introducing the attitudes and techniques of the donor country into the different economic and social circumstances of the recipient, for which they are inappropriate. One writer asks rhetorically: Will it ever be possible to evaluate the cost of the wastage and of the ill-conceived decisions due to technical assistance based on the application of inappropriate techniques prompted… by the unimaginative transplantation of western experience into vastly different sociological surroundings…?10

Controversies Over Aid 197 Although, as we remarked in Chapter Six, these strictures do not by any means apply to all technical assistance personnel in Kenya, such adverse effects of technical assistance are not unknown. After all, the quality of technical assistance personnel varies from the highly competent to the barely employable, and it would be surprising if technical assistance had always worked well. But the difficulty with a general condemnation of technical assistance on the grounds that it introduces inappropriate techniques, is the implicit assumption that appropriate techniques are available. The argument reflects the inverted arrogance that the donors inevitably know less than the recipients. If the recipient countries did without technical assistance, the argument implies, they would do things the right way. In reality, many instances of ‘inappropriate’ technical assistance—as of inappropriate technology in general—are cases where what is ‘appropriate’ is known only in principle and not in practice. In such instances, doubtless there is a better way of doing things—methods more appropriate to the factor endowment, products more appropriate to consumer needs—but no one has yet made it concrete. The argument for ‘appropriateness’ is, therefore, often an argument for more investigation and research, backed by the necessary resources and political will, rather than an argument against aid and technical assistance, which necessarily operate within the constraints of existing knowledge. There may, indeed, be circumstances in which an investment in ‘inappropriate’ technology would pre-empt future opportunities when a more appropriate technology is available. It must rarely be wise, however, to forgo a beneficial investment in favour of a better one that may (or may not) become available in the future. Aid for education—an important sphere of British aid to Kenya—has come in for particular criticism for introducing inappropriate ideas. Again there are the rhetorical questions: …do these teachers…dispense the kind of education the host country really needs? Teaching little Africans…the history and geography of their former colonizers may be a long term investment on behalf of the donors, but how much does it contribute to local development? How can it be determined to what extent the teaching of metropolitan curricula has contributed to progress, or has rather helped to maintain or even broaden the cleavage between the intellectually expatriate ruling groups and the majorities…?11

This statement no more describes the reality of British educational aid to Kenya than the novels of Dickens describe life in modern Britain. There are, of course, immensely important issues about the appropriate structure of education and curricula in a country like Kenya. Education can be socially divisive. It can detract from a willingness to accept the life of the traditional society. It can produce half-educated, unemployed school-leavers. But a serious consideration of educational issues is not helped by the suggestion that what is wrong is that the schools teach the history and geography of the former colonial power (which, in fact, they do not in Kenya). Britain’s educational aid to Kenya is not, of course, above criticism (see Chapter Six) but it would be absurd to believe that the responsibility for the major faults in the Kenya educational system and in the curricula can be laid at the door of British aid. It would be quite mistaken to assume that if there had been no aid for education, even though the quantity of education provided might have been reduced, its quality and relevance would have been greater. The enormous political pressure for educational expansion would have

198 Aid and Inequality in Kenya: British Development Assistance to Kenya made it difficult to keep down the numbers at school, though fees might have been higher and the children of the poorest kept out in consequence. But the effect of the removal of educational aid would have primarily been a decline in quality within the same conceptual framework. Some expatriate teachers might have been funded by the Kenya government itself, but there would inevitably have been far fewer, and for a considerable period there would have been insufficient equally qualified Kenyan teachers to replace them. The total number of teachers would have been smaller, and on average they would have been less well qualified. The precise pattern must be conjectural, but what is certain in our judgement is that the absence of expatriate teachers would not have resulted in an educational system more relevant to Kenya’s needs. It would not have been different and better, but the same, only worse. However inappropriate the curricula of the Kenya schools may still be, it is not because there is a powerful lobby of expatriate teachers hindering reform. Much effort has been, and continues to be, devoted to curriculum development. There is no simple explanation of remaining inadequacies; pressures and vested interests have their place, but they are not created by aid. More Kenyans than Englishmen have intellectual capital invested in the existing system, and the existence of a neocolonial conspiracy is not necessary to explain conservatism. The pressure from the consumers for the maintenance of international comparability of the content and standards of education should not be underestimated. As a result it is difficult for changes to be introduced by expatriates. Too easily is ‘different’ interpreted as ‘inferior’, and if the structure of education is a colonial heritage, it is one from which it is equally difficult for Kenyans and expatriates to break away. And there is another important reason for the survival of ‘inappropriate’ curricula: there is no universally accepted answer to the question ‘What is an “appropriate” curriculum?’ It is not because of the baleful influence of the outmoded ideas of expatriate teachers that Kenya continues to produce unemployable—or at least unemployed—school-leavers. It is because this is a problem to which no one has yet found a solution, in Kenya or elsewhere. Certainly, many of the suggestions of foreign reformers have been too general to be useful. Technical assistance has improved the quality of teaching within a framework it was impotent to change. The fact that the framework was established in the colonial era, and is protected by attitudes that are influenced by colonial experience, does not mean that it either benefits the ex-colonialists now, or that they can do anything about it. Aid has been accused of fostering corruption, and though the accusation might seem to be more plausibly made against foreign investors and businesses, which are competing for opportunities, than against aid agencies, it is obvious that the availability of aid funds makes their misappropriation possible. It cannot be denied that there is corruption in Kenya: it is widely discussed in the press, and in 1975 a Select Committee of the Legislature was set up to investigate the matter. It would be surprising if some British aid funds had not been corruptly diverted from their intended purpose, though the administration of aid, with tying to projects and payment in arrears for expenditures incurred, makes the more blatant forms of misappropriation difficult, if not impossible. In principle, the involvement of government in the expenditure of large funds widens the scope for corruption, but in practice it does not seem likely—except to the extent that the government has more money—that the availability of aid has led to significantly greater corruption than would otherwise have existed.

Controversies Over Aid 199 There is, however, a wider and less literal sense in which it is argued that aid corrupts. Aid corrupts morally: it establishes dependency, destroys self-reliance, and enervates government and society. On this view: ‘there seems to be no short cut consistent with dignity which bypasses frugality today as a necessary condition for self-reliance and real progress tomorrow.’12 For this school of economic analysis, strength can come only through suffering. It can indeed happen that ‘pre-occupation with aid diverts attention from the basic causes of poverty and the possibilities of acting on them’,13 and that aid ‘enables those in power to evade and avoid fundamental reforms’14 (why mend the bone when the broken leg can be cut off?). But it is not inevitable that aid should on balance be detrimental by distracting a government from its development task, rather than beneficial by increasing the resources of the recipient country. Although, as was concluded in Chapter Six, the receipt of aid has probably reduced resource mobilisation by the Kenya government, it does not follow that this has been detrimental to development. Even if reduced resource mobilisation had resulted in increased consumption, that would not necessarily have been a reason for condemning aid. The developmental effect of aid is in part determined by the extent to which there is increased investment as a result of aid. Even though aid is given nominally for particular investment projects, this does not mean that total investment will necessarily be greater by the amount of the aid. The receipt of aid could be accompanied by a lower level of domestic saving brought about, for example, by a relaxation in the revenue-collecting efforts of the government. It is only in the extreme case that investment will rise by the full amount of the aid, just as the other extreme is approached when ‘aid is essentially a substitute for savings and…a large fraction of foreign capital is used to increase consumption rather than investment’.15 More commonly, it would be expected that the increase in available resources resulting from the provision of aid would be partly consumed and partly invested. The developmental effect of the aid would probably be the greater, up to a point, the greater the rise in investment. An increase in consumption, on the other hand, might have a greater effect on welfare, and there may be important developmental benefits from a rise in what is classed as consumption. Aid is not necessarily wasted, therefore, if it increases consumption rather than investment. It was pointed out in Chapter Six that although the lower level of resource mobilisation by government than might have occurred if aid had been smaller had released resources to the private sector, the resources could have gone into investment rather than into consumption. Some attempts at a statistical demonstration of the effect of aid on consumption and investment have been vitiated by a failure to disentangle the effect of aid from that of nonaid capital inflows. No judgement about aid can properly follow from such an analysis of changes in consumption and investment.16 The effect of all inflows, aid and commercial, might be judged from changes in the distribution between consumption and investment of total available resources. A rising proportion of consumption would suggest that the situation was nearer to the extreme in which the inflows were a substitute for domestic saving; an increasing proportion of investment would suggest that the situation was nearer to the other extreme, and that the effect of the inflows on investment predominated.

200 Aid and Inequality in Kenya: British Development Assistance to Kenya The statistics for Kenya indicate a rising proportion of total resources being devoted to investment, and a declining proportion to private consumption (see Table 19). Such figures are not conclusive, because it cannot be known what would have happened if the inflows had been smaller. Nevertheless, in the light of them we find it impossible not to conclude that the inflows benefited investment rather than consumption. Table 19: Use of resources available in the monetary economy

Investment Public consumption Private consumption

1964

1969

16 20 64

23 21 56

(percentages) 1970 1971 1972 25 20 55

27 21 52

1973

1974

29 20 51

29 16 55

26 22 53

Sources: Economic Survey 1974, Table 1.6 and Economic Survey 1975, Table 2.5.

Indeed, it was argued in Chapter Six that the procedures and requirements of aid donors have resulted in a diversion of Kenya government funds to an undesirable extent away from recurrent expenditures into investment. Another undesirable effect of the administration of aid was identified in the same chapter. It was concluded that aid had had unfortunate effects on economic administration by undermining central control of operational ministries. But these arguments deal with the details of aid administration and cannot be used as ammunition for a general attack on aid. They do not carry the implication that Kenya would have been better off either without aid, or with less aid. They are concerned with practicalities and are far removed from the concerns of those who condemn aid because it makes possible the avoidance of ‘fundamental reforms’. Kenya can perhaps be criticised for being in some respects one of those recipients which ‘pursue courses of action which patently reduce the level of income or retard its increase’.17 Aid recipients do this, it has been argued, in various ways: ‘For instance, they expel the most productive groups of the population from their countries, or restrict the inflow and the deployment of private capital’.18 Kenya has been criticised, however, for giving too much, not too little, encouragement to private capital. Nor can it be supposed that the attitude to the commercial activities of non-citizen Asians would have been less discouraging if aid had been less abundant; on the contrary, donor leverage has perhaps achieved a slower exodus of non-citizen Asians than would otherwise have been permitted. There is, indeed, one field in which the availability of aid does appear to have encouraged, or at least permitted, a laxity of effort on the part of the Kenya government. It was explained in Chapter Six that the Kenya government has seemed unwilling to reduce its use of technical assistance staff under OSAS at as fast a rate as the British government thinks desirable. But it would be fatuous to generalise from this instance and to conclude that Kenya’s development effort would have been greater and better directed in the absence of aid. Another argument against aid, the persuasiveness of which is diminished by its failure to distinguish aid from foreign private investment, is that ‘aid may have retarded development…by frustrating the emergence of an indigenous entrepreneurial class’.19 This effect is, of course, even more difficult to identify than the effect of aid on savings, and

Controversies Over Aid 201 it is not obvious that the development of entrepreneurial skills is inevitably frustrated by competition more than it is stimulated by example. Be that as it may, it is difficult to believe that foreign enterprise, let alone aid, has been a major barrier to the development of Kenyan entrepreneurship. In fact, the situation in Kenya is too complex to be analysed in simple terms of foreign enterprise versus indigenous entrepreneurs. If the emergence of an African entrepreneurial class has been frustrated it has been by Asian entrepreneurs, citizen and non-citizen, rather than by aid and foreign investment. The policy under which non-citizen Asians are being removed from business activity (for which aid and foreign investment can attract neither credit nor blame) is likely to protect African entrepreneurship much more than aid and foreign investment are likely to frustrate its emergence. In fact, foreign business, operating as it does under rules for Africanisation (even though to some extent these are satisfied by the employment of ‘front men’), stimulates rather than frustrates the emergence of an African entrepreneurial, or at any rate a managerial, class. We have characterised Kenya government policy as aiming to foster the entrepreneurial society, and we do not believe that the inflow of aid and foreign investment has hindered the achievement of that aim. That aid promotes such inequality in income and wealth between individuals and regions within the recipient country is a plank in the platform of the critics of aid. There are many possible links between aid and inequality. Aid funds may be corruptly diverted into the pockets of those who are already rich, and as a result powerful and influential, to make them still richer. Aid may provide services which benefit the rich, and not the poor—aid for improving a road along which the rich will drive, but not for roads which lead to the villages of the poor. Aid for education may help produce recruits for the rich from the ranks of the children of the rich, as when a university education guarantees employment at a ‘European’ level of pay, and when entry to a university requires an earlier education which few but the better-off can afford to buy their children. Aid may foster inequality by providing economic opportunities which only some are able to grasp. In fact, the concern of donors with the ‘efficiency’ of aid will encourage this effect by leading to a concentration of aid on projects and in places where people are able to grasp the opportunities created. It is widely believed that inequality in the distribution of income and wealth has increased in Kenya since independence. Although no definitive statistical demonstration of a growth in inequality is available, we do not wish to argue that it has not occurred. Indeed, the conclusion was reached in Chapter Six that aid has indirectly contributed to the growth of inequality, at any rate before donors’ recent emphasis on rural development. The question that remains to be considered is whether the growth of inequality is wholly to be deprecated, and aid criticised for contributing to it. Is the increased inequality resulting from aid to be seen, at best, as the price to be paid for the benefits of aid? Could not, alternatively, the growth of inequality be positively welcomed as a reflection of the fact that some people—in fact many people—are becoming better-off? It would be better still if all were becoming better-off, equally fast. But is not improvement, though unevenly spread, to be preferred to an equality of poverty? The emphasis on equality versus inequality is in danger of leading to a neglect of absolute changes in income and consumption. A growth of inequality can take very different forms. Inequality increases when a few get very rich while the absolute position of the majority remains unchanged or deteriorates. Inequality also increases when a large number, perhaps the great majority of the population, becomes

202 Aid and Inequality in Kenya: British Development Assistance to Kenya absolutely better-off, though some become better-off faster than others, and a very few become very rich. Although undoubtedly some Kenyans have become worse-off with the disruption of traditional ways by economic development, the growth of inequality in Kenya conforms more closely to the second than to the first of these patterns. Our conclusion, therefore, is that responsibility for the growth of inequality in Kenya cannot be laid at the door of aid in any crude or direct sense, though to the extent that aid has contributed to political stability and economic growth, it has contributed to the growth of inequality. Moreover, though there are features of the growth of inequality in Kenya which we—making our own value judgement—deplore, the conspicuous inequality—the stark contrast in Nairobi of Mbenzi20 and beggar—is a more superficial, less fundamental feature of the situation, of post-independence developments, than the widely varying but for very large numbers significant improvement in the standard of life. Nor do we think that because some have become substantially better-off (we do not refer to the very rich) while others remain at little above the standard of life provided in the subsistence economy, this is a condemnation of the process of development, though it should be taken as a guide to policy, including the policy of aid donors, to focus on improving the position of the poorest of the poor. Inevitably, some will benefit more than others—the benefits will be unevenly spread—if only because resources are insufficient to benefit all equally at the same time, and because of differences in the ability and willingness of different individuals and the people of different areas to take advantage of opportunities for improvement. Inequality is increased, for example, when water supplies are introduced into one area, but not yet to another. The aim of policy should be, it appears to us, not to endeavour to prevent the growth of inequalities of this kind, but to see that there is no tendency for the cumulative favouring of particular areas or individuals. It appears to us that Britain and other aid donors, and the Kenya government, are aware of the possibility that policies will foster the cumulative growth of inequality beyond that which we see as a ‘natural’ and inevitable characteristic of the process of development. It remains to be seen whether understanding is, or can be, adequately translated into achievement.

Notes 1. Determined believers in the Machiavellian view will see this reason for the toleration of foreign capital as confirmation of the compliance theory. In fact, a more obvious reason for favouring foreign private capital is that it brings in foreign private capital rather than because it brings in aid. 2. K.B.Griffin, ‘Foreign Capital, Domestic Savings and Economic Development: A Reply’, Bulletin (Oxford University Institute of Economics and Statistics), Vol. 33, No. 2, May 1971, p. 157. 3. K.B.Griffin and J.L.Enos, ‘Foreign Assistance: Objectives and Consequences’, Economic Development and Cultural Change, Vol. 18, No. 3, March 1970, pp. 315–16. 4. Griffin, 1971, op. cit., p. 160. But compare the view that aid is fully fungible, so that it always finances the marginal expenditure, including consumption (K.B.Griffin, ‘Foreign Capital, Domestic Savings and Economic Development’, Bulletin, Vol. 32, No. 2, May 1970, p. 103) which is compatible only on the particular assumptions of excess capital and an administrative constraint. See Chapter Six. 5. P.T.Bauer, Dissent on Development, London, Weidenfeld and Nicolson, 1971, p. 103. 6. Griffin and Enos, op. cit., p. 326.

Controversies Over Aid 203 7. Bauer, op. cit., p. 104. The further contention that because of its effect on the concentration of power aid ‘tends to favour and strengthen governments which lean towards the Soviet bloc’ (p. 128) cannot with respect to Kenya be taken seriously. 8. The increase, according to the statistics (see Chapter Two), has been more modest than is often believed, and it has been large only in investment expenditure. 9. T.Mende, From Aid to Re-Colonization, London, George G.Harrap, 1973, p. 156. 10. Ibid., p.51. 11. Ibid., p.52. 12. Ibid., p. 158. 13. Bauer, op. cit., p. 101. 14. Griffin and Enos, op. cit. 15. Ibid. 16. It really will not do, as a way out of the difficulty, to say that ‘public and private capital flows to the underdeveloped countries are often linked through the practice of aid-giving agencies of making their assistance contingent upon the willingness of the recipient countries to encourage foreign private investment’ so that ‘the effects of aid upon development cannot, in practice, be separated from the effects of private foreign capital on development’. (Griffin and Enos, op. cit., p. 324.) Even if their effects cannot be separated, it does not mean that the differences between them can be ignored, or that judgements about aid can be based on the effects of total inflows. 17. Bauer, op. cit., p. 101. 18. Ibid. 19. Griffin and Enos, op. cit., p. 326. 20. Mbenzi, the man who drives in a Mercedes Benz.

INDEX

Abaluhya 147, 155 Aden 75 African Socialism 20, 33, 37 Agriculture advisory and extension services 17, 21, 24, 26, 121, 169, 175–7; Agriculture Act (1955) 12, 116, 126; Agriculture, Ministry of, see Ministries, Kenyan; Agricultural Development Corporation (ADC) 26, 112–4, 224; Agricultural Finance Corporation (AFC) 22–3, 114, 176, 224; artificial insemination project 97; cash-cropping 13, 15, 24, 26, see also Mumias Sugar Company; commercial 14, 22, 25; foot-and-mouth control project 97; grain storage policy and projects 54, 200, 207–9, 249; irrigation 66; land tenure and reform 13–15, 22–6, 52, 54, 241; land transfer and resettlement 11, 13, 15–16, 20–25, 42, see also Land Transfer Programme; livestock-marketing and development projects 54, 66, 204; maize breeding scheme 204–5; marketing and credit 13, 15, 17, 21, 23, 24, 66, 169; National Agricultural Research Station, Kitale 204–5; pre-independence 12–16; research 13, 91; Areas (White Highlands) 12–15, 17, 22, 24–5, 105 Aid administration British, 79–80, 82–3, 88, 91–3, Chaps. 5 and 6 passim, 237–41, 243–4, 247–50 Kenyan, 68–70, Chap. 6 passim other, 91–3, 95–9 Aid policy

British 74–81, 84–91, 118–23, 135, 206–8, Chaps. 7 and 8 passim; German (West) 96–7; Swedish 67, 97–8, 232; U.S. 95–6; World Bank 98–9 Angaine, Jackson 121 Australia 51, 135 Belgium 95 Birth control schemes 206 Booker McConnell Ltd 145–51, 154, 156–7, 160–4 British Expatriate Supplementation Scheme (BESS) 33, 63 British High Commission, Nairobi 79–85, 87, 93, 119, 205 Bura Bura Housing Estate, Nairobi 65 Canada 50, 66, 90, 215 Capital aid British 49–50, 52–61, 76–80, 83, 85–90, Chap. 5 passim 204–5, 207–8, 223–7; other 50, 65–8, 95–9, 184–209, 215–7, 222–3, 227 Census of Industrial Production 30 Central Bank 121, 134 Central Land Board 107–8 Central Province 14, 26, 127, 204, 207 Chemelil sugar factory 54, 66, 148 Civil service, Kenyan and academics 167; and aid 70, 89–93, 198–9, 210–17; and corruption 210–11, 220; and expatriates 199, 210–12; and planning 21–24; Kenyanisation 33, 211; pay and conditions of service 36, 210; relations with politicians 40, 213–4 Coast Province 86 Colonial Development and Welfare Act (CD&W) 52, 74, 77–8, 204

Index Commonwealth Development Corporation (CDC) 15, 54, 63–5, 99–100, 108–10, 145, 162, 239 Commonwealth Finance Ministers’ Conference (Montreal) 74 Construction 16–17, 27 Cooperatives marketing 123; settlements 21, 108, 114–6, 118 Corruption 210–11, 220, 254 Country Policy Papers, see Overseas Development, Ministry of (ODM) Cyprus 75 Debt repayments 21, 41, 49, 54–7, 60, 67, 71, see also Land Transfer Programme Denmark 50–51, 69, 93, 181, 208–9 Dependence and aid 227–8, 232–4, 246–8, 254–5 Development Divisions, see Overseas Development, Ministry of Development Finance Company of Kenya (DFCK) 38, 64, 100 Development Plans 78, 213; (1966–70) 26; (1970–4) 21, 28, 42, 128, 165, 167, 199; (1974–8) 21, 26, 32, 44, 64–5, 122, 165, 180, 191, 195, 199, 203, 205 Digo 171, 173–4 Donor cooperation and competition 93–4, 99, 148, 165–7, 181, 184–9, 202–5, 214, 243–4, 246; Consultative Group for East Africa 98, 198, 244 Duruma 171–4 East African Community (formerly common market) 16–17, 27, 32, 39, 56, 61, 67, 147, 205 Economic policy, Kenyan, Chaps.2 and 5 passim, Chap. 6 190–209, Chap. 8 passim; attitude to aid 40–2, 77, 202–5; investment and development expenditure 37–40, 52, 131–2, 192, 195–205, 250–1, 255–6; public sector borrowing and taxation policy 190–6; recurrent expenditure 52, 192, 196–200, 203–4 Endarasha 127 Education 33, 40, 42–3, 81, 206–7, 213, 253–4

205

Egerton Agricultural College 78 Emergency Regulations 14 Exchange control 130, 194 Expropriation hypothesis 116–7, 124–34, 136–41, 246 Finland 51 Food and Agriculture Organisation (FAO) 158, 165, 175 ‘Food for peace’ 66 Ford Foundation 51 Foreign Investment Protection Act (1964) 27 France 51, 95 Fungibility of aid 11, 87–8, 104, 184–9 197–8, 239–41, 250 Germany (West) 38, 49, 50–51, 54, 64, 66–7, 77, 91, 96–7, 110, 148, 184, 202–4, 227 ‘Ghana 74, 200 Haraka (squatter settlements) scheme 21, 116–8, 223 Harambee settlement scheme 21, 112–4, 119, 121, 206–7, 219 Hart, Rt. Hon. Judith, 57, 232 Housing 35, 201 Income distribution 12, 24–6, 30–2, 36–7, 43–4, 110, 156–7, 165, 222–7, 249, 257–9 Independence settlement 53, 74–8 India 51, 158–60 Industrial and Commercial Development Corporation (ICDC) 38, 66 ICDC Investment Company 38 Industrial Court 38 Industrial Development Bank 38, 86 Industrialisation 16, 27–33, 37–8, 66 International Development Association (IDA) 49, 50, 67, 99, 171, 206 International Finance Corporation (IFC) 67 International Labour Organisation (ILO), see Reports International Monetary Fund (IMF) 98 Investment, private foreign 27–30, 32–3, 38, 75–6, 100–1, 125, 131–2, 162, 194–6, 205–6, 209, 221, 257 Investment, public see Economic policy, Kenyan Italy 51

206

Index

Japan 51, 200 Kaggia, B. 111–2 Kalenjin 218 Kamba 171, 173–4 Kariuki, J.M. 231n Kenya African Democratic Union (KADU) 75, 106–7, 219 Kenya African National Union (KANU) 75, 106, 107, 111–2, 135, 200, 217, 221 Kenya Commercial Bank 145, 162 Kenya Cooperative Creameries 40 Kenya Industrial Estates 38 Kenya Land and Agriculture Bank 109 Kenya Meat Commission 64 Kenya National Trading Corporation (KNTC) 34, 146 Kenya National Youth Service 66 Kenya People’s Union (KPU) 112, 117, 119, 122, 135 Kenya (formerly East African) Power and Lighting Company 37, 64–5, 251 Kenya Public Service Commission 80 Kenya Science Teachers Training College project 97 Kenya Tea Development Authority (KTDA) (formerly Special Corps Crops Development Authority) 64–5 Kenyanisation 33–8, 42 of civil service 33, 211; of informal sector’, 34; of land 12–17, 20–24; see also Agriculture, land tenure and Land Transfer Programme; of manufacturing 27, 33, 257; of trade 17, 34, 38 Kenyatta, Jomo 18, 27, 75, 76, 86, 107, 112, 122, 149 Kenyatta National Hospital 200, 249 Kenyatta University College 207 Kericho Conference 166 Khandsari 158–60 Kibaki, M. 230n. Kikuyu 14, 26, 75, 106, 107, 211, 218–9, 221 Kinangop 242 Kindamura power station 65 Kisumu 67, 100, 200 Kitale 204–5 Kwale, see Rural development Lancaster House Conferences 75, 105–7

Land Acquisition Act (1968) 116 Land Development and Settlement Board 113–4 Land reform, see Agriculture Land Transfer Programme (LTP) 11, 15, 52, 54, 57, 75–6, 104–44, 223–5, 241–3, 248–50; compassionate farm purchases 108–9, 242; low-density schemes 15, 21, 108–13; high-density schemes 15, 21, 106, 108–13, 249; ‘Million Acre Scheme’ 15, 21, 107–11, 118–23, 127; ‘New Scheme’ 106; Special Scheme 114; ‘Yeoman and Peasant Scheme’ 106, 108; ‘Z plots’ Scheme 249 Latin America 95 Leverage 44, 179–82, 201, 207, 234, 237, 248–51 Local costs 41, 55, 59–61, 67, 68–70, 86, 89, 149–51, 181, 234, 241 Luo 75, 174, 211, 218–9 MacArthur, J.D. 117 Machakos 126 Majimbo Constitution 107 Malindi 200 Manufactures Export Compensation Scheme 28 Manufacturing, see Industrialisation Masai 202 Mau Mau rebellion 14, 75, 125 Mbere 167, 169 Mboya, T.J. 112 McKenzie, B.R. 113, 147 McNamara, R. 98 Migori 168, 178 ‘Million Acre Scheme’, see Land Transfer Programme Ministries, British Department of Trade 57, 59, 79; Foreign Office 76–79; Overseas Development, see Overseas Development, Ministry of (ODM); Treasury 59, 77, 79, 86, 232 Ministries, Kenyan Agriculture 49, 89–90, 107, 112, 168, 172–3, 175–7, 201, 203, 208, 220, Project Planning Unit of 90; Cooperatives 49; Defence 49; Education 49, 52, 69, 90, 172;

Index Finance and Planning 41, 49, 52, 70, 82, 88, 91–3, 94, 107, 113, 167–9, 171 190–7, 213–5, External Aid Division of 91–2, 215–6, Farm Economic Surveys of 113, 118, 128; Foreign Affairs 49; Inland Revenue 52, 121; Labour 66; Lands and Settlement 106–8, 119–21, 242; Works 49, 89, 172–3 Mombasa 34, 66, 170, 200, 226, 250; TV project 54, 202, 248 Mumias Sugar Company (MSC) 54, 65, 101, 145–64, 225, 250 Munyu 117 Nairobi 17, 34, 64, 66, 69, 90, 94, 119, 165, 179, 193, 200, 204 University of, Institute of Development Studies, see Reports Nairobi-Mombasa road project 54, 226, 250 Naivasha-Suswa pipeline project 54, 202 Nakuru, 66, 67 Nandi salient 108–9 National Rural Development Coordinating Committee, see Rural development Nationalisation, 20, 37, 76, 251 Ndegwa Commission, see Reports Netherlands 38, 49, 50–1, 181 Non-African communities 12, 16, 17, 28, 34, 106; Asians 12, 28, 34, 46n, 85, 100, 160, 256–7; European farmers 12, 20–22, 53, 74–8, see also Land Transfer Programme North-East Region road building project 54 Norway 50–1, 69, 93, 167 Nyanza Province 26, 146, 149, 204 Odinga, Oginga 75, 112 Oil 42–3, 61, 85, 191, 213 Ol Kalou salient 108 Overseas Development, Ministry of (ODM), Chaps.4, 5, 6 and 7 passim; Country Policy Papers (CPP) 64, 79–86, 238; Development Divisions 80–85, 87, 92–3, 179, 205, 234, 236–7; Guide to Project Appraisal 91;

207

Projects Committee 80, 87, 92; relations with other UK ministries 57, 59, 79–80, 86–7, 232; technical assistance administration 80–4, 179–80, 214, 235–7 Overseas Service Aid Scheme (OSAS) 33, 63, 69–70, 76, 206–7, 209, 210–11, 214–5, 223, 235 Peace Corps 66 Political system 217–22 Polytechnic and university colleges project 54 Population 12, 43, 111, 165 President 40, 190, 217–9, 221; Office of, 169, 201 Programme aid 61, 66, 85, 90, 96–7, 98, 233, 241 Recurrent financing 90, 97, 166, 238–9 Renison, Sir Patrick 75 Reports Department of Lands and Settlement Annual Reports, 111, 122–3; East Africa Royal Commission (1955) 14, 25; ILO 24–5, 34–7, 44, 85, 113, 135, 165; Maize Board, 208; Ndegwa Commission 36, 220; ODM 112–3; Stamp 108, 111–3, 118; University of Nairobi, Institute of Development Studies 166–70, 175; van Arkadie 107, 110, 120–3, 128; World Bank 85, 165, 205 Rockefeller Foundation 51 Rural development 85–7, 89, 98, 154–8, 165–83; access roads scheme 54, 85; District Development Planning 180–2; local involvement 169, 173–9, 181; National Rural Development Coordinating Committee 166–7, 172–3; Rural Development Fund 98, 181, 234, see also SRDP; Special Rural Development Project (SRDP) 94, 104, 165–83, 250 Sectoral allocation of aid 85–7, 95, 98, 201–4, 206–9, 223, 233–4, 241 Scheduled Areas, see Agriculture

208

Index

Seroney, J. 230n, 231n Sessional Papers No. 10 (1965) On African Socialism 20, 23, 37, 41, 76, 105, 112, 133; No. 4 (1975) 43–4; No. 10 (1973), On Employment, 222 Shikuku, M. 231n Shirika settlement scheme, 21, 114–6, 119, 122, 242–3 Shunting, shuffling, switching of aid, see Fungibility Special Commonwealth African Assistance Plan (SCAAP) 63, 211, 214–5, 235 Special Rural Development Project, see Rural development Squatters, Commissioner of, 116 Stamp, Maxwell, see Reports Sterling 55, 60–1, 111 Suez Canal 75 Sweden 50–1, 67, 77, 93, 97–8, 171, 181, 202, 227 Switzerland, 51 Swynnerton Plan (1954) 14, 24–5, 147, 209–10 Tana River Development Authority, 65 Tanzania 16, 29, 32, 97 Tariff policy 16, 27–8, 32–3, 205 Taxation 27, 191, 195 Technical assistance British 33, 49, 51–2, 61–3, 69–70, 76, 80–4, 89–90, 99, 120, 206–7, 209–15, 235–7; other 33, 49, 51–2, 61–3, 66, 69–70, 89–91, 96, 168–70, 177–82, 199 Treaty for East African Cooperation 29 Tying of aid

procurement 54–61, 66, 77, 86, 97, 99, 185–6, 238; project 58–9, 74, 78, 82–3, 87–8, 186–9 Uganda 16, 29, 32, 85, 147, 149, 151, 205 UNCTAD 58, 60 Underspending of aid British 52, 78, 83, 88–93, 232–3; other 52, 89–93, 96–7 Unemployment 42–3, 70, 111, 165 165 UNDP 50–1, 69, 94 UNIDO 158 United States 49, 50–1, 65–6, 82, 94, 95–6, 179, 202, 206, 216 U.S.S.R. 51 van Arkadie, see Reports Vihiga 168–9, 178 Wanga 145–64 Water supply and sewerage scheme, 66–7 Western Province 204 White Papers British Private Investment in Developing Countries 100–1; The Changing Emphasis in British Aid Policies: More Help for the Poorest, 232 WHO 51 World Bank 15, 22, 38, 41, 49, 50–1, 66–8, 85, 90, 93, 97–9, 108–110, 200 Yugoslavia 51

DOES AID WORK IN INDIA? A country study of the impact of official development assistance Michael Lipton and John Toye

ROUTLEDGE LIBRARY EDITIONS: DEVELOPMENT

ROUTLEDGE LIBRARY EDITIONS: DEVELOPMENT

DOES AID WORK IN INDIA?

DOES AID WORK IN INDIA? A country study of the impact of official development assistance

MICHAEL LIPTON and JOHN TOYE

Volume 10

LONDON AND NEW YORK

First published in 1990 This edition first published in 2011 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 This edition published in the Taylor & Francis e-Library, 2011. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to.eBookstore.tandf.co.uk. Routledge is an imprint of the Taylor & Francis Group, an informa business © 1990 Michael Lipton and John Toye All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 0-203-84015-1 Master e-book ISBN

ISBN 13:978-0-415-58414-2 (Set) eISBN 13:978-0-203-84035-1 (Set) ISBN 13:978-0-415-59269-7 (Volume 10) eISBN 13:978-0-203-84015-3 (Volume 10) DOI: 10.4324/9780203840153 Publisher’s Note The publisher has gone to great lengths to ensure the quality of this reprint but points out that some imperfections in the original copies may be apparent. Disclaimer The publisher has made every effort to trace copyright holders and welcomes correspondence from those they have been unable to contact.

Does Aid Work in India?

How has India benefited from 30 years of official aid programmes? Michael Lipton and John Toye examine the impact of foreign aid to developing countries, focusing on India as an example of a very important recipient country. Drawing on their wide knowledge of different facets of Indian development, including planning, agricultural development, public expenditure, and India’s relations with the UK, they evaluate in an authoritative way the successes (and some undoubted failures) that foreign aid has contributed to the history of Indian development. Lipton and Toye examine the impact of aid at both the macroeconomic and the microeconomic levels, and give special attention to aid as a means of poverty alleviation, the evaluation of aid projects, and the broader effects of aid on institutions and private sector markets. Their book will be essential reading for officials concerned with aid policy both in governments and international financial institutions, for economists and political scientists who analyse the aid process, and for students of economic development. The Authors Michael Lipton is a Professorial Fellow of the Institute of Development Studies, University of Sussex. John Toye is the Director of the Institute of Development Studies at the University of Sussex.

Does Aid Work in India? A country study of the impact of official development assistance

Michael Lipton and John Toye

London and New York

First published 1990 by Routledge 11 New Fetter Lane, London EC4P 4EE This edition published in the Taylor & Francis e-Library, 2011. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to.eBookstore.tandf.co.uk. Simultaneously published in the USA and Canada by Routledge a division of Routledge, Chapman and Hall, Inc. 29 West 35th Street, New York, NY 10001 New in paperback 1991 © 1990 Michael Lipton and John Toye All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data Lipton, Michael, 1937– Does aid work in India?: a country study of the impact of official development assistance 1. India (Republic). Economic development. Foreign assistance I. Title II. Toye, John, 1942– 338.91′0954 ISBN 0-415-07160-7 Library of Congress Cataloging in Publication Data Lipton, Michael. Does aid work in India?: a country study of the impact of official development assistance/Michael Lipton and John Toye. p. cm. Includes index. ISBN 0-415-07160-7 1. Economic assistance—India—Evaluation. I. Toye, J.F.J. II. Does aid work? III. Title. HC435.2L58 1989 338.91′0954–dc19 89-3532 CIP ISBN 0-203-84015-1 Master e-book ISBN

Contents

List of Tables Preface Glossary

1

2

3

x xi xiii

Introduction

1

India’s Aid Resources in Macroeconomic Context Introduction

3 3

Measurement of aid flows

4

Size of aid flows to India

5

Impact of aid on economic growth

19

India’s future capital inflows: aid or commercial lending?

26

Conclusions

32

References

33

Aid and Poverty in India Dilemmas in using policy against poverty

35 35

Scale and structure of aid: impact on poverty

41

Aid to agriculture: poverty-oriented?

44

Aid other than to food and agriculture: poverty impact

58

Learning from India

60

References

62

Policy Dialogue Surprising facts and aid processes

65

From bilateral leverage to multilateral sector dialogue

65

Before leverage

69

Towards the crisis of bilateral macro-leverage

71

65

viii Contents

4

5

6

7

Aid 1965–85: towards multilateral sector dialogue

75

The problem of ‘ideology’ in policy dialogue

79

Risk of rigidity

82

Changing fashions

84

Choosing sectors and topics for dialogue

86

Can the dialoguer deliver?

89

References

90

The Systemic Effects of Aid and Donor Procedures Introduction

93 93

Many donors: help or hindrance?

96

Systemic aspects of multilateral aid

99

Systemic aspects of bilateral aid

103

Local cost finance

109

Donor procedures

112

References

114

Project Aid to India Role and evaluation of project aid

116 116

The World Bank Group’s project sample

118

Some projects from other donors

151

Appendix: Does India’ s project aid matter?

164

References

166

Resource Management, Institution Building, and Technical Assistance Aid’s contribution to resource management and institution building: the organized sector Resource management and institution building: the rural sector

170

Aid to remove administrative and manpower deficiencies

180

References

184

Aid and Market Forces The ‘control syndrome’ and the market critique

186 186

The effects on private foreign investment

189

The control syndrome and domestic business

190

170 174

Contents ix Donors and the control syndrome

192

How much aid goes to the private sector?

193

Aid and the private sector: recent policy changes

196

References

200

Appendix: Conclusions of the India Aid Effectiveness Study Notes Index

202 205 217

Tables

1.1 1.2 13 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12 1.13 1.14 4.1 4.2 4.3 4.4 5.1 5.2 5.3 5.4 5.5 7.1

External assistance to India, 1966–85 Untied credits as a percentage of gross aid to India, 1965–82 Average external assistance to India, 1958–72 India’s loan disbursements and debt service 1975–6 to 1985–6 India’s grant disbursements, 1975–6 to 1985–6 India’s net transfer of concessional finance, 1975–6 to 1985–6 Official development assistance disbursements to India, 1971–85 India’s balance of payments, 1948–70 Ratios of aid utilizations to imports in India, 1966–84 India’s balance of payments, 1975–6 to 1984–5 Sectoral financial surplus/deficits in India, 1951–75 Ratios of gross aid utilizations to saving and investment in India, 1975–86 Ratios of gross aid utilizations to Government expenditure categories in India, 1966–85 Key growth rates of the Indian economy, 1951–85 India: aid utilization by source Gross loan and grant disbursements by Aid-India Consortium Gross disbursements of Consortium aid in 1979–80 Tying status of DAC bilateral aid, 1982–3 World Bank projects in India Bank role in funding some projects Time and cost escalations in World Bank-aided projects in India Some UK-aided projects in India UK gross bilateral aid to India Public/private-sector shares in foreign loans and grants to India up to 1966

7 9 9 10 11 12 13 14 15 16 17 18 20 22 97 100 104 105 119 127 129 156 160 194

Preface

Since 1974, the World Bank and the International Monetary Fund have had a joint committee, called informally the Development Committee, which advises both institutions on the broad issues involved with resource transfers to developing countries. In 1982, the Bank/Fund Development Committee appointed a Task Force on Concessional Flows, whose membership consisted of nine representatives of developed countries and nine from developing countries. Their task was to report on ‘problems affecting the volume and quality and effective use of concessional flows of financial aid both in the shorter and in the longer term’. As a result of the activities of the Development Committee’s Task Force, several documents have already been published. On behalf of the Development Committee, in December 1985 the World Bank issued the Report of the Task Force as Development Committee Pamphlet No.7 whose twenty-six pages provide the essential findings in a form suitable for widespread dissemination. The Bank also prepared a companion volume of 135 pages (Development Committee Pamphlet No.8) composed of supporting materials which help to explain and substantiate the essential findings of the Report. So far so good. In carrying out its remit, the Task Force also engaged a group of twentythree consultants, headed by Professor Robert Cassen then of IDS, Sussex, but now Director of Queen Elizabeth House, Oxford. The work of the Task Force’s consultants under Robert Cassen, was published in 1986 by the Clarendon Press as Does Aid Work? Report to an Intergovernmental Task Force by Robert Cassen and Associates. Does Aid Work? was based on, among other things, a set of seven country case studies: Bangladesh, Colombia, India, Kenya, Malawi, Mali and South Korea. The present authors were among the consultants to the Task Force and co-authored the country case study of India. This was a two- volume mimeograph document which is cited in Does Aid Work? as ‘Lipton, M. and J.Toye with R.H.Cassen, 1984, Aid Effectiveness: India’. This document is the starting point of the present book. Since the general conclusions which emerge from the Cassen volume in part rest on the findings of the country case studies, and since at present none of these is available except in very limited quantities and impermanent form, it seemed to us desirable at least to publish the India study. Since it was completed in 1984, it needed up-dating and revision. This task took us longer than we had initially expected. One reason was that the aid debate has dramatically revived since 1984, with important contributions such as Roger Riddell’s Foreign Aid Reconsidered (London, James Currey) and Paul Mosley’s Overseas Aid: Its Defence and Reform (Brighton, Wheatsheaf) making their appearance, as well as much specific material on aid to India. Another reason is the usual one of a variety of competing projects and activities to which both authors were committed. The effect is cumulative: the longer the delay, the more radical are the revisions which are required by unfolding events. Even now,

xii Preface we do not claim that we have incorporated everything of relevance to our subject which has appeared since 1984, although we have looked at a great deal of new material. The division of labour has been that Michael Lipton originally wrote Chapters 2, 3, and 5 and John Toye wrote Chapters 1, 4, 6, and 7. However, the end product is much more of a joint effort than such a split would imply. Each author read and commented extensively on the other’s chapters through several drafts, and we have both tried to follow a principle of excluding statements with which our co-author strongly disagreed. No doubt we have not achieved a uniform style of presentation, nor is each author equally knowledgeable on all topics covered by the book. But it does represent a consensus between us on what might be sensibly concluded about the impact of aid on India. In writing this book, many debts have been incurred, which it is a pleasure now to acknowledge publicly. Our first debt is to Robert Cassen for interesting us in this project and for co-ordinating the work on the original consultants’ report Aid-Effectiveness: India, Robert Cassen also contributed an ‘Overview’ to the original report which is not reproduced in this book although it served a valuable integrating purpose at the stage before the authors had had time to develop the joint approach mentioned above. During the preparation of the original report, we benefited from many interviews and conversations with officials of donor governments, aid agencies and international financial institutions, as well as with past and present representatives of the Government of India. We have followed a rule of preserving the anonymity of our informants unless they have already put their views in the public domain. We are extremely grateful for all the assistance generously given to us by these informants. In the process of revision, we have drawn on the work of several research assistants, Olive Hamilton, Jonathon Perraton and Herman Prieto, all of whom discharged their tasks with accuracy, thoroughness, and speed, much to their credit. Many hands have helped in word-processing. Michael Lipton wished to salute the prodigious efforts of Ann Watson and John Toye is grateful to Hazel Lewis, Jo Stannard, Marion Huxley, Rosalyn Skilton, and Francine Spencer for their invaluable contributions to the production of the final typescript. Michael Lipton Brighton John Toye May 1988

Glossary

AICC

All India Congress Committee

ARDC

Agricultural Refinance and Development Corporation

ATP

Aid and Trade Provision (UK)

BDO

block development officers

BMWZ

Bundesministerium für Wirtschaftliche Zusammenarbeit (German aid ministry)

CD

community development

CDC

Commonwealth Development Corporation (UK)

CENTO

Central Treaty Organization

CGIAR

Consultative Group for International Agricultural Research

DAC

Development Assistance Committee (OECD)

DEA

Department of Economic Affairs (GoI)

DPAPs

drought prone area programmes

EFF

extended financing facility (IMF)

EGS

employment guarantee scheme

EROR

economic rate of return

EV

evaluation report (UK)

FAO

Food and Agriculture Organization (UN)

FY

financial year

GoI

Government of India

HY Vs

high-yielding varieties (of cereals)

IADP

Intensive Agricultural District Programme

IAS

Indian Administrative Service

IBRD

International Bank for Reconstruction and Development (‘Hard’ loan section of World Bank)

ICB

international competitive bidding

ICICI

Industrial Credit and Investment Corporation of India

ICOR

incremental capital-output ratio

IDA

International Development Agency (‘Soft’ loan section of World Bank)

IDS

Institute of Development Studies (Brighton)

IEL

Indian Explosives Ltd

IFAD

International Fund for Agricultural Development (UNO)

IFFCO

Indian Farmers’ Fertilizer Company

xiv

Glossary

IFPRI

International Food Policy Research Institute (Washington DC)

IFS

Indian Forestry Service

IIT

Indian Institutes of Technology

IMDP

Indian Manpower Programme

IRDP

Integrated Rural Development Programme

ISICO

Iron and Steel Company of India

ITDG

Intermediate Technology Development Group

LDB

land development bank

LDCs

less developed countries

LICs

low-income countries

MLD

medium- and long-term debt

MRTP

Monopolies and Restrictive Trade Practices Act (India)

NAB ARD

National Bank for Agriculture and Rural Development (formerly ARDC) (India)

NDDB

National Dairy Development Board (India)

NGO

non-governmental organizations

ODA/ODM

Overseas Development Administration (formerly Ministry), UK

OECD

Organization for Economic Cooperation and Development

OED

Operations Evaluation Department (World Bank)

OF

Operation Flood (EEC Indian Dairy aid project)

OOF

other official flows

PCRs

project completion reports (of World Bank)

PFI

private foreign investment

PLDB

primary land development bank

PPAs

project performance audits (of World Bank)

RBI

Reserve Bank of India

RTA

Retrospective Terms Adjustment (UK)

SAL

structural adjustment loan (World Bank)

SF

school feeding

SIDA

Swedish International Development Agency

SNAB

Swedish National Audit Bureau

SROR

social rate of return

SSA

sub-Saharan Africa

STEPs

short-term expert pools

T and V

training and visit (system of agricultural extension)

Glossary TC

technical co-operation

TISCO

Tata Iron and Steel Company (India)

US AID

US Agency for International Development

VLW

village level workers

WFP

World Food Programme

xv

Introduction

In the 1980s, official development assistance, or ‘aid’ for short, has been in the doldrums. After a generation of aid-giving, politicians and the public in donor countries had become somewhat weary with it, and cynical about what aid had achieved, or could achieve. Enthusiasm for aid-giving had flagged, and the voices of those who opposed the whole enterprise—from left and right of the political spectrum, but mainly from the right—became increasingly influential. Giving foreign exchange to the governments of poor countries where economic infrastructure is weak and public administration is fragile is self-evidently a risky undertaking. It was therefore not a task of great difficulty to find colourful evidence of waste, corruption, and disastrous outcomes on at least some aid projects, nor to discover contradictions or anomalies in the aid policies of donors which could serve as grist to the anti-aid mill. Aid-weariness also came at a convenient time for the donors. The recession of the early 1980s was accompanied by powerful pressures to cut public spending and the aid budget was often a soft target for cuts. The first half of the 1980s witnessed a steady decline in the real value of the UK’s aid programme. But a recession is an extraordinarily bad time, from the point of view of poor countries in receipt of aid, for aid flows to fall. This simply adds to the misery caused by dramatic falls in the prices of primary commodity exports, and the return to positive real interest rates on debt. The question of the effectiveness of aid arose out of this contradiction. It was hoped either that developing countries could be shown a way to get more benefits from less aid, or that by shifting aid money towards more effective uses, the political support for aid in developed countries could be revived and aid-cynicism dispelled. India was in many ways an obvious choice of country to turn to, in order to examine in empirical depth the question of aid-effectiveness. India had a long history of receiving aid, which had progressed through a number of quite distinct phases. India had also been in receipt of large absolute amounts of aid (although, because of her huge population, these translated to small amounts per head). India had been singled out in the debate on development strategy as the archetype of the country which had adopted the wrong development strategy—rapid industrialization on the basis of import substitution: this raised the question of the role of aid in the formation of economic policy and development strategy. India also appeared likely to suffer from a redistribution of aid towards sub-Saharan Africa: this raised the questions of whether India would not then be ‘punished’ by the aid donors for a relatively good aid performance and whether elements of India’s experiences with aid management could not be transferred to the African situation, so that an India-toAfrica redistribution of aid did not result in a lowering of its overall effectiveness. This then is the general background to the writing of this book, and some indication of the kind of questions which it addressed. The general conclusions which the original study DOI: 10.4324/9780203840153-1

2

Does Aid Work in India?

arrived at are set out in the Appendix, and they remain broadly the conclusions which the book in its revised form points to. Two more preliminary remarks are necessary. One is to emphasize the limits of the scope of this enquiry, which confines itself to ‘official development assistance’. It does not encompass non-governmental flows of aid to India, through Oxfam, Save the Children Fund and the many, many other charities and agencies which operate there. Within official aid, we also say very little about the smallish amounts given for immediate disaster relief. The other remark is to point to a limitation of form. We have adhered closely to the format of Cassen’s Does Aid Work? This has led us to exclude some questions which could be discussed under our general title. But it should be helpful to those who want to see clearly how much of the findings of the Cassen study rely on our interpretation of the Indian experience of aid.

Chapter 1 India’s Aid Resources in Macroeconomic Context

Introduction India is a very big country, with 765 million people in 1985 inhabiting an area of over three million square kilometres—233 persons per square kilometre, compared with 52 for all developing countries and only 19 for sub-Saharan Africa (World Bank, 1987, 202). The sheer size of India has conditioned the relationship between aid resources and the Indian economy as a whole. In absolute terms, India has always been one of the three largest recipients of net economic aid.1 From the major western donors in OECD—who in 1985 disbursed US$29.4bn of net economic aid to developing countries (bilaterally and via multilateral agencies), of which US$25.6bn could be clearly allocated to particular countries—India received US$1.5bn in 1985, as against US$1.8bn for Egypt (population fifty million) and US$2.0 billion for Israel (population four million). Net aid from other donors in 1985 provided a further US$7.6bn for developing countries, but almost nothing for India, since a net inflow of about US$100m from Comecon was offset by a US$91m net outflow to OPEC (i.e. capital repayments of past aid loans outweighed new aid receipts) (World Bank, 1987, 242–5; OECD, 1986, 77–82, 258, 284). These same figures show that aid donors have never been willing to provide the massive absolute sums of money required to render India’s aid, per person or as a proportion of GNP, comparable with that of other low income countries. In 1985 aid constituted 0.7 per cent of India’s GNP as compared with 7.8 per cent for low-income countries other than India and China; per capita aid was US$1.9 for India compared with US$16.2 for other low-income countries excluding China (World Bank, 1987, 244). This ‘large-country effect’ results partly from the geo-political motivation behind aid-giving. Most donors wish to appear generous in as many different countries as possible; and can win friends in twenty medium-sized poor countries, each with 35–40 million people, by economizing on aid to one large country like India. Apart from this ‘large-country effect’, the amount of aid that India has received has depended on the state of its bilateral relations with major donors. Among western donors, which have provided the bulk of aid to India, India’s relations with the United States—at least from 1960 to 1972—were a major determinant of the size of her aid flows (Chapter 3). India’s aid flows remain small compared with those of neighbouring (and similarly poor) Pakistan, reflecting both India’s greater size and its more distanced, and at times turbulent political relations with the United States. Thus, in a macroeconomic context, aid resources have always played a much smaller part in India than in Asian countries like Korea and Taiwan in the 1950s and 1960s, or in most African countries in the 1970s and early 1980s.2 Yet the macroeconomic effects of DOI: 10.4324/9780203840153-2

4

Does Aid Work in India?

aid on India have not been negligible. To explain these effects some preliminary problems have to be tackled. Solutions must be found for various problems of measurement of aid resource flows; The size of these flows over time to India must then be presented, so that trends, fluctuations and turning points can be identified; Aid resources must then be seen relative to key macroeconomic variables such as imports, government expenditure categories, and domestic investment; such ratios provide indicators of the dynamic role of aid in financing development (its static contribution to potential welfare, in a given year, being crudely approximated by its ratio to GNP); The econometric evidence on the link between aid resources and economic performance should be examined to see whether it suggests any conclusions on the nature of this link; Any additional aspects of aid’s contribution to the macroeconomic performance of India must be specified and evaluated. The remainder of this chapter addresses each of these tasks and draws a set of conclusions on the impact of aid on India at the aggregate level. In the light of these conclusions, a final section also looks at one important question for future financial policy: should India continue to expand its borrowings from commercial banks and progressively reduce its absorption of concessional flows of aid? Measurement of aid flows International financial flows may be classified as follows: Aid: Multilateral aid Bilateral aid Other concessional finance: Multilateral loans on ‘hard’ terms (IBRD and IMF loans) Official export credits Officially guaranteed suppliers’ credits Non-concessional finance Loans from the Eurocurrency market Other commercial bank lending

The definition of aid used here follows that of the OECD’s Development Assistance Committee’s definition of Official Development Assistance: i.e. official flows of finance for development purposes containing a grant element of at least 25 per cent. Among ‘other concessional finance’, multilateral loans contain a grant element of less than 25 per cent; official export credits and officially guaranteed suppliers’ credits contain varying grant elements, which are offset by their being tied to specific sources at prices above the cheapest available on the world market—typically by about a fifth (Riddell, 1987,209). The giving of Western foreign aid to India follows a regular set of procedures. Donors meet together with Government of India officials at regular intervals, in a ‘Consortium’ organized by the World Bank, in order to discuss the current state of the Indian economy and the government’s plans and projects for future development There are four distinct stages

India’s Aid Resources in Macroeconomic Context 5 in the aid process. First, at these meetings, the donors make ‘pledges’, that is, promises of amounts of aid to be given within a specified time period. Second, these pledges must be ratified in the donor countries. From this follows, thirdly, specific action by donors to make the resources available—the stage of aid authorization or ‘commitment’. Not all available aid resources are used up straight away, however. Only when they are drawn down, to pay for components in an actual project or programme, do we reach the fourth stage: ‘disbursement’ or utilization. Aid flows through an invisible bureaucratic pipeline, part foreign and part national, from pledge to final utilization. This has a political pay-off for donors, incidentally. They can be thanked for the same unit of aid several times over as it makes its way along the invisible pipeline. However, the longer the gaps between stages the more the value of donors’ commitments is eroded by inflation. Flows of aid can be measured at a number of points in this pipeline, Both the OECD and the Indian Government publish figures for both aid authorizations and aid utilization. This book concentrates on utilizations; authorizations represent specific resources available from a donor but such commitments are not always embodied in formal agreements. Note that the ratio of authorization to utilization, in a sector or area, can measure the recipient government’s progress in implementing aid-financed activities there; and declining or rising authorizations can show declining or rising concern, by one or all donors, for a particular country or sector. However, a historical study or a study from a macroeconomic perspective should focus on the actual transfer of resources: the aid utilization figures. Since much aid is in the form of loans, which entail repayment of principal and interest, it is helpful to show gross utilized aid alongside figures of debt service payments on concessional loans. Net aid is: gross utilized aid minus capital repayments of past aid loans; net transfers are: net aid minus interest payments on past aid loans. Such net calculations are feasible only for total aid to India, with which we are concerned here. It is not possible to show aid figures net of debt service payments at the sectoral level, only gross aid flows to a sector—because debt service repayments are a central government responsibility. Figures of gross and net utilized aid, and net transfers, to India are given in Table 1.1. Size of aid flows to India Absolute flows The figures in Table 1.1 are given in Indian rupees. This does not indicate the changes due to aid in India’s purchasing power over imports. Yet it is this purchasing power and its fluctuations which mainly interest us, if we want to know how much help the resources have been to the Indian economy. This is because, at least until the mid-1970s, the great bulk of aid resources were then spent on the purchase of imports. From the mid-1970s onwards, restrictions on using aid to meet locally incurred, rupee-denominated costs were somewhat relaxed, thus complicating the question of estimating the purchasing power of aid. To estimate the changing impact of aid inflows on India’s access, each year, to imported goods and services, we must adjust the rupee figures of aid to allow for:

6

Does Aid Work in India? (a) Changes in the exchange rate of the rupee; (b) Differences in the inflation rate between India and the countries from which she imports, unless offset by exchange rate changes; (c) Changes in the extent of aid-tying—i.e. restrictions on possible suppliers of the desired imports, which will alter the difference between the world price of the goods to be imported and the actual price that India has to pay.

As for (a), the rate of exchange between the rupee and foreign currencies has altered during the thirty-year period with which we are concerned. Until May 1966, a fixed rate of US$1=R 4.76 was maintained. Then the rupee underwent a major devaluation to US$1= R 7.50, which lasted until December 1971. At that point, the rupee was pegged to sterling and the nominal exchange rate remained constant until 1975, when the rupee was pegged to a basket of different foreign currencies important in India’s pattern of trade, combined together by an unannounced weighting system. The margin of variation permitted around this peg was initially 2.5 per cent but this was widened to 5 per cent in January 1979 (Joshi and Little, 1986,35). As for (b), India’s policy preference for maintaining a relatively fixed or only gradually changing nominal exchange rate (apart from the big devaluation of 1966) has meant that changes in its effective exchange rate, i.e. in the quantum of imports that can be purchased with one rupee, have depended largely on the differences between India’s domestic inflation rate and the world rate of inflation. In 1973–83, India’s inflation rate was lower than those of most export competitors and import suppliers. It was 7.7 per cent per year, compared with 13.8 per cent for other low-income countries (excluding China), 8.0 per cent for industrial market economies, and 29.3 per cent for middle-income countries (World Bank, 1985, 174). As a result, India’s effective exchange rate fell by 23 per cent between 1974 and 1979, both stimulating exports and making imports much more costly in terms of rupees. Between 1979–83, moreover, the industrial market economies’ inflation rate fell sharply, eroding India’s export advantage and reducing the rupee cost of imports (Joshi and Little, 1986,36). As for (c), the third element which affects the purchasing power of aid resources is the stringency of donor restrictions on import procurement. Procurement-tying of aid raises the costs of associated imports because the aid recipient cannot shop around for the best buy but must purchase from the donor only. The effect on import prices—i.e. the cut in the value of aid—depends on the size, competitiveness and exchange-rate policy of the donor, but has been consistently estimated as averaging around 20 per cent (Bhagwati, 1970; Chaudhuri, 1978, 103). But the effects are hard to pin down. Table 1.2, which shows untied aid as a percentage of gross aid loans (but gives no indication of the extent of tying among grants), suggests that tying fell to historically low levels in the early and middle 1970s, but rose towards the end of the 1970s. Taken together, these three influences on the purchasing power of aid make it very difficult to estimate the ‘real value’ to India of the nominal flows, or how much it changed over nearly thirty years of aid. We can, however, take into account exchange-rate fluctuation and express aid flows in US dollars (Table 1.3). These data for 1958–9 to 1971–2 (Chaudhuri, 1978)3 indicate that, before the oil shock and floating exchange rates, net aid utilizations in current US dollars peaked during the Third Five Year Plan period (i.e. 1960–1 to 1965–6) and went into a marked decline thereafter.

2388.7

648.9

759.2

649.3

628.6

617.2

642.6

615.0

987.9

1,185.2

1,429.1

1,216.4

877.2

792.9

927.8

1,089.0

1,142,4

1,623.5

1,478.5

To end Third Plan

1966–7

1967–8

1968–9

1969–70

1970–1

1971–2

1972–3

1973–4

1974–5

1975–6

1976–7

1977–8

1978–9

1979–80

1980–1

1981–2

1982–3

1983–4

Loans

303.4

339.4

350.6

396.4

304.4

273.3

260.6

245.8

283.3

93.9

20.7

12.0

505

43.5

26.1

65.2

60.7

97.1

336.9

Grants

1,781.9

1,962.9

1,493.0

1,485.4

1,232.2

1,066.2

1.137.8

1,462.2

1,712.4

1,279.1

1,008.6

627.0

693.1

660.7

654.7

714.5

819.9

746.0

2,725.6

Total

























8.8

37.7

107.5

84.5

310.9

359.6

1,403.2













21.9

67.8

92.3





4.3

103.1

51.3

62.0

73.1

30.8





PL480/665b Rupees Convertible

Table 1.1 External assistance to India 1966–85 (Rupees crores)a

1,781.9

1,962.9

1,493.0

1,485.4

1,232.2

1,066.2

1,159.7

1.530.0

1,804.7

1,279.1

1.008.6

631.3

805.0

749.7

824.2

872.1

1,161.6

1,105.6

4,128.8

Gross aid

1,032.5

947.5

849.1

803.9

800.7

796.0

820.7

754.7

686.9

626.0

595.8

507.4

479.3

450.0

412.5

375.0

333.0

274.6

685.8

Debt service

749.4

1,015.4

643.9

681.5

431.5

270.2

339.0

775.3

1,117.8

653.1

412.8

123.9

325.7

299.7

411.7

497.1

828.6

831.0

3,443.0

Net aid transfer

390.4

3,954.2

1,619.3

21,318.7

25,272.9

2,009.7 2,312.2

– 506.6

– 28,091.7

2,009.7 13,407.6

1,176.2 14,684.1

833.5

Source Government of India, Economic Survey 1985–6, Table 7.1 (172) and Table 7.4 (179). Notes a Conversions in rupees are at the pre-devaluation rate of exchange (US$ 1=R4.7619) up to the end of the Third Plan and at the post-devaluation rate of exchange (US$ 1=R7.50) for the subsequent years up to 1970–1. For the year 1971–2, pre-May 1971 exchange rates have been retained for conversion into rupees. For 1972–3, the rupee figures have been derived on the basis of the central rates which prevailed following the currency realignment of December 1971. From 1973–4, the quarterly average of the exchange rate of the rupee with individual donor currency has been applied to the corresponding quarterly data in respect of utilization for arriving at the equivalent rupee figure. For 1974–5, utilization figures have been worked out at current rates, which is the monthly average exchange rate of the rupee with individual donor currency. Utilization figures for 1975–6,1976–7 and 1977–4 are based on actual daily rates of the rupee with the donor currency on the respective dates, b PL480/665 is aid given by the USA under US Public Laws Nos 480 and 665. It has two different forms; one which generates counterpart funds in rupees; the other is ‘convertible’, and carries repayment obligations in hard currency, c Crore=10 million, d Figures for external assistance are minus loans from the IBRD and IMF.

1984–5

India’s Aid Resources in Macroeconomic Context 9 Table 1.2 Untied credits as a percentage of gross aid to India, 1965–82 To end Third Plan

33.4

1966–7

14.2

1967–8

18.9

1968–9

14.4

1969–70

19.9

1970–1

15.9

1971–2

18.5

1972–3

38.4

1973–4

42.0

1974–5

47.9

1975–6

45.4

1976–7

53.4

1977–8

13.6

1978–9

14.7

1979–80

13.8

1980–1

16.0

1981–2

13.4

Source Government of India, Economic Survey, 1982–3, Table 7.4 Note Figures recalculated net of IBRD loans.

Table 1.3 Average external assistance to India, 1958–72 (US$m)

Gross aid Debt service Net aid

3-year average 1958–9 to 1961–2

5-year average 1961–2 to 1965–6

5-year average 1966–7 to 1970–1

1971–2

733

1,208

1,330

1,123

73

232

492

626

660

976

838

497

Source Chaudhuri (1978), Table 32,99 Since the 1973 oil shock, roughly comparable data are available from the World Bank, which follows the OECD definition of aid. Table 1.4 shows India’s loan disbursements and debt service repayments (of both principal and interest), distinguishing between concessional (i.e, ‘aid’) loans and non-concessional loans. Table 1.5 shows India’s receipts of outright grants (also ‘aid’) by main donor source. The net transfer from the concessional loans is then aggregated with grants, where the netting off procedure does not apply, in Table 1.6. This indicates the aid transfers, in millions of current US dollars, for 1975–6 to 1985–6—and confirms (together with Table 1.2) that, after the major peak in net aid flows during the Third Five Year Plan, two further peaks occurred. One was during the 1975–6 ‘Emergency’, the other in 1980–1. Both, as we shall see, were reflected in ‘real’ peaks, though more modestly.

110

Commercial loans

112

90

21

61

258

30

24

90

92

76

752

IBRD

Bilateral loans

Commercial loans

Total publicly guaranteed

757

86

81

88

255

472

502

Non-concessional

91

810

77

76 960

96

62

127

285

260 107

637

38

675

1,236

61

1,175

60

14

180

254

14

580

327

921

78–9

515

35

550

1499 1,348

494

1,915

470

Multilateral loans

59

33

1,803 1,538 1,257

Bilateral loans

IDA

Concessional

Debt service payments

Total medium and long-term loans (excluding IMF)

Private non-guaranteed

Total publicly guaranteed

51

Bilateral loans

155

266

76

39

168

200

IBRD

636

Non-concessional

837

1,112 22

Bilateral loans

333

991

Multilateral loans

533

491

1,603 1,370

Concessional

IDA

75–6 76–7 77–8

Loan disbursements

1,012

93

51

130

274

695

43

738

1,361

79

1,282

79

37

149

265

28

443

546

1,017

79–80

422

1,037

90

48

174

29

176

379

5

679

72

756

1,010 1,135

84

35

138

257

4

60

753

1 141

585

2,442 3,000

689

279

473

14

288

775

55

476

2,020 2,415

311

7

421

739

24

471

786 1,109

1,281 1,640

1,279

260

24

244

528

7

653

91

751

3,176

639

2,537

708

11

470

1,189

17

457

874

1,348

1,458

474

34

274

782

8

559

109

676

3,687

835

2,852

1,070

185

291

1,546

30

453

823

1,306

81–2 82–3 83–4 84–5

707

50

758

2,769

285

2,484

252

36

174

462

718

652

652

2,022

80–1

Table 1.4 India’s loan disbursements and debt service, 1975–6 to 1985–6 (US$m)

1,881

740

66

325

1,131

34

592

124

750

4,586

1,135

3,451

1,179

296

329

1,804

28

572

1,047

1,647

85–6

830

78 838

81 890

80 1,055

95

15.1

19.8

112.9

Multilateral consortium

60.0

128.2

120.9

81.4

154.3

UK

USA

2.5

41.9

Switzerlanda

2.9

58.2

b

Sweden

9.6

9.0

Norwaya

Netherlands





a

Japan



4.2



12.1

W. Germany



10.0

62.3





394.8

1976–7

Italya

5.1

4.3

France

Canada

Denmarka

0.7

65.4

Belgium



413.1

1975–6

Austria

Bilateral consortium

117.8

111.1

175.6

2.9

67.3

15.2

6.5









4.9

22.4





406.4

1977–8

135.3

133.1

258.3

6.0

57.0

21.2

5.1





6.3



14.2

5.8





506.9

1978–9

93 1,105

164.2

165.3

282.7

16.8

52.6

22.1

30.4

27.2



1.3



15.2

18.4





632.0

355

1,132 1,490

122

184.3

124.9

307.7

17.1

64.9

22.4

28.7

35.7



7.8



19.6

3.3





632.0

206.0

143.9

199.1

5.3

49.1

19.7

23.6

5.2



0.8



10.0

7.6





464.4

1980–1 1981–2

1,158

121

1979–80

Table 1.5 India’s grant disbursements, 1975–6 to 1985–6 (US$m)

Source World Bank

Total medium and long-term loan (excluding IMF)

Private non-guaranteed

233.0

93.2

164.5

18.8

56.9

20.9

50.2

14.7

0.4

1.9



16.7

10.7





448.9

2,799

918

182.1

109.7

128.4

22.3

53.1

20.4

29.8

9.7

0.2

0.1



16.0

14.1





403.8

1983–4

1,991

533

1982–3

1,740

461

143.6

102.6

132.5

21.7

35.1

19.9

36.8

13.6

4.2

0.3



17.2

15.8





399.7

166.4

84.4

151.9

11.9

43.0

23.0

40.9

11.8

5.9

1.4



11.2

12.3





397.8

1984–5 1985–6

462.3

7.5

454.8 524.2



524.2 642.3



642.3



114.0

21.3

797.7

1.4

796.3



109.8

54.4

822.0

5.7

816.3



103.4

80.9

670.4



670.4



112.0

94.0

681.9

1,109

541

1,650

Net position on loans

Total grant disbursement

Net position on loans and grants

Source Tables 1.3 and 1.4

494

1,603

of which debt service

Concessional loan disbursements

1,330

462 965

524

441

550

502

868

991

1,370

888

642

246

675

921

1,071

798

273

738

1,017

1975–6 1976–7 1977–8 1978–9 1979–80

2,086

822

1,264

758

2,022

1980–1

585.9



585.9



116.4

65.7

543.3



543.3



101.5

42.1

1,198

670

528

753

1,281

1,566

682

884

756

1,640

1,183

586

597

751

1,348

1,173

543

630

676

1306

1,461

564

897

750

1,647

1981–2 1982–3 1983–4 1984–5 1985–6

Table 1.6 India’s net transfer of concessional finance, 1975–6 to 1985–6 (US$m)



681.9



110.4

122.6

Source World Bank Notes a Relates to calender years; e.g. the figures for 1980–1 relate to calender year 1980, and so on. b Relates to the year July-June; e.g. the figures for 1980–1 relates to period July 1980-June 1981, and so on.

541.3

15.3

Grand total

526.0

Total consortium



85.0



60.0

60.0



32.8

52.9

Others

IFAD

UN

a

EECa

564.2



564.2



108.0

58.4

India’s Aid Resources in Macroeconomic Context 13 The ‘Emergency’ aid peak (1975–7) was a lagged response to India’s severe balanceof-payments difficulties in 1972–5.4 It was not a response of western aid donors to the declaration of a State of Emergency in India in mid-1975, as some people suspected at the time. The Aid-India Consortium did meet very soon after the Emergency was declared, and the UK raised the question of whether aid flows should be cut in response to the political changes of June 1975, but this suggestion was not acted on. The second minor peak of the early 1980s in aid transfers to India was a response to the serious deterioration of India’s economic performance in 1979–80.5 After the return to power of a Congress (I) government in 1980, it was decided both to go to the IMF for a large medium-term loan, and to seek greater aid financing. In November 1981, a standby agreement was reached with the IMF for aid credits totalling US$5.8bn over the years 1981–2 to 1984–5. In the end only 534 rupees crores (about US$650m) were actually utilized (Government of India Economic Survey, 1986,175).6 Simultaneously, aid loans from IDA and other multilateral sources were increased from 1980–1, and sustained through 1981–2 and 1982–3. At this point, India turned, much more than ever before, toward commercial loans for balance-of-payments support. How much could India purchase with these aid flows of nominal dollars? That depended on the fluctuating purchasing power of the dollar. The current-dollar figures should ideally be deflated by dividing, in each year, the ratio of a dollar price index of Indian imports in that year to the same price index in the base year.7 This would give the trend in the purchasing power of aid to India. Not much work has been done in this area. Table 1.7 lists OECD estimates of net aid flows (i.e. net only of capital repaid) to India, including estimates for flows from OPEC and Comecon countries. These net flows are deflated by the US producer price index as a proxy for import inflation facing India. These figures suggest that the nominal aid peaks discussed above were reflected in real aid peaks around 1974–6 and 1980, the latter peak being smaller and shorter. Another estimate (Rubin, 1982) for real non-food aid produced broadly similar results. He took his calculations back to 1950, and concluded that in real terms 1963–6 experienced the highest ever aid inflows. Table 1.7 Official development assistance disbursements to India, 1971–85 (US $m) Gross ODA

Nett ODA

1,083.1

1,002.8

972.1

1972

728.6

614.0

569.6

1973

1,009.0

779.5

638.2

1974

1,568.4

1,365.6

941.7

1975

2,004.6

1,708.4

1,078.6

1976

2,056.8

1,820.6

1,098.8

1977

1348.9

1,078.3

613.0

1978

1,749.8

1,338.6

796.2

1979

1,755.0

1,370.1

643.6

1980

2,664.0

2,256.2

927.3

1981

2,389.0

1,920.4

722.8

1971

Net ODA (1970 US$m)

14

Does Aid Work in India?

1982

2,022.0

1,545.1

570.1

1983

2,203.2

1,742.8

635.0

1984

2,022.1

1,541.1

548.7

1985

1,946.0

1,469.7

525.3

Source OECD, Geographical Distribution of Financial Flows to Developing Countries: Disbursements and Commilments, 1971–7, 1978–81, 1982–5. Note Calculated using US producer price index as deflator.

Aid relative to macroeconomic categories Although, for reasons already explained, India receives only small amounts of aid relative to its area and its population, the macroeconomic impact may still be significant. Aid can be seen as supplementing a constrained supply of foreign exchange and/or adding to total domestic investment. In this connection, one would be interested in the size of aid relative to the level of imports and the balance of payments and/or domestic savings and investment. The role of aid in permitting India to grow relatively unconstrained by deficits on the balance of payments in the years 1956–70 is hinted at in Table 1.8 (derived from Goldsmith, 1983). This shows that, between 1956 and 1970, India’s deficit on trade in goods and services (falling from three to two per cent of GNP) was largely financed by a surplus of just less than two per cent of GNP on central government capital account. This surplus does not, of course, include only aid inflows; some of it comprises non-concessional borrowing by the central government. However, this was relatively small; so the surplus on ‘capital, central’ is not merely a maximum estimate, but in fact a good proxy, for the true contribution of net aid to India’s balance of payments. Table 1.8 India’s balance of payments 1948–70 (as a percentage of GNP)

Commodity trade Imports

1948–50

1951–5

1956–60

1961–5

1966–70

−1.33

−1.04

−2.52

−1.81

−0.97

8.32

9.29

7.21

6.61

0.08

0.37

−0.43

−0.79

−0.89

a

Services Transfers, private

0.32

0.38

0.38

0.16

0.35

Transfers, government

0.00

0.21

0.23

0.28

0.13

Capital, private

−0.21

−0.04

0.07

0.02

−0.02

Capital, central

−1.03

0.08

1.43

1.96

1.86

2.13

0.25

0.98

0.19

−0.33

Monetary authorities Miscellaneousb Net errors and omissions





0.03

0.00

0.01

0.04

−0.21

−0.17

−0.06

−0.15

Source Goldsmith (1983), Table 3.16, 162 Notes a Calculated from: United Nations, Yearbook of National Accounts Statistics, 1957–75; Government of India, Economic Survey, various issues b ‘Miscellaneous’ includes deposit money banks and allocations of SDRs

India’s Aid Resources in Macroeconomic Context 15 In the second half of the 1960s, when aid inflows were at the height of their importance relative to key macroeconomic variables, aid was financing about a third of India’s imports of goods and services. By the mid-1980s, the aid/imports ratio had fallen to 5 per cent (Table 1.9). The aid peak of 1975–6 financed less than 25 per cent of imports in that year, while the 1980–1 peak financed only 5 per cent. The falling aid/import ratio partly reflects the fact that imports rose from less than 7 per cent of GNP in the late 1960s to over 8 per cent in the late 1970s and over 10 per cent in the early 1980s (Tables 1.8, 1.10). By 1980–1, India had to face an import bill which suddenly had jumped by almost half while exports remained stagnant due to the rising real exchange rate. Aid was no longer available to India in sufficient quantities to cope with all her external finance requirements, and from this point both non-concessional public lending and commercial loans entered significantly for the first time into India’s external financing package. This picture is further illustrated by Table 1.10. This uses World Bank figures of India’s balance of payments in a form which compares the situation in 1975–85 as closely as possible with Goldsmith’s estimates in Table 1.8 for the period up to 1970. A strong positive factor in the balance of payments in the middle and late 1970s had been the inflow of private remittances from abroad. In 1975–80, current transfers to India (net of transfers out) exceeded one per cent of national income for the first time. This was partly as a result of increased labour migration to West Asia to supply labour shortages in new-rich OPEC countries, and partly a response to a government decision in November 1975 to allow non-resident Indians to open interest-bearing accounts of foreign exchange in India. These rapidly rising remittances tended to divert attention from the slowdown in export growth, and a sense of balance-of-payments security was built up as the foreign exchange reserves accumulated— from two months’ worth of imports in 1974–5 to nine months’ worth in 1978–9. Table 1.9 Ratios of aid utilizations to imports in India, 1966–84 Year

Imports

Ratio

1966–7

Aid utilizations 831.0

2,194.5

37.9

1967–8

828.6

2,135.7

38.8

1968–9

497.1

1,858.4

26.7

1969–70

411.7

1,677.1

24.5

1970–1

299.7

1,770.3

16.9

1971–2

325.7

2,002.2

16.3

1972–3

123.9

2,146.6

5.8

1973–4

412.8

2,729.3

15.1

1974–5

653.1

4,156.9

15.7

1975–6

1,117.8

4,744.1

23.6

1976–7

775.3

4,816.9

16.1

1977–8

339.0

5,541.0

6.1

1978–9

270.2

7,397.5

3.7

1979–80

431.5

9,575.7

4.5

16

Does Aid Work in India?

1980–1

681.5

12,543.6

5.4

1981–2

643.9

13,886.5

4.6

1982–3

1,015.4

14,913.2

6.8

1983–4

749.3

16,039.3

4.7

Source Table 1.1; and Government of India, Economic survey, various issues.

The vulnerability of India’s balance of payments was, however, cruelly exposed when the second oil shock of 1979 was followed by the severe drought of 1979–80. The response of aid inflows in this crisis was far too weak to finance India’s unprecedented deficit on commodity trade (almost 4 per cent of GDP in 1980–5). Aid in the form of grants remained constant in nominal dollar terms, and fell as a proportion of nominal GDP. Aid in the form of loans at concessional rates rose quite sharply in nominal dollars, and probably also as a proportion of nominal GDP. But even so, in order to finance the deficit of the early 1980s, India had to tap other sources. These were medium- and long-term public loans (less concessional than aid); commercial loans (by definition at full market rates); its facilities with the IMF; and its own foreign exchange reserves. Although aid still finances a significant proportion of India’s imports, particularly those directly accessible to the public sector, its lower level than during the 1970s—and perhaps its reduced responsiveness to crises—now render aid insufficient as India’s main instrument for coping with balance-ofpayments crises. Table 1.10 India’s balance of payments, 1975–6 to 1984–5 (as a percentage of GDP at factor cost) 1975–6 to 1979–80 Commodity trade Imports

−1.44

1980–1 to 1984–5 −3.97

8.68

10.95

Services (non-factor)

+0.66

+0.70

Factor incomes (net)

−0.04

−0.13

Current transfers (net)

+1.10

+1.63

Capital, private

+0.00

+0.02

Capital, central

+1.23

+1.39

official grant aid

+0.34

+0.24

medium and long loans (net)

+0.89

+1.15

Net credit from IMF

−0.12

+0.56

Capital flows NEI

−0.13

−0.24

Errors and omissions

+0.02

n.a

Changes in reserves

−1.29

+0.14

of which:

Source Author’s calculation based on World Bank figures Note A minus sign indicates increases in the reserves

India’s Aid Resources in Macroeconomic Context 17 Aid is said to be important not only as a supplement to scarce foreign exchange, but also as a means of permitting the recipient country to invest (i.e. add to physical capital) more that it can finance out of domestic saving.8 In particular aid permits governments to finance higher levels of investment than can be financed out of the current surplus of taxation over public current outlays or from borrowing from the private sector. What then has been the statistical record on aid flows to India relative to its investment? Many of the available data on saving and investment are over-aggregated, and include some non-aid foreign inflows.9 Only in special conditions can one regard such inflows from the ‘rest of the world’ (ROW) as a good proxy for the contribution of aid. They are when the non-aid elements in the ROW surplus are very small, as they were in India for most of the period until the mid-1970s. Table 1.11 Sectoral financial surplus/deficits in India, 1951–75 (Rbn and percentages of GNP) 1951–5

1956–60

1961–5

1966–70

1971–5

1951–75

Absolute figures (Rbn) Financial institutions

0.41

0.94

2.79

3.50

1178

20.42

Private corporations

−4.16

−8.24

−15.96

−18.07

−29.73

−76.16

Government

−8.82

−28.81

−44.26

−70.18

−123.86

−275.93

Households

8.81

18.49

33.07

45.63

130.74

236.34

1.40

17.19

20.32

29.02

5.45

73.39

23.60

73.67

116.40

166.40

307.57

682.24

Rest of the world All sectors; absolute

Distribution (per cent) 1.7

1.3

2.4

2.1

4.2

3.0

Private corporations

−17.6

−11.2

−13.7

−10.9

−9.8

−11.2

Government

−37.4

−39.1

−38.0

−42.2

−40.9

−40.4

Households

37.4

25.1

28.4

27.4

43.2

34.6

Financial institutions

Rest of the world All sectors; absolute

5.9

23.3

17.5

17.4

1.8

10.8

100.0

100.0

100.0

100.0

100.0

100.0

Relation to gross national product (per cent) Financial institutions

0.08

0.14

0.28

0.21

0.44

0.30

Private corporations

−0.78

−1.25

−1.60

−1.07

−1.02

−1.22

Government

−1.69

−4.35

−4.45

−4.15

−4.24

−4.07

Households

1.69

2.81

3.32

2.70

4.47

3.48

Rest of the world

0.27

2.61

2.04

1.71

0.19

1.08

All sectors; absolute

4.51

11.81

11.67

9.83

10.35

10.15

Source Goldsmith, 1983, Table 3.15, 161

Historical figures (Goldsmith, 1983) for 1950–75 are set out in Table 1.11. Of course, for India as a whole, investment in any year must be equal to (i.e. precisely financed by)

18

Does Aid Work in India?

savings in that year, but there can be surpluses or deficits for each major sector’s saving over its investment. Private corporations and the government have consistently invested more than they saved, while financial institutions, households and the rest of the world have consistently saved more than they have invested.10 Between 1951 and 1975, the rest of the world supplied 20 percent of the combined deficits of the government and private corporations, or 25 per cent of the government’s deficit on its own. Dependence on a foreign surplus was heaviest during the Second Plan, when it supplied just under half of the combined government and private corporate deficits. From 1961 onwards, corporate and public deficits came progressively to depend, for finance, less on foreign inflows and more on domestic private savings,11 as the household sector’s surplus steadily increased. This increase itself represents partly the growth of household financial assets associated with increases in the scope and sophistication of financial intermediaries and, after 1975, interest-bearing foreign exchange accounts of non-resident Indians. It may also partly reflect the improved accounting of household savings. Table 1.12 Ratios of gross aid utilizations to saving and investment in India, 1975–86 (Rupees crores) 1 Aid Utilizations

2 Domestic Savings

3 1 as % of 2

4 Domestic investment

5 1 as % of 4

1975–6

1,804.7

16,362

11.0

16,418

11.0

1076–7

1,536.0

18,845

8.1

17,705

8.6

1977–8

1,159.7

19,935

5.8

18,621

6.2

1978–9

1,066.2

22,709

4.7

22,984

4.6

1979–80

1,232.2

25,372

4.9

26,143

4.7

1980–1

1,485.4

29,239

5.1

31,457

4.7

1981–2

1,493.0

33,411

4.5

36,229

4.1

1982–3

1,962.9

37,730

5.2

40,476

4.8

1983–4

1,781.9

44,559

4.0

47,255

3.8

1984–5

2,009.7

50,365

4.0

53,844

3.7

1985–6



56,682



61,518



Sources Column 1 from Table 1.1; columns 2 and 3 from World Bank figures.

From 1975–6 to 1978–9, foreign savings made a negligible, or even negative, contribution to India’s disposable income (Table 1.12, column 4 minus column 2). (The negative contribution arises because repayments of interest on part-aid loans and negative non-aid foreign savings outweighed net-aid utilizations, which remained positive though small). It is in 1980 that the role of foreign savings changed dramatically. In the early 1980s foreign savings climbed back to about 1.4 per cent of disposable income (Table 1.10), within sight of the 2 per cent peak of the early 1960s. However, aid contributes much less to foreign savings in the 1980s than it did in the 1960s; much more is contributed by the swelling proportions of remittances, non-aid multilateral loans, and commercial borrowings.

India’s Aid Resources in Macroeconomic Context 19 To sum up: as a result of the rise in domestic saving and investment relative to sluggish aid, domestic capital formation depends much less on aid in the 1980s than under the Third Plan. Table 1.12 confirms the dwindling size of aid utilizations in relation to domestic savings and investment, even between 1975 and 1985; both ratios fell from about 11 per cent to about 4 per cent. The recent diversification of foreign finance, and the longerterm rise in domestic savings and investment, have greatly eroded the contribution of aid to India’s accumulation process, since the heyday of post-Independence development planning. The role of aid in financing government development expenditure exhibits a similar pattern, as indicated in Table 1.13.12 Aid utilizations as a percentage of public-sector development outlays fell from around one-third in the mid-1960s to less than 5 per cent in the mid-1980s. The ratio of aid utilization to gross capital formation out of central government budgetary resources (which includes finance for capital formation elsewhere in the economy) also fell (Table 1.13), from almost 70 per cent in 1967–8 to just above 10 per cent in the mid-1980s. As with the ratio to development outlays, the decline is marked after the end of the mid-1970s aid peak. However, some gross public-sector capital formation—now probably at least close to one-half—is not normally eligible for aid.13 Of ‘new gross public development investment’, we judge that utilized aid still comprises about a fifth. Impact of aid on economic growth As we have seen, because of statistical difficulties, it is a complex task even to set out the historical record on aid flows to India. Additional problems arise in assessing whether, let alone how, economic performance has been determined by aid. First, good (or bad) performance can cause changes in the level of aid: to reward success (or compensate for failure). Second, even if this can be allowed for, as in ‘simultaneous’ econometric models where growth14 can be represented as part cause and part effect of aid (Mosley, 1987), aid levels are only one of many things that may determine growth. The macroeconomist cannot, unfortunately, hold constant all the other variables, apart from aid, that are relevant to economic performance—nor is it clear that one should, since some such variables may themselves be affected by aid.15 One cannot inspect changes in economic growth, compare them with earlier changes in the inflows of aid, and draw simple causal conclusions, on the assumption that one variable must be influencing the other because nothing else has changed. Everything else is changing at the same time: sometimes erratically (weather, power supply); sometimes as a cause or effect of past aid or growth themselves. Although econometric models can cope with some of this—and do provide some indications of aidgrowth relationships—any attempt to distil the impact of aid in such a multiplex scene must be more a matter of judgement than a matter of precise calculation. This is not only because of the under-development of econometrics (Learner, 1982) and the complexities of reciprocal or of exogenous causation. It is also because of the difficulty of judging what the Indian Government, households, firms, and trading and lending partners would have done—and with what impact on growth—if aid levels had been substantially different, or had sharply changed (not the same thing).

20

Does Aid Work in India?

Table 1.13 Ratios of gross aid utilizations to government expenditure categories in India 1966–85 (Rupees crores) Gross aid Development Ratio utilizations outlays (per cent)

Gross capital formation out of central government budgetary resources

Ratio (per cent)

1966–7

1,105.6

3,416

32.4

1,793

61.7

1967–8

1,161.6

3,645

31.9

1,675

69.3

1968–9

872.1

3,937

22.2

1,660

52.5

1969–70

824.2

4,166

19.8

1,612

51.1

1970–1

749.7

4,858

15.4

1,917

39.1

1971–2

805.0

5,405

14.9

2,217

36.3

1972–3

631.3

6,550

9.6

2,628

24.0

1973–4

1,008.6

6,864

14.7

2,665

37.8

1974–5

1,279.1

9,403

13.6

3,677

34.8

1975–6

1,804.7

11,574

15.6

4,663

38.7

1976–7

1,530.0

13,134

11.6

4,991

30.7

1977–4

1,159.7

15,005

7.7

5.899

19.7

1978–9

1,066.2

17,994

5.9

6,913

15.4

1979–80

1,232.2

20,298

6.1

7,229

17.0

1980–1

1,485.4

24,426

6.1

9,012

16.5

1981–2

1,493.0

28,653

5.2

10,799

13.8

1982–3

1,962.9

33,591

5.8

12,404

15.8

1983–4

1,781.9

39,274

4.5

14,702

12.1

1984–5

2,009.7

48,003

4.2

17,872

11.2

Source Government of India, Economic Survey, 1968–9 to 1985–6.

Before considering the aid-growth link, we furnish some indicators of India’s growth (Table 1.14). Over 1950–75, GNP grew at over 3 per cent a year which, with population growth of 2 per cent a year, allowed income-per-person to rise by an average 1 per cent a year. Although this represents a much better performance than was achieved in pre-Independence India (Heston, 1982), it is nevertheless modest when compared with the absolutely low starting level of income, with the plans of the government and with the experience of many other poor countries in Asia. (It should be noted that the growth rates in the 1960s are misleading: 1965–6 and 1966–7 were years of disastrous drought—the worst years for agriculture in the whole period. This leads to an understatement of the agricultural growth achieved from 1960–1 to 1965–6, and a corresponding overstatement of growth in the subsequent five-year period—in fact 1965–7 saw the beginning of the ‘green revolution’ in Punjab, Haryana, and Western Uttar Pradesh.) Economic growth in 1971–5, however, was clearly lower than the 1950–75 average rates. This lends some plausibility to the

India’s Aid Resources in Macroeconomic Context 21 hypothesis of secularly slackening growth between 1950 and 1975. However, this result is very sensitive to the precise periods chosen. Further, growth over 1976–85 appears to be significantly higher than the 1950–75 average.16 Juxtaposing Table 1.14 with Tables 1.7 and 1.3, we might conclude that growth seems to have been highest when real aid utilizations were lowest! Of course, this does not establish a causal association; high aid might be a response to growth difficulties; or rapid growth could even be a lagged response to aid flows in earlier periods.17 The heart of the matter is the question of the dynamic of growth. This dynamic is still heavily influenced by the performance of agriculture, although statically this might seem to be less important today; after all, agriculture’s share of total output fell, from 45 per cent in 1960–4 to 41 per cent in 1975–6, and to 31 per cent in 1985–6. Nevertheless, it is still not easy for the economy’s overall growth rate to exceed greatly the rate of growth of agriculture. Two in three Indians depend on agriculture for income. It is these Indians who, being for the most part relatively poor, are likely to spend (rather than save) extra income. If agriculture does not grow, there may be inadequate demand for the products of growth in other sectors. In particular, Indian workers—if in the poor half of India’s population— spend over two-thirds of extra income on food. If food production does not grow, then rising output and income involving many workers—e.g. in parts of the service sector— becomes a source of unsustainable inflationary pressure. (There is some evidence that the latter is happening in the India of the 1980s.) If agriculture constrains growth, there is a further problem. Since the new possibilities for extending the area of cultivation, e.g. by irrigation, are constantly shrinking or becoming more costly, the growth rate of agriculture itself depends on continuously finding new methods to raise the productivity of land. Such methods must be profitable for farmers if they are to be adopted. But falling world and Indian agriculturalprices undermine the potential for profitable innovation. The falling prices themselves are due in part to subsidization of food production in the EEC, Japan and the USA; in part to India’s own past adoption of improved methods; but mostly to the concentration of the gains from growth in India on richer people, who have a low incomeelasticity of demand for food. India’s industry, for both reasons, is still influenced strongly by the performance of agriculture: through its supply of food as an urban wage-good, its supply of raw materials,and its major role as a source of demand for industrial consumer goods. A strangely steady 70 per cent of India’sworkforce is engaged in agriculture (World Bank, 1987, 264); that proportion usually drops with growth, but not in India. One possible ‘escape route with growth’ would be to encourage labour intensive small industry, which would reduce this proportion. However, centrally-run or—influenced projects—private or public, aid-financed or not—have so far made little contribution to rural non-farm activity.18 Apart from this, growth is influenced through a variety of mechanisms which determine whether the total productivity of the resources invested in industry rises or falls. This can be viewed in terms of the familiar Harrod-Domar identity: the rate of growth equals the savings ratio divided by the incremental capital-output ratio, which in turn is the sum of the sectoral ICORs weighted by sectoral shares in output. As the share of manufacturing in GNP rises with development, its ICOR becomes increasingly important. In India it

Industrial production

Labour force

Capital stock

Area under cultivation

4.

5.

6.

7.

2.74

1.21

8.00

4.22

3.70b 1.12

1.51

7.16

3.80

2.10

1.89

4.03

1956–60

0.30

1.97

9.06

-1.39

2.39

0.60

2.90

1961–5

Source Goldsmith (1983), Table 3.1, 139; and World Bank figures. Notes a 1951–70 b 1950–60 c 1950–70 d Applies to major crops only, 1977–86

Agricultural production

3.

1.53

1.87

b. Per head

Population

3.45

a. Total

Real GNP

2.

1.

1951–5

5.77 1.24

2.40

3.43

6.65

2.23

2.40

4.82

1966–70

Table 1.14 Key growth rates of the Indian economy, 1951–85 (per cent per annum)

0.60

3.60

2.39

2.10

0.89

3.00

1971–5

1.20

4.93c

1.78a

6.22

3.10

2.07

1.53

3.63

1951–75

0.09d

n.a

n.a

5.07

1.84

2.30

1.91

4.21

1976–85

India’s Aid Resources in Macroeconomic Context 23 appears to have suffered a twenty-year increase (1960–80) across the board, not related to compositional changes in the structure of manufacturing (Ahluwalia, 1985, 132), nor to particular industries which are especially capital-intensive in their operating technology (Ahluwalia, 1985, 141). The fairly stable growth of GNP in 1950–80 reflects the fact that though the savings ratio doubled, the incremental capital-output ratio almost doubled too. These findings have raised a number of crucial policy questions,19 some specifically related to aid. It almost certainly raised the savings ratio; did it also reduce capacity utilization, or the efficiency with which utilized capital produced output, thus raising the ICOR? Aid—through its effects on the level of demand; by selection of sectors; or by choosing simple or troublesome technologies—can affect the level of capacity utilization. Perhaps most importantly, the allocation of aid resources—both between and within sectors—may concentrate them unduly on activities with high capital-intensity and low labour share. This would lead to slower and less equitable growth, both by raising the capital-output ratio, and by placing extra income in the hands of persons relatively less likely to spend it on extra Indian production. We calculated comparable figures to Goldsmith’s for 1950–75 from World Bank data for 1976–85, and added them as a final column in Table 1.14. What do they tell us of India’s recent economic performance? The dominance of agriculture continues to be eroded, and economic growth is less tightly tied to agriculture’s own growth performance. Although the figures indicate indifferent agricultural growth from 1976 to 1985, this is critically dependent on the years chosen. A collection of detailed studies of Indian agriculture in 1951–81 (Jakhade, 1983) rejected the hypothesis of secular deceleration; since 1981 (except in the disastrous 1987 drought) there has, if anything, been acceleration. Overall, India’s total real GNP grew faster in 1976–85 than in the relatively buoyant 1950s, and certainly much faster than during the long years of slowdown and stagnation from 1960–75. The hypothesis of secularly slackening growth seems implausible with respect to the last ten years. The medium-term growth of real per capita product was also faster than in any comparable period since Independence, despite an apparent upturn—already being reversed20—in the rate of population growth. Year-to-year growth, however, was still significantly influenced by the state of the harvest; possibly the variability of food (and farm) output, though not its worst-case levels per person, had deteriorated after 1960 or so (Lipton with Longhurst, 1985, discussing work by Hazell and others). Cycles with troughs in the drought years of 1979–80 and 1982–3 are evident when the decade’s growth rates are disaggregated by sector. The industrial growth rate overall recovered from the previous period of stagnation, but not dramatically. Services have been the fastest growing of the sectors, which must raise some questions about how robust recent growth will prove to be, particularly with respect to the growth of government services. The sedate pace of industrial growth was disappointing, especially given the rise in the investment rate from 22 per cent in 1975–6 to 25 per cent in 1985–6. So much for major indicators of India’s macroeconomic performance. What of the impact of aid upon it? Although economic growth was fastest in 1955–70, when aid made its greatest relative contribution to the balance of payments and to the financial deficits of the government and corporate sectors, aid was falling relative to these aggregates throughout this period. The peak in aid around 1980, when there may have been some acceleration in the underlying growth rate, was nevertheless low relative to macroeconomic

24

Does Aid Work in India?

aggregates by historical standards. The associations cannot by themselves be regarded as conclusive evidence of an aid/growth link. Economists have applied econometric tests to the aid and growth data, seeking more rigorous evidence for or against a link. Predictably, with different researchers testing differently specified equations with different data sets, the results have appeared to be contradictory. One might have thought that econometric studies of the aid/growth link would have taken changes in national income as the dependent variable and aid utilizations (net of repayments with a balance of payments constraint, gross with a savings constraint: Lipton, 1972), along with domestic saving and non-concessional foreign finance, as the independent variables, thus linking the three main sources of investment funds with output changes in a simple Harrod-Domar type of formulation. This has been attempted in crosssection studies, which generally show some positive effect of aid on growth (Mosley, 1987, especially 130). Econometric studies of aid to India, however, take a different tack, using investment as the dependent variable and aid (along with various government budget indicators and sometimes the level of national income itself) as the independent variable. For example, Chaudhuri (1978; 107–8) tested an equation of the form: It=a+bYt +cAt where I is investment, Y is net national product and A is gross utilized aid (including a small but unspecified amount of concessional finance with a grant element too low to constitute aid) minus food aid. Food aid was excluded on the grounds that it was not intended to finance public investment.21 The following result was obtained:

This result had encouragingly high values of the correlation coefficient and the t-statistic on the A variable, while the Durbin-Watson statistic is sufficiently near to 2 that autocorrelation of the Y and A variables need not be a serious worry. The analysis thus suggests that aid had a fairly strong positive effect on the level of investment. Other analysts relying on econometric studies have not been in agreement. A Task Force of the Asian Development Bank used time-series data for India, Malaysia, Sri Lanka, and Thailand to examine whether foreign aid added to gross government investment (ADB, 1977, 305–6). Government capital formation was regressed on two different measures of foreign aid, plus the government budget deficit, the size of government domestic borrowing and two measures of the foreign exchange gap. The conclusion was that foreign savings added to government capital formation in only one of the four countries—Sri Lanka. This result seems odd, for it has hardly ever been seriously questioned that aid raises government investment; the debate has been over whether aid ‘crowds out’ private investment. Not unexpectedly, with such radically differing results, research has continued in an attempt to resolve the conflict The most elaborate recent attempt is by Rubin (1982). He regressed Indian public sector gross fixed capital formation on foreign aid inflows and an array of other variables—government consumption, agricultural output and domestic

India’s Aid Resources in Macroeconomic Context 25 and foreign inflation. His definition of aid was similar to Chaudhuri’s, in that it excluded food aid, but his figures were net, rather than gross, of repayments and deflated (by the US producer price index) to a 1970 price base. His initial finding concurred with that of the ADB study quoted above—that foreign aid had ‘no significant effect on public investment in India’ (Rubin, 127). However, when the flow of foreign aid was added to the stock of foreign exchange reserves, the combined variable—admittedly a bastardized one called ‘foreign resources’—did have a significant explanatory effect (25 per cent of the change in government capital formation) with an R2 of 0.88. This result was interpreted as follows: Aid does not simply ‘finance’ projects, in the sense that it provides for expenditures that would not have been undertaken otherwise… Foreign aid appears to be as much a subsidy to maintain reserves as a way of financing new spending (Rubin, 1982,129–30)

However, this conclusion can just as well be interpreted to mean that—given the permissible depletion of reserves—new aid meant new public investment. Any excess of investment over saving is financed by aid, other foreign inflows, or depletions of foreign-exchange reserves. These and similar econometric studies may be too narrowly focused. In the first place, as the above quotation suggests, the effectiveness of aid in changing economic performance must judged against some set of assumed objectives—including permissible levels of reserves—which policy-makers are trying to achieve. Looking only at the impact of aid on investment tends to suggest that policy-makers are (or should be) trying to maximise the amount of investment; it is far from obvious that they do, or should. We need to know whether aid-financed projects have increased or reduced the return to investment. Secondly, to look at investment as a single category is too aggregative. Aid may stimulate different categories of investment differently. Also the marginal productivity of investment will not (unless we assume an economy free of all types of distortion) be equal in different sectors. In such conditions, one needs a much more elaborately specified model of the ‘pathways of aid effectiveness’ (e.g. Mosley, 1987, 142–8) before econometric techniques are able to yield much insight. Some would go further than this to argue that the impacts of aid are so sector-selective that, in principle, econometric methods applied at the macroeconomic level are inappropriate to capture them. We do not go deeper into that question here. However, as one illustration of some of the difficulties involved in tracing aid impacts, it is worth noting the relationship between growth and aid in the form of emergency food imports to India. If the government—absent such emergency aid—would not have acquired commercial food imports, the extra food aid does not directly influence investment, but boosts consumption. This helps to break important bottlenecks in productive activity and adds substantially to the productivity of existing resources, including labour. Certainly food aid, which was excluded from most of the econometric studies, at one time played this role in the Indian economy. Sen has pointed out that, since Independence in India, it has become almost impossible for a famine to take place: No matter how and where famine threatens—whether with a flood or a drought, whether in Bihar in 1967–8, in Maharashtra in 1971–3, or in West Bengal in 1978—an obligatory policy response prevents the famine actually occurring. (Sen, 1983,757)

26

Does Aid Work in India?

This excellent record of swift public action to avert famine must have been facilitated to some degree by authorised foreign aid funds, which could be utilized at critical moments (with donor agreement) to pay for additional imports of food for the public distribution system. If one argues that, without aid, the Government of India would have had to pay for such food imports itself, then this kind of food aid can be credited with relaxing the foreignexchange constraint on investment which existed until the mid-1970s—a role increasingly played, as food aid dwindled and releasable public-sector food stocks grew, by depletions of such food stocks. Food aid, in either case, maintains both working capacity and mass demand for non-food products. In India, success in averting acute hunger co-exists with the ability to tolerate endemic malnutrition (Sen, 1983). If emergency food imports financed by aid contributed to the former, can one argue that recurrent food imports, for example under US Public Law 480, contributed to the latter, by blunting the incentives of domestic food producers? Before 1965, when PL 480 wheat imports were available against rupee rather than foreign exchange payments, it seems likely that such aid did indeed have some effect in depressing supplies from domestic wheat producers. The required calculation is the net effect of the increase in supply due to aid minus the loss in supply due to depressed domestic prices. The most reliable estimate of this net effect (see p. 51) is that every ton of food aid added five-sixths of a ton to total Indian supplies of grain. This is simply a supply-side calculation and neglects the demand shifts resulting from subsidization. In any case, it has been argued that the depressive effect on supply arose not from food aid as such, but from the system of urban food distribution at subsidized prices (Chaudhuri, 1978, 102–3). The criticism applies not to food aid per se but to food aid when channelled through the particular Indian environment of distributional controls. In this sense, it is analogous to Kidron’s criticisms of the effects of private foreign investment in India (see Chapter 7). Thus recurrent food aid, like emergency aid-financed food imports, had, on balance, beneficial effects. But the real point of the food aid illustration is that these effects would not show up at all in the results of the kinds of econometric investigation of the aid/growth link which have been commonest in the literature on India. India’s future capital inflows: aid or commercial lending? So far we have said little about sources of finance for capital formation, other than aid and domestic savings. Is there no non-aid foreign inflow which could offer an alternative to aid? Indeed, throughout the 1960s and 1970s: Without doing much violence to reality, the role of foreign capital in resource mobilisation can be discussed in the form of the role of foreign aid in Indian economic development. (Chaudhuri, 1978,95)

That is to say, non-aid sources of external finance, such as private foreign investment, other concessional finance, and commercial lending to government or public sector organizations, were relatively insignificant. The flow of private foreign investment (PFI) had always been small since Independence. This was understandable given the Indian Government’s policy towards PFI, which varied from deep suspicion to merely moderate suspicion. Private foreign investors were faced by the Indian Government with a large

India’s Aid Resources in Macroeconomic Context 27 and fluctuating apparatus of official controls, including the Foreign Exchange Regulation Act, over their activities, which most found discouraging. The result was a flow of PFI which was, in net terms, probably negative (Chaudhuri, 1978, 170–2) as pre-Independence foreign assets were gradually repatriated. At the same time, until the second oil price shock of 1979–80, the Government of India, unlike many other developing countries in the mid1970s, did not take on large quantities of commercial debt. This again seems to have been a conscious policy decision. As a result, India’s ratio of debt service payments (capital and interest) to export receipts fell through the 1970s, from 28 per cent (1971–2) to 12 per cent (1978–9) (Patel, 1986,59). By 1980, however, India’s position changed abruptly. The combination of the second oil shock with the drought-affected harvest of the previous year led India to seek a substantial IMF stand-by facility to help to finance a period of structural adjustment. The US$5.8bn from the IMF’s Extended Fund Facility, only a fraction of which was used, covered the period 1981–2 to 1984–5. A major purpose was to tide over a transition to lesser dependence on imported oil by means of greater exploitation of domestic sources of oil, natural gas, and coal. Given the deterioration of India’s terms of trade, volumes of agricultural exports had to grow rapidly as, in the longer run, did exports of manufactured goods. All of these changes needed to be accompanied by considerable improvements in India’s economic infrastructure (roads, ports, railways, and power supply) and changes in institutions and economic policies. Apart from the IMF facility designed to underwrite the costs of these structural adjustments, India had to tap commercial sources of borrowing much more extensively than it ever had to do in the past. India was relatively well placed for this. It was not, like many other developing countries, already enmeshed in the gathering debt crisis with the commercial banks. Indeed, as the net outflows of commercial bank funds from Latin America and sub-Saharan Africa in the 1980s shows, such banks were looking for safer places to put their funds—and recession-hit developed countries were unpromising, leaving India quite attractive. It had already secured the support of key agencies such as the IMF and the World Bank for its policy of structural adjustment. This meant that India was relatively credit-worthy in the eyes of the commercial banks. To some extent it was also well placed because the foreign trade sector is small in relation to the rest of the economy. Thus, a small fall in the proportion of resources servicing the domestic sector—being large relative to exports—can have a large impact on the resources gap and on the ability to service foreign debt incurred partially to cover that gap. However, precisely because the foreign trade sector is relatively small, a fairly low debt:GNP ratio can give rise to a high debt:exports ratio. The ability to prevent or rectify this depends on the ease of movements of resources from the domestic to the foreign trade sector. India began to tap the international credit market in a serious way in 1980–1. For the public sector, large-scale Eurocurrency financing was arranged for the National Aluminium Company’s projects (R544 crores) and for the Oil and Natural Gas Corporation (R160 crores). Private-sector companies were selectively permitted to sell bonds abroad through the International Finance Corporation to finance approved projects. Oil-exporting developing countries have been encouraged both to advance loans and to make equity investments. During 1981–2, according to the Economic Survey 1982–3, the total commitment of commercial borrowings was R 1,204 crores. This was equivalent to about US$1.5bn, which was in line with India’s undertaking to the IMF at the time of

28

Does Aid Work in India?

the Extended Fund Facility loan agreement. These signs of increased reliance on foreign commercial borrowing were not much publicized. But should India, with its good credit rating, resort increasingly to commercial borrowing from the international capital markets? Most shrewd judges are inclined to give a qualified ‘yes’ to this question, for example, Patel (1986, 59). The qualification is that caution has to be exercised in borrowing commercially because a high credit rating crumbles rapidly as soon as a developing country deviates from its planned trajectory of growth and debt repayment Problems can occur with ‘blips’ in repayment which can suddenly boost debt service ratios; if the ‘blip’ coincides with domestic problems, confidence may collapse. Aid is probably still necessary both as a safeguard and attractor of commercial loans, rather than as an alternative. The relative shift from aid to less concessional finance and commercial loans, and from longer-term to shorter-term loans, has profound implications for India which are considered below. Any likely future Indian level of foreign commercial borrowing will represent a balance between the forces creating the demand for such credit and the forces creating the supply. On the demand side, the major influences are: India’s demand for foreign exchange to buy imports, given assumptions about (i) the likely rate of growth of Indian GNP; (ii) the ratio of extra imports to extra GNP (at various price levels); (iii) exchange rates; (iv) import prices relative to prices in India; and (v) India’s import controls. The availability of foreign exchange from India’s exports and inward remittances, given assumptions about (i) the rest of the world’s demand schedule for India’s goods and services (including the services of Indian emigrant labour); (ii) India’s ability to supply them at various prices; (iii) exchange rates; (iv) India’s inflation relative to that of its customers abroad; and (v) their import controls. The projected difference between India’s foreign-exchange demands for imported goods and services, and its foreign-exchange availability from exports, gives the projected deficit on the current account of the balance of payments. This can be financed by, for example, a run-down of gold and foreign exchange reserves. India’s projected demand for aid, other concessional finance and foreign commercial credit will thus depend on (i) the currentaccount balance-of-payments deficit, and (ii) the difference between the existing and the desired level of gold and foreign-exchange reserves. The major influences on the supply schedule of commercial credit to India are the price and quantity available of competing non-commercial credit, i.e. various kinds of concessional flows—and the ‘confidence’ of commercial investors in India’s ability and willingness to meet her commercial obligations. Obviously the supply of concessional flows affects India’s demand for the commercial flows. In the case of concessional flows, aid as a source of finance has proved highly resistant to any significant expansion in the 1980s (save perhaps in once-for-all response to drought, as in 1987). But commercial credit can have a highly unpleasant ‘whiplash effect’, whereby yesterday one was expansively encouraged to borrow, today one hits a minor economic hazard and tomorrow one’s creditworthiness has vanished,

Trying to place precise numbers on the above determinants of India’s future level of commercial borrowing is extremely hazardous. Some years ago, a careful attempt was made by the World Bank (1981b). It examined various alternative blends of aid, other concessional finance, and non-concessional finance (the Bank’s own blends reviewed were the existing blend; 80 per cent IDA/20 per cent IBRD; an intermediate blend 50 per cent

India’s Aid Resources in Macroeconomic Context 29 IDA/50 per cent IBRD; and a hard blend 25 per cent IDA/75 per cent IBRD). These were juxtaposed with alternative assumptions about India’s future size of resources gap (i.e. the annual deficits on the current balance of payments). Each case would imply various levels of the balance-of-payments deficit and of the debt service ratio. The main or central projection was made on the assumptions that the overall growth rate would be 4 to 5 per cent per annum (i.e. somewhat above its long-term historical rate of 3 per cent p.a); that import volumes would grow at about the same rate as GNP, and export volumes at about their 1970s rate of 7 per cent p.a.; and that foreign exchange reserves were required to cover 3 months’ worth of imports. On these assumptions—and given plausible changes in prices—India’s current-account deficit on the balance of payments could be kept down to ‘manageable’ proportions, defined as being less than 2 per cent of GNP. It was then judged likely (World Bank, 1981b) that India would be regarded by foreign lenders as credit-worthy for the absolutely large sums that would be required to cover about half this deficit by the end of the 1980s. In other words, by the World Bank’s calculations, India’s new commercial loans abroad would then be as large, each year, as those from IDA and IBRD taken together. Whether a switch towards commercial borrowing of this magnitude really is feasible— i.e. whether India can reduce its reliance on aid in line with falling relative disbursements and maintain growth—depends on whether the above assumptions are realistic. The most doubtful may prove to be an export volume growth rate of at least 7 per cent a year through the 1980s. That rate was achieved in 1970–1 to 1978–9, but even by 1981 it was clear that this fact should not breed excessive optimism. The world trading environment suffered a general recession in the early 1980s and an increase in specific measures of protection against Third World imports; the halting recovery of the western economies in 1985–7 did little to improve these matters; and the liquidity crises of late 1987 posed renewed threats. Moreover, India’s export performance in the latter half of the 1970s reflected its relatively low inflation vis-à-vis its developing-country competitors and the developed countries. Its exports suffered in the first half of the 1980s when global inflation rates, especially in developed countries, fell sharply while India’s was stable or slightly higher (see p.9). Much depends on India’s ability to continue a rapid change in the structure of its exports away from staple primary exports and towards manufactured goods. There have been some hopeful early signs of this in chemicals and engineering, but the growth of iron and steel exports has been disappointing. If the required dynamism in exports cannot be achieved, the balance of payments deficit can in principle be contained by reducing the overall growth rate and thus the growth rate of imports. But since the improvement of infrastructure itself requires imports, this would inevitably lengthen the time during which structural adjustment takes place and that, in turn, may have a deleterious effect on foreign lenders’ confidence. In brief, the less well India’s exports perform, the greater the deficit, and hence the need to supplement concessional with commercial inflows; but the less the readiness of commercial lenders to supply such inflows, and hence the higher the interestrate supplements that India will have to pay. These supplements, in turn, would burden the future balance of payments, impeding both creditworthiness and growth in the long-term. Since the World Bank’s 1981 projections mentioned above were made, the other key supply side factor—the a vailability and terms of aid and other concessional finance—has changed for the worse. The 1986–7 decisions to replenish ID A less generously than had

30

Does Aid Work in India?

been sought by the World Bank—despite the Bank’s prior concurrence with the largest donor to IDA (the USA) that India’s share in IDA lending had to fall sharply—must raise the volume of required commercial borrowing by India. There are thus reasons for suggesting that the Bank’s projections may have been somewhat optimistic. Much thus rests on the amounts and terms offered by bilateral aid donors and by the commercial banks themselves and by suppliers of non-aid (but still slightly concessional) finance, mainly the IBRD; and by commercial banks.22 IBRD money seems unlikely to compensate for the loss of IDA money at an appropriate substitution ratio (see p.36). Nor is it yet clear whether bilateral aid donors will take a lead from the Bank in hardening their terms to India. Such uncertainties compound those affecting India’s ability to make her exports grow strongly. India herself is reluctant to go so rapidly as envisaged by the Bank in the direction of private commercial borrowing. The considerations here are more political than economic: the government believes that it will encounter less political pressure by relying on multilateral lending. It fears to accumulate a heavy bank debt, because it fears that in the event of, say, two successive bad harvests, it would be seriously exposed to political pressures on its economic policy as a price for receiving a financial package to bail it out Drought still arrives in India unpredictably, but seldom less than once a decade. The failure of the 1987 monsoon means that a second successive year of drought in 1988 would place the country in a most vulnerable position, that could be used, as it was after 1965–6, to try and force the government’s hand on the nature and speed of economic liberalization. What has been painfully learned through the aid policy dialogue (see Chapter 3) might not be heeded when worried commercial bankers get together with their Foreign Offices. As indicated in the earlier sections, Measurement of aid flows (pp. 4–6) and Size of aid flows to India (absolute flows) (pp. 6–14), no single statistic can summarize the value of aid, or of total concessional finance, relative to non-concessional finance. Thus there can be no single numerical answer to the question: ‘What volume of non-concessional finance would balance, in its value to the Indian economy, a given volume of concessional finance?’ The primary measure of the value of aid relative to non-concessional finance is the size of the grant element (see note 1), but the costs of tying must also be taken into account. Apart from tying, loans have other hidden costs which affect their value to the recipients: most notably, the costs of negotiating the loan, of fees, of meeting any special requirements of the lender, and of delay, i.e. of reduction in the present value of returns to an investment because it has to be held up until the loan is finally authorized. Just as the costs of tying vary as between different types of aid, so these costs of negotiation and delay vary between other types of finance. Eurocurrency loans are usually less costly, in this respect, than IBRD loans (Harvey, 1983, 25), though of course the formal interest rate is higher. Finally one has to note that the rate of interest payable on loans from the Eurocurrency markets, since 1985 from the IBRD, and frequently from private banks, is not fixed. Thus one does not know ex ante what such a loan will cost over its term, although the initial rate of interest is, of course, known. This is an element of uncertainty, rather than a calculable risk. The future course of interest rates is extremely difficult to predict, and an over-optimistic prediction leads to short-term debt service problems.23 When these complications do not apply, the problem becomes more tractable. When we are comparing two different types of loan which have a common term and whose interest

India’s Aid Resources in Macroeconomic Context 31 rates, though different, are fixed over the life of the loan, the size of the grant element in the loan can be related to the ‘substitution ratio’ of the loans. ‘Substitution ratio’ here means that number by which the volume of the concessional loans must be multiplied to give the volume of non-concessional finance which is of equivalent value. As the grant element ranges from zero to 100 per cent, so the substitution ratio varies from one to infinity.24 In the Indian case, the substitution ratio lies somewhere between two and three, if we are thinking of the substitution of loans on ‘hard terms’ loans from IBRD for loans on ‘soft terms’ from IDA. A switch towards commercial borrowing by India raises several problems (Anagol, 1987). The commercial terms imply higher borrowing for the same level of effective resources at higher interest rates, raising both the debt and debt service ratios. Further the loan terms are usually shorter than with various forms of concessional finance (with IDA terms longer than IBRD ones), again raising the debt service ratio. Not only are interest rates variable, but so far in the 1980s they have been very high in real terms (by historical standards). The level and speed of returns to projects must therefore be assessed carefully in determining what level of commercial borrowing India can safely undertake. One estimate (Anagol, 1987, 31) suggests a level of US$1–2bn per annum, roughly comparable to current gross aid utilizations. This is close to the World Bank’s conclusion in 1981 (see p.34). Of course, the export growth rate may itself partly depend on the level of finance—and may substantially affect the readiness of commercial lenders to provide new cash (p. 34). More past lending and disappointing current exports both would imply rising debt service ratios alongside falling credit ratings. The dangers of this ‘Latin American scenario’, especially if the economy deviates from its growth path, are such that it would be reckless to project expansion of commercial loans, in the late 1980s and early 1990s, on a scale sufficient to replace aid to India; they can, however, be an increasingly valuable complement to such aid. Indeed, India’s requirement for concessional finance—to safeguard and ‘soften’ the total debt position created by the growing IBRD and commercial elements—has for several years been stressed by the World Bank. David Hopper, then Vice-President for South Asia, told the 1984 All-India Consortium meeting: As the composition of capital flows to India hardens, the risk of severe deterioration in the external economic environment and the impact that this could have on the debt service burden are and will remain sources of deep concern to the Government of India…. [T]o minimise these risks, India needs our more active participation through the provision of more aid on the best possible terms. (Badhwar, 1984, 22)

As will be shown in Chapter 7, this macroeconomic judgement is reinforced by microeconomic considerations. Resource allocation throughout the economy is still distorted by India’s pervasive ‘control syndrome’. India has begun to tackle this, but such policy reforms create some losers. Some of these people—particularly if they are poor—need to be cushioned by the continued availability (if required) of aid. Otherwise the reform process will prove politically unsustainable and will be reversed, again damaging exports and eroding India’s credit-worthiness. This is yet another way in which access to aid is a complement, not an alternative, to commercial borrowing by India.

32

Does Aid Work in India? Conclusions

Our conclusions on the scale, role, and impact of the aid inflow to India over the last thirty years may be summarized as follows. Scale 1. Although India’s aid receipts are the third largest in the world, she receives 11.6 per cent of the global net disbursements of aid to low-income countries (yet contains 31.4 per cent of their population), and only 5.7 per cent of total disbursements to all developing countries (for 20.8 per cent of population) (World Bank, 1987, 202–3, 244–5).26 2. In current rupees, India by the mid-1980s had utilized nearly R150bn of foreign aid, net of debt service and repayments. The peak occurred in 1961–6 (during the Third Five Year Plan), with smaller peaks in 1975–6 and around 1980 (Table 1.1). 3. Figures in current US dollars for the period since 1971 (Table 1.6) indicate that aid net of repayments peaked in 1974–6 and again around 1980. 4. None of these figures reveals the purchasing power of these concessional flows. That would require (i) an estimate of the inflation of prices of goods purchased with aid; and (ii) the level of aid-tying, and changes in that level over time. Allowing for the inflation factor only, we confirm that aid flows under the Third Plan represented the greatest annual transfer of purchasing power; the minor peaks also survive; but the 1984 and 1985 purchasing power of India’s net aid disbursements appear to be barely half the levels of 1974–6 (Table 1.6), for a recipient population over 20 per cent higher. Role 5. Figures for the central government capital account items in the balance of payments indicate the maximum contribution which aid could have made to closing the currentaccount gap. This gap reached nearly 2 per cent of GNP in 1961–5—mostly covered by net aid. From 1975–6 until 1980, aid flows (although reduced) were greater than required to cover the much smaller current-account deficit, and permitted the building up of foreignexchange reserves to levels unprecedented for India since 1958. Since 1980, aid flows have been insufficient to cover the greatly increased current-account deficit, and have had to be supplemented by non-concessional flows and a rundown of reserves. 6. Figures of foreign savings show the maximum contribution of aid flows to the deficit of Indian investment over Indian saving. Between 1951 and 1975, there were few other foreign inflows; aid—roughly speaking, foreign savings—amounted to one-quarter of the deficits of the government and the private corporate sector. Aid then played an important role in permitting public and corporate investment to be undertaken well ahead of the development of the household sector’s propensity to save. Since 1980, however, non-aid foreign savings have dramatically increased their contribution to India’s total savings. The average contribution of total foreign savings in the 1980s has been around 7 per cent, although the aid element in foreign savings has sharply declined: aid’s average contribution to India’s total saving in the early 1980s was less than 4 per cent, half the proportion of the mid-1970s (Table 1.12). Aid’s contribution to government investment fell from over 50 per cent in the late 1960s to just above 10 per cent in the mid-1980s (though its contribution to ‘aidable’ public investment is probably around a fifth). Meanwhile, commercial loans and official non-aid finance have played an ever-increasing role.

India’s Aid Resources in Macroeconomic Context 33 Impact 7. India’s economic growth had a good start in the 1950s; perhaps slowed a little from the early 1960s to the mid-1970s; and achieved a more dynamic tempo—through still not as rapid as in some other Asian economies (including China)—in the 1980s. Agriculture is still the dominant cause of fluctuations in growth; its share in GNP, i.e. its statistical dominance of the overall growth rate, is being increasingly eroded, but not so its demand and supply roles in constraining that rate. 8. The attempts to find the link between aid and economic performance by econometric methods have been diverse and have produced conflicting results. Several look for the impact of aid on aggregate investment without considering the objectives of Indian policy-makers, without disaggregating investment (or aid) by type and sector, and without focusing on the indirect effects of aid on productivity via extra food consumption. Food aid to India is often ignored, although in the 1960s and 1970s it was probably helpful in relaxing the foreign exchange constraint. As a whole, aid almost certainly raised India’s ratio of total savings (domestic and foreign) to GNP—i.e. aid did little to counter the steady rise in India’s domestic savings ratio (and much to add foreign savings). Alongside this rise, India suffered an almost offsetting fall in the utilization and/or productivity of capital (so its growth rate shows little or no secular rise). The concentration of aid in the 1960s on steel, heavy electricals, etc. certainly contributed to this, but the less capital-intensive and more rural emphases of aid in 1973–81 may have reversed the trend. However, it may well be that aid now being channelled to new industrial investment will be less effective because of the pervasive inefficiencies which some have detected in Indian public sector industry (e.g. Weiner, 1986,609: cf. Chakravarty, 1987). References Ahluwalia, I.J. (1985) Industrial Growth in India: Stagnation since the Mid-Sixties, Delhi, Oxford University Press. Anagol, M. (1987) ‘Implications of commercial borrowing for India’, International Journal of Development Banking, 5:1, January. Asian Development Bank (1977) Rural Asia: Challenge and Opportunity, New York and London, Praeger Publishers. Badhwar, I. (1984) ‘Letter from Washington’, India Today, 15 August. Bhagwati, J.N. (1970) ‘The tying of aid’, in J.N.Bhagwati and R.Eckaus (eds), Foreign Aid, Harmondsworth, Penguin. Chakravarthy, S. (1987) Development Planning: the Indian Experience, Oxford, Clarendon Press. Chaudhuri, P.K. (1978) The Indian Economy: Poverty and Development, London, Crosby Lockwood Staples. Goldsmith, R.W. (1983) The Financial Development of India 1860–1977, London and New Haven, Yale University Press. Government of India (1983,1986) Economic Survey 1982–3 and 1985–6, New Delhi, Government of India Press. Hamilton, C. (1983), ‘Capitalist industrialization in East Asia’s Four Little Tigers’, Journal of Contemporary Asia, 13:1. Harvey, C. (1983) Analysis of Project Finance in Developing Countries, London, Heinemann Educational Books. Heston, A. (1982) ‘National Income’, in D.Kumar (ed.) Cambridge Economic History of India, vol.2.

34

Does Aid Work in India?

Jakhade, V. (ed.) (1983) Special Issue of the Indian Journal of Agricultural Economics, 38:4, December. Joshi, V. and Little, I.M.D. (1986) ‘Indian macroeconomic policies’, Oxford (mimeo). Learner, H., (1982) ‘Let’s take the con out of econometrics’, American Economic Review. Lipton, M. (1972) ‘Aid allocation where aid is inadequate’, in T.Byres (ed.) Foreign Resources and Economic Development, London, Frank Cass. Lipton, M. with Longhurst, R. (1985) Modern Varieties, International Agricultural Research and the Poor, Study Paper 2, Consultative Group on International Agricultural Research, World Bank, Washington DC. Mosley, P. (1987) Overseas Aid: its Defence and Reform, Brighton, Wheatsheaf Books. OECD (1980 et seq.) Geographical Distribution of Financial Flows to Developing Countries, Paris. OECD (1986) Development Co-operation, Paris. Patel, I.G. (1986) Essays in Economic Policy and Economic Growth, Basingstoke, Macmillan. Riddell, R, (1987) Foreign Aid Reconsidered, Baltimore, Johns Hopkins University Press. Rubin, B.R. (1982) Private Power and Public Investment in India, University of Chicago, unpublished PhD thesis. Sen, A.K. (1983) ‘Development—which way now?’, Economic Journal, 93: 372, December. Weiner, M. (1986) ‘The Political Economy of Industrial Growth in India’ World Politics 38:4, July. World Bank (1980,1985 and 1987) World Development Report, New York, Oxford University Press. World Bank (1981a) Accelerated Development in sub-Saharan Africa, Washington DC. World Bank (1981b) ‘India’s Balance of Payments Prospects: Resource Requirements and Credit worthiness’, unpublished memorandum. World Bank (1984) Situation and Prospects of the indian Economy—A Medium Term Perspective, Report 4962-IN (3 vols) for official use only. World Bank (1986) India: Economic Situation and Development Prospects, Report 6690-IN (2 vols).

Chapter 2 Aid and Poverty in India

Dilemmas in using policy against poverty Problems and performance In assessing the effects of aid on India’s performance in reducing poverty, we face several problems. (i) The performance has not been very good. Between the late 1950s and the mid1980s, the absolute number of Indians below a (fixed and harsh) poverty line has increased sharply. There has been at best a very small decline—masked by major fluctuations—in the proportion of the (growing) population below this line. There has been little or no easing in the intensity of their poverty. This is despite growth, rapid by historical standards, in GNP per person. (ii) Aid to India was not obviously relevant. Until the late 1960s, it was concentrated on activities—or on types of aid flow—not very likely to affect poverty directly. Since then, aid has been small relative to Indian investment. Aid is therefore not likely to have had much direct impact on poverty in India. (iii) However, it would be mistaken to ‘blame’ aid for India’s unsatisfactory performance in poverty alleviation. The smallness of the aid/investment ratio, and the nature of India’s planning procedure, ensure that many aid-financed activities would have happened anyway. (Even if all would, this does not, as will be shown, mean that aid makes no difference). But for aid, poverty might have got worse. The most important example is that, beyond reasonable doubt, Indian poverty would have been much worse without the spread of the high-yielding cereal varieties. This is indirectly linked to aid, both for Indian agriculture in particular and for tropical agricultural research more generally. (iv) Owing to (i) and (ii) above, assessment of the aid-poverty nexus must concentrate on specifics; there is little point in testing aggregate relationships ‘linking’ small or possibly irrelevant aid to very slow progress in reducing poverty. With the labour force growing at over 2 per cent annually in 1951–81—so that its bargaining power becomes weaker as more workers compete for jobs—such progress, in a deeply unequal society, is likely to be slow. Yet in India a wide range of activities and organizational forms—public and voluntary, central and local—is nominally directed at the reduction of poverty. Despite the numerous evaluation efforts, little is really known about which efforts have succeeded, let alone which can be supported by aid. Such support could make a real difference, either if aid added to the volume of government (or Indian private) anti-poverty outlays, rather than merely releasing them for other uses; or if aid accompanied technical inflows, or a joint learning process between Indian and donor agencies, that increased the efficiency of anti-poverty activity, and/or the extent to which the benefits of DOI: 10.4324/9780203840153-3

36

Does Aid Work in India?

such activity reached the poorest. Conversely, inappropriate aid could damage some poor groups or places. There are three main ways for aid to a project or programme to benefit poor people. First, aid may benefit them as producers, by helping to provide the poor with land, credit, capital, or jobs. Second, aid may—by increasing the volume, accessibility, or quality of poor people’s health or education—raise their productive performance (World Bank, 1980; Jamison and Lau, 1982; Chaudhri, 1979) or their general well-being. Third, aid may benefit the poor as consumers, either by raising their direct access to consumption benefits from public goods, or by lowering the price (or raising the secure availability) of what they buy—mainly, in India, cheap food: over 70 per cent of spending by the poorer half of India’s households is on food, and over 50 per cent on cereals and pulses alone. Overwhelmingly the main contribution of aid to poverty reduction in India has been in the third area: through the indirect contribution of aid-supported agricultural research, in conjunction with aid-supported farm inputs (especially irrigation and extension), in helping to increase the availability of low-cost calories. However this contribution has been largely offset by two problems, one of policy and one of politics, which aid can to some extent help to resolve. The policy problem arises if, as food production per person rises, real incomeper-head of poor people—largely determining their ‘exchange entitlements’ to food (Sen, 1981)—does not. Since only poor people spend a lot of extra income on food, where is the extra demand to come from? Initially, the Indian Government used extra ‘green revolution’ food production to replace imports, a process from which poor food consumers gained little. Since the mid-1970s, extra wheat output has been used to build up public stocks to unprecedented levels of 20–30 million tons. Up to a point (say 10–12 million tons) such stocks are needed to increase the safety of the poor. However, the expense of tying up savings in grain stocks, at the cost of devoting them to real investment, is large. Another option—followed for rice and sorghum—is simply to let the extra ‘green revolution’ output restrain the price of grains. Poor consumers benefit; but many of the very poorest depend for income on landless labour to produce these food items. If prices of these grains fall, their production (and hence employment) are discouraged. The apparent solution to the policy problem is to raise the purchasing, as well as the producing, power of the poor, who are net food buyers. But that brings us to the political problem of poverty alleviation. Discussion of this, and of aid’s role in alleviating it, is deferred to p. 47, after our review of what has happened to poverty in India. Trends in poverty in India: a political dilemma? The detailed analysis of poverty in India is fraught with controversies. These have led people to lose sight of the four key facts described below, which are hardly in dispute. Large and detailed surveys, notably the Indian National Sample Survey, provide pretty reliable results about the key trends. First, the proportion of Indians in poverty, and the average severity of that poverty, have not changed much since Independence. On nationwide data the proportion below a fairly rigorously defined poverty line—while fluctuating—can be shown from many data sets not to have changed systematically, probably since the mid-1950s, almost certainly since the early 1960s (Kumar, 1974,35; on the rural data summaries, see Ahluwalia, 1978, 302;

Aid and Poverty in India 37 1986; and Saith, 1981, 197–201). The rural incidence fluctuates between 40 and 55 per cent, mainly depending on whether food is cheap, and on whether farm output is high enough to provide ample work (Ahluwalia, 1986; Minhas et al., 1987). Of the Indian poor, about onethird (15–20 per cent of Indians) are at serious risk of eating too few calories (Sukhatme, 1978; Lipton, 1983). Since the ultra-poor families contain a much larger proportion of children, the proportion of under-fives affected by nutritionally dangerous poverty is even higher than that of adults. In the poorest one-fifth of Indian households there are twice as many. children per adult as in the best-off one-fifth, and their risk of death is probably about double. About one-third of these infant deaths are due to the interactions of hunger and infection (Mitra, 1978; Lipton, 1983; and 1983a, 43–5). The life of a typical poor Indian has certainly improved—and lengthened—since Independence, even though his or her real private consumption has not increased in an average year. Better public health, together with public works and food distribution to anticipate famine, have brought age-specific death rates steadily down. Nevertheless, the proportion of Indians unable to afford the basic minimum package of privately purchased consumer goods—to buy enough to eat, even if they spend only the meagre 20–30 per cent of outlay on non-foods typical of their income and household composition—was not much different in 1986 from 1960, when average real income in India was 50 per cent lower. All this relates to incidence of poverty—proportions of Indians below an inevitably arbitrary poverty line. Recently much work has been done on intensity of poverty—the gap between the poverty line and the outlay, income, or calorie intake of the average poor person—and on the distribution of income among poor persons (Sen, 1981). Since about 1960, intensity and distribution of poverty show no clear trend. This, therefore, must be said of the ‘severity of poverty’ also—since severity is the resultant of intensity, incidence, and distribution among the poor. One major bright spot is that extreme fluctuations in poverty have been reduced. This is probably not the result of reduced variability in foodgrain output; modern varieties of cereals are nowadays usually safer than the varieties they displace, but they do tend to ‘concentrate’ production in areas subject to similar weather shocks, and population growth has pushed grain production out towards riskier lands. But, in the 1960s and early 1970s, grain stocks supported by food aid reduced the damage done by bad seasons and years. Since the mid-1970s, huge (and costly) grain stocks were increasingly accumulated without net imports, as the supply of food from the new cereal varieties outpaced demand from the near-stagnant incomes of the poor; such stocks were ready to be mobilized for times and places of food crisis. Rural foodgrain distribution via fair-price grain shops has lagged behind urban, but state schemes have lately offset this, by providing public employment or school food as an emergency reserve. Thus since 1943 India, unlike China, has avoided massive famine deaths due to the sharpening of already severe poverty in below-average years (Sen, 1983). A consequence is that the linkage of extreme poverty, as such, to very high death-rates has been weakened. This is partly due to aid, both to build food reserves in the 1960s, and to develop agriculture and irrigation in the 1970s; but the vast bulk of the efforts have been Indian. Nevertheless, poor people’s purchasing power in normal times shows no clear uptrend since Independence. A second undisputed fact is that this was accompanied by steady growth in real GNP per person. Although slow, and although interrupted by occasional

38

Does Aid Work in India?

bad harvests, this growth has been about 1.5 per cent per head per year ever since the early 1950s. Since the 1973 oil crisis, and especially since 1980, the performance has improved somewhat. So India has experienced just over 80 per cent growth of real GDP per person since Independence (1.5 per cent compound for forty years); probably, no significant improvement in the average annual real private income (and hence nutrition) of the ‘poorer half’; and major improvement in its access to publicly provided services, many of them–above all health, and access to emergency famine prevention–improving the efficiency with which nutrition is converted into welfare, especially in bad times. The third undisputed fact is that India’s poverty-reducing performance shows marked variation among different parts of India (Ahluwalia, 1978, 305–6, 314; Rao, 1985). Trends in real wage-rates suggest significant improvements for the rural and urban poor in the ‘green revolution’heartland states, Punjab and Haryana (Bhalla, 1979). In some other states, especially Kerala, overall policies–on land reform, labour legislation, health and education– kept the levels of poverty and under-nutrition much lower than state income-per-person would suggest; elsewhere, rural employment guarantees (Maharashtra) or free school meals (Tamil Nadu) had some impact. Conversely, the parts of India untouched by either major agricultural innovation or systematic redistribution–most of South Bihar, Orissa, Madhya Pradesh–must have experienced major worsening of poverty during 1960–88, since other areas improved yet the national average did not. Even there, fluctuations were softened; there was not a massive Bihar famine in 1967. But high infant and child mortality–and lethargy due to adult poverty and hunger–plague much of Bihar every year. The fourth fact is that this bad performance in reducing poverty has persisted despite a rather favourable policy environment. Huge sums have been spent on poverty programmes. If a large part of the money has gone to the non-poor, or to administration, or to purely temporary respite for the poor, the question is why this was not prevented by democratic procedures and pressures. These, alongside growing literacy, should have led to much more, and more effective, political organization of poor people than has been witnessed. As for governmental attitudes, the First Plan (1951–2 to 1956–7) was the first major Third World planning document explicitly to stress employment and poverty-reduction even if there were a cost in foregone growth, The slogans of garibi hatao (‘banish poverty’), from all political quarters, have been backed with overt, costly actions through a whole series of official programmes. In the Seventh Five Year Plan for 1985–90 (vol. 1, 1985, pp 27–9), of the R1,800bn of allocated public outlays, R35bn are for the Integrated Rural Development Programme (IRDP, distributing non-farm assets to the rural poor); R47bn for rural employment schemes; R31bn for ‘special area programmes’ in impoverished regions; R25bn for welfare of scheduled castes and tribes; and R17bn for nutrition programmes–apart from major poverty-directed components in the plans for housing and urban development (R43bn), family welfare (R33bn), and water and sanitation (R65bn). Each such programme generates a career-stracture that rewards success in reaching the poor. In the past, each has registered some real achievements, and some failures. Most programmes have passed on, absorbed in a new programme or fashion. Why has the poverty outcome not been better? Two main sorts of explanations are offered, with different policy implications. India’s political parties have been generally unsuccessful in organising the rural poor; for example, fewer than 5 per cent of all rural Indians are members of any political organization or

Aid and Poverty in India 39 peasant movement, even on very generous definitions (Alexander, 1981). One explanation claims that India’s social structure creates a ‘dilemma of politics’ for any group that undertakes actions to advance the poor. Before explaining this dilemma, we look at an alternative demographic explanation. It is India’s poorest couples who have the largest, fastest-growing families, with the highest proportion of children too young to work. Each such couple finds that it pays to produce many children, as a low-cost source of future income (Becker, 1981, ch. 5). However, when these children do join the labour force, their increasing numbers mean that they compete ever more fiercely for scarce jobs, or for scarce land to work. Yet among the owners of capital, and of farmland (for hire or for working), population grows more slowly (Lipton, 1983a). Employers–owners of land and capital–thus become fewer and more concentrated, relative to poor workers in the labour market. Labour becomes more plentiful, and hence weaker, relative to employers. Yet the poor rely increasingly on labour income, not owned land; and the real wage and the volume of available work per person are both depressed by demographic pressures. Anti-poverty schemes (and ‘green revolutions’) can at best offset this baleful trend. Thus, for example, in its first six years of life IRDP, according to its severest and most articulate critic (Rath, 1985), pulled about 3 per cent of India’s rural poor above the poverty line—yet the proportion of India’s rural people in poverty by 1983 did not appear to be significantly lower than in 1960. Does this alternative, non-political explanation account for the persistence of poverty in India alongside growth? The policy responses of the Government of India and donors do not suggest they believe so. The short-run policy suggested by the demographic explanation involves concentration on labour-intensive activities, so as to strengthen the employment position of the growing numbers of poor. Despite intermittent efforts, however, aid projects, especially outside agriculture, have generally been less labour-intensive than other projects. The Indian Government, whose procedures allow it considerable choice among projects for limited aid support, has increasingly preferred aid to go to projects that are heavy on technology and capital, where it sees donors as having most to contribute. In the long run the demographic explanation also suggests efforts in family planning. India’s First Plan carried one of the developing world’s earliest commitments to it, and there has been a major learning process subsequently (Cassen, 1978, 146). Massive, well-funded, and partly successful GoI efforts have received some aid support. For every thousand Indians, there were 45 births in 1963, but ‘only’ 33 in 1985. In 1970, one in eight married couples used contraception; by 1984 it was one in three (World Bank, 1987, 256). However, family planning efforts cannot succeed unless they are overwhelmingly Indian, and for political reasons, only small parts of family planning programmes in India can be supported by aid. Nor is it automatic that family planning affects mainly the poor rather than the accessible urban non-poor. The demographic ‘explanation’, anyway, cannot explain much of the growth-withpoverty paradox. Real wage-rate averages across India—being close to subsistence—have no down-trend. In any event, there is no obvious relationship, across Indian regions, between population pressure and failure to reduce poverty. (If anything, the evidence (Repetto, 1979) suggests that, as in Kerala, better distribution of the gains from growth brings down fertility.) So one turns to another explanation: to what, we have suggested, is a political dilemma. This dilemma is as follows. A ‘pro-poor’ agency of the state may either enlist or exclude powerful groups from its anti-poverty efforts. Enlistment risks subversion; exclusion risks blocking.

40

Does Aid Work in India?

It is regularly observed that the better-off are able to obtain benefits of schemes labelled ‘For the Poor’: subsidized credit and fertilizers; ‘low-cost’urban housing; food rations; even about one-third of the small-scale assets distributed in IRDP (Rath, 1985). The dilemma of politics, that has severely limited India’s poverty-reducing efforts, is this. Assume that a disinterested, or career-orientated, part of the state genuinely seeks to implement a law or policy aimed at helping some weak, disorganized, inarticulate people. Such actions are likely to harm some strong groupings. Should the ‘pro-poor’ part of the state, then, work with or against such groups? If enlisted to support its policy against poverty, they will seek to subvert that policy—either by emasculation (e.g. by writing loopholes into land reform legislation) or by capturing the benefits (e.g. by obtaining the credit allegedly subsidized ‘to help the poor’). If excluded from anti-poverty policy, the privileged (in the absence of powerful and articulate pressure from below) can often block the anti-poverty policy— e.g. by using the courts against genuine land reform; or by insisting on central actions to impede radical reforms by state governments. Rath (1985) recognizes an economic counterpart to this dilemma; he seeks to escape it by arguing that the better-off should be included in programmes to develop a poor region as a whole, and that ‘assets for the poor’ are foredoomed without overall market development—an argument also implicit in the ‘linkages’ approach (Mellor, 1976; Hazell and Roell, 1983). But the central dilemma is political: economic ‘solutions’ that incorporate the rich may achieve little for the poor if they do not first gain some power. Of course the state—even a more or less independent anti-poverty agency of the state—is not simply a neutral, passive agent If its officials ally themselves with the powerful (or the poor), the problem of subversion (or blocking) of pro-poor actions becomes even worse. Such an account could well be accepted both by advocates of free-market styles of development, and by centralizers. The former see the state as the main source of privilege; the latter indict private power and ownership; but both blame the interaction of state rentcreating activity (including poverty programmes) with unequal power-balances (the rich being strong in seeking rents, the poor inarticulate or weak in doing so) for the imperfect performance of India’s anti-poverty programmes despite growth. Nevertheless, aid donors must conduct a ‘policy dialogue’ with those in power: officials of the Indian state. These, in turn, can move the pro-poor inputs—cash, food, or whatever—only through channels substantially influenced by existing power structures. Aid may ultimately reach semiindependent (e.g. parastatal) public agencies, or voluntary organizations, or the private sector; but the state is the first recipient. However, this is far too gloomy a view, for three reasons. First, a genuinely federal, multi-ministerial democracy like India has numerous and varied power-holding agencies. Many important ones include direct incentives to get benefits to poor constituents. Aid donors can, to some extent, steer resources to agencies of this type. Second, many sorts of producer and consumer assets are relatively likely to be used by the poor. Wealthy people will seldom want to eat broken rice or sorghum, for example. The better-off already take for granted such facilities as primary schooling or basic health care; providing more of them, especially with a subsidy, is likely to reach mainly the poor. Even such facilities as site-and-service housing or small-scale rural credit—in principle, readily diverted to the non-poor, or blocked by competing private suppliers—are in practice ‘aidable’ in most societies in ways that reach mainly the poor. Hence it is not credible—even

Aid and Poverty in India 41 in India, where aid is a small and tightly-programmed part of public expenditure—that aid is wholly fungible, and can do nothing to help the parts of the Indian State who wish to increase the share of its outlays (however financed) that reaches the poor. Third, the vigour and openness of debate in India overstates the failures of each antipoverty programme in turn. All pro-poor programmes, given the ‘dilemma’ and the demography, have limits and leakages, but also many beneficiaries. Aid can support some of these programmes. Indeed—though the main determinants of progress out of poverty are internal in almost every country—if aid cannot provide some net gains to the poor in India’s relatively open and pluralist society, it has little chance of doing so any where. Scale and structure of aid: impact on poverty Smallness of aid The evidence of Chapter 1 might suggest that aid to India is, relative to population and GNP, just too small to affect poverty. At such low levels, extra aid for a specific purpose, e.g. a new anti-poverty project, does seem likely in part to be used for the relief of GoI domestic funds. However, gross project aid is about 15 to 20 per cent of public sector project investment; and donors retain considerable leverage through selection of which agencies or locations to help (see Chapter 5). However, declining real aid ever since the late 1960s probably does mean declining capacity by donors to support (or influence) Indian policies against poverty. The big cuts in food aid in the 1970s did not specially harm poor food consumers—it was low income, not low food availability, that denied them claims over India’s food stockpiles. However, donor support for Indian anti-poverty actions declined because dwindling food aid was not replaced by other forms of aid. When, as in India, aid in real terms is small and declining, both the difficulty and the importance of focusing it on poverty are likely to increase. Recency of sectoral anti-poverty emphasis The sectoral composition of Indian aid receipts has substantially shifted towards activities more likely to reduce poverty. Unfortunately, however, this improvement took place during the very period—from about 1971–2—when downward pressures on total Indian net aid were most severe. Until the end of the Third Plan (1965–6), aid to industrial development alone, excluding steel aid, comprised 59.8 per cent of authorized and sectorally allocable aid (i.e. excluding the 2 to 3 per cent of ‘debt relief and miscellaneous’), and 57.5 per cent of aid utilized. In 1966–74, the respective proportions were 65.1 per cent and 68.4 per cent; by the period 1974–5 to 1978–9 they had fallen to 29.5 per cent and 40.8 per cent respectively, and by the period 1980–1 to 1984–5 to 28.2 per cent and 31.6 per cent (RBI, 1984–5, statement III). ‘Industrial development’ means overwhelmingly–especially for aid donors, who have been unable to steer much support to small-scale or rural industry–a form of investment that, whatever its virtues, produces rather few jobs, and rather low values of mass consumption goods, per rupee of investment.1 The same applies to aid for transport and communications, power and iron and steel. With other industrial development, these comprised 96.2 per cent of sectorally allocable aid authorization (and 98.4 per cent of utilizations) in the first

42

Does Aid Work in India?

three Plans, i.e. prior to 1965–6. In 1966–74 the proportions were 73.7 per cent of aid authorized (81.3 per cent of aid utilized); in the period 1974–5 to 1978–9, 58.6 per cent (66.6 per cent); and in 1980–1 to 1984–5, 52.7 per cent (61.7 per cent). Authorizations to the ‘heavy’ sectors turned up sharply after 1982–3, and for the three most recent years with available data stood at well over 73 per cent of all aid authorized (RBI, 1984–5, statement III). Although poverty impact monitoring in the World Bank group is beginning to cover these ‘heavy’ sectors more fully, they remain relatively little poverty-focused. Donors’ involvement with poverty alleviation in India is likely to have been much more, at project level, in 1975–82 than before or since, on these data. The ‘non-heavy’ forms of ‘sectoral aid’—those with the greatest potential to benefit poor people–are probably food aid and aid to agriculture. (Other areas where aid has had a significant poverty impact are examined on pp. 71–3.) Food aid Food aid was barely 3 per cent of total aid during the first three Plans. In 1966–74 it comprised almost 9 per cent—still mainly wheat. Its importance dwindled with the US withdrawal from the Indian aid scene (1972–8), and with the attainment by India of virtual food self-sufficiency, albeit at low levels of nutrition. In recent years EC skimmed-milk powder has been given to the GoI for sale to provide counterpart funds in support of dairy development through ‘Operation Flood’ (RBI, 1984–5, statement III). Has all this helped the poor? (i) There is no doubt that the short-run effects of cereal aid, in years when India combined bad harvests with foreign-exchange crisis (notably 1965–6 and 1966–7), saved the lives of many thousands of poor people. Nor is there much doubt that, in the early years of food aid, the long-run effects of food aid on poor people were bad—the discouragement of some domestic food production and employment. (ii) Until the late 1960s, India’s food aid managers were learning by their mistakes. Rural areas, especially remote ones, contained few fair-price shops and therefore received relatively little consumption support from food aid. Yet these areas suffered from glutted markets and price disincentives. (iii) The best estimate of the ‘disincentive effect’ of an extra ton of food aid per year, via extra food supply and lower farm-gate prices, upon Indian farmers’ output remains one made in 1977: that every ton of sustained annual cereals aid reduced Indian annual grain output by about one-sixth of a ton (Plocki and Blandford, 1977). (iv) This does not allow for the further depression of farm-gate prices by food aid on the demand side—by the consumers switching from market purchases of food to subsidized food aid in fair price shops. However, since such consumers are likely to be very poor, this effect is small: they enjoy extra real income (i.e. via subsidized, aid-backed food) mainly as extra food consumption, and do not much reduce market demand for home-grown food. If we allow for this small extra cut in price incentives, it might be reasonable to suppose that each ton of food aid reduced domestic cereals output in India by 0.2 tons. (v) The main cost to the poor was via employment. If the 0.2 tons of ‘lost’ domestic cereals output corresponded to land taken out of production, it meant some four to eight days of extra unemployment; if to land switched to other (usually less labour-intensive) crops, some two to four days; and if to lower inputs into cereals production and hence lower yields, some one and a half-to-two days.

Aid and Poverty in India 43 (vi) Apart from employment effects, there were consumption effects. If each extra ton of food aid indeed meant 0.2 tons of lost domestic output this still left 0.8 tons as a favourable balance.2 The poorest half of Indians spend well over 70 per cent of income on food, mostly on cereals; the richest fifth spend well below 40 per cent. Actions to stabilize food availability and to prevent consumer price upsurges, therefore, are bound to help the poor as consumers. Also, the use of food aid as wage goods, in support of agriculture-linked public works—and the learning of lessons about the need to stabilize, not depress, producer prices through cereal aid releases—improved the impact on poor producers. With India at least ‘self-sufficient’ in foodgrains in normal years, and with EEC’s wish both to remedy its past relative neglect of India3 and to dispose of its dairy surpluses, food aid to India has shifted towards powdered milk and butteroil. First around Anand in Kaira district, Gujarat, and recently elsewhere, ‘Operation Flood’ (OF) has sought to follow up the apparent success of Indian dairy development programmes. In Kaira at least, such programmes do seem to have reached many poor, even landless, owners of one-tothree milch buffaloes or cattle, and to have provided them with reliable and competitivelypriced marketing and processing facilities. ‘OF’ involves the sale of EEC milk products, mainly in cities, and the use of counterpart funds to support such dairy development. Unfortunately, the project is surrounded by intemperate attacks and excessive claims. Certainly, few projects have got gains to the landless, and this one has. But glutting markets with subsidized dairy-product aid, to raise counterpart funds, is an odd means to finance dairy-producer expansion. Milk producers in areas not benefited by Operation Flood suffer twice, as prices are reduced by EEC milk powder and by dairy products from OF areas, And one is worried about diverting increasingly scarce land from grains to dairy products, which (even if employment-intensive in production) remain too costly to be important in poor people’s diets. Most serious is the lack of genuinely independent evaluation of the poverty impact, or cost-effectiveness, of these large and controversial projects.4 We discuss the issues further in the context of project aid (Chapter 5). Volume of aid to agriculture But the main development since about 1970, affecting India’s poor via aid has not been in the sphere of food aid, but in crop production. Increasingly, donors came in the 1970s to believe that aid must attack the syndrome of ‘growth without poverty reduction’ by increasing the productivity of the poor (rather than by welfare schemes), and that expanded smallholder food production was the best path. Until the end of the Third Plan, agriculture had received only 1.8 per cent of sectorally allocable aid authorized for India (and 1.0 per cent of aid utilized). Even in 1966–74, the period of the big (and aid-backed) breakthroughs with new wheat varieties and of India’s main thrust to foodgrain self-sufficiency and stockbuilding, the proportion was only 9.0 per cent (4.6 per cent). In 1974–5 to 1978–9, it leapt to 29.6 per cent (19.7 per cent) (RBI, 1984–5, statement III). Meanwhile, the gap between the proportion of aid authorized and utilized for agriculture was steadily shrinking, due largely to gradually improving implementation of irrigation plans. By 1980–1 to 1984–5, indeed, agriculture’s share in authorized aid funds (29.4 per cent) was below its share of utilization (39.3 per cent). However, new authorized aid to agriculture, in real terms, was edging down again, to about 22 per cent of total aid authorized in 1982–5. This is partly because of a

44

Does Aid Work in India?

sense among donors that low world commodity prices are reducing returns to agricultural aid, and partly because GoI increasingly sees aid donors’ comparative advantage in hightechnology sectors and investments. Aid to agriculture: poverty-oriented? Aid to agriculture in India has been substantially, and increasingly, for irrigation, credit, extension, and other inputs in support of high-yielding variety (HYV) cereal production.5 Probably, a substantial rise in the share of Indian resources used in that way is necessary to raise the level of living of India’s hungry poor, but it is not sufficient. First, such aid might benefit mainly big farmers. Second, to the extent that it does reach smallholders, it might displace, not increase, Indian Government resource allocations to smallholder food production. Third, aid to smallholder food production might be inefficiently applied. Finally, it might leave untouched major or growing proportions of poor Indians, perhaps including most of the poorest—largely landless farm labourers, their jobs perhaps threatened by aid-linked mechanization of ploughing or post-harvest processing. Overarching all these problems is the fact that donors have overstressed the somewhat vulgar-Malthusian perspective that poverty and hunger are due mainly to inadequate available food per head of (growing) populations. Add the misperception that India’s growing numbers of hungry people are mainly ‘small farmers’ and one readily gets stuck with the perspective (Ford Foundation, 1959) that ‘more food from progressive smallholders’ holds the key to the reduction of poverty. In much of Africa and parts of India, this is valid; increasingly, however, poverty reduction (especially in Eastern India) depends on increased employment, and hence more real income—leading to extra exchange entitlements to food (Sen, 1981, 1986)—for near-landless workers. Agriculture is crucial here too, and aid can help; but the formulations and formulae of the early 1960s are no longer quite right. ‘Green Revolution’ and aid Before assessing the required changes, we must sketch the content of aid to Indian agricultural production. The most dramatic and cost-effective component was the support of research into HYV wheats, and recently other cereals. Even when based in India, such research was substantially helped by aid to international centres under the auspices of the Consultative Group for International Agricultural Research, CGIAR. Also, significant aid supported some of the bases of domestic Indian research and extension, including US support for India’s network of agricultural universities, and World Bank support for the ‘training and visit’ extension system. Yet the development, application, and extension of fertilizer-responsive wheat and rice HY Vs did not absorb even 2 per cent of India’s aid, and 2 to 3 per cent of world aid. Without such HYVs Indian cereal output, now typically enough for low-level self-sufficiency, would be 15 to 20 per cent less. This does not, however, ‘prove’ that the remaining 97.5 per cent of aid is unacceptably unproductive. First, research (and therefore research aid) inevitably seeks an adequate average return by mixing extremely successful activities with failures, and one obviously cannot predict exactly which will be which. Second, aid to agriculture also falls into this category to some extent; average rates of return are somewhat higher on Bank projects in this sector than in others, but so is the proportion of projects (and to a lesser extent, of money) with very

Aid and Poverty in India 45 bad returns (Lipton, 1987, 5–6). Third, many agricultural-sector investments—aided or not—showed decent but unsensational returns prior to HYVs, which themselves did very well partly because such prior projects, e.g. tubewells in Western Uttar Pradesh, already existed. The outstanding returns to and for research on HYVs, then, does not prove that all aid can achieve similar returns, nor that returns on most aid are too low. However, such research is indeed a powerful example of the cost-effectiveness, both in growing more food and (we shall argue below) in reducing poverty, of aid for research-based, food-linked biological inputs. But four cautionary notes must be added. (i) This dramatic success in parts of India—as in parts of several other Asian and Latin American countries—depended on a long-standing domestic network of agricultural research. There is statistical evidence that national research capacity6 determines the usefulness, for a developing country, of inputs from the (largely aid-financed) international research system (Evenson and Kislev, 1985). (ii) The anti-poverty success of the high-yielding varieties was due partly to the economic attractiveness of their product types for consumers and smallholders. In the early stages of the green revolution, this attractiveness was largely fortuitous. It certainly cannot be assumed for other countries or future developments. (iii) The high returns to small outlays of aid for agricultural research would have been impossible without large prior outlays (partly aid-backed) for other agricultural investments. Such outlays, especially on irrigation, prior to HYVs produced decent but seldom sensational returns. (iv) The early successes of HYVs, on reliably watered rice and wheat, have proved harder to achieve elsewhere. The green-revolution inputs were developed to meet a problem described in the following terms by the Ford Foundation (1959). Until the early 1960s, modest yield increases could suffice to keep food production slightly ahead of population, because the cropped area could still increase. After about 1963, as was clear well beforehand, almost all economically arable land would be in use. To feed India’s growing population at existing levels of demand—let alone those generated by growth of income-per-head—would increasingly use up foreign exchange, thus impeding imports seen to be needed for industrialization. Food aid was not reliable in bad years, and might in any case depress local food output. Thus it was the perceived foreign-exchange consequences of area exhaustion alongside population growth, not considerations of poverty, that drove the yield-oriented green revolution strategy, and especially donor support for it. Nevertheless, contrary to much ill-informed discussion, the HYV technologies were oriented towards inputs and effects that—perhaps more by luck than judgement—were, in their techno-economic nature and potential, dramatically pro-poor, even apart from their main effect in raising food availability (Lipton, 1978; Lipton with Longhurst, 1989). On the side of production, the HYVs—being short-strawed, fertilizer-responsive and high-yielding—greatly7 increased demand for labour; thus both the small farmer (with ready access to family labour) and the hired worker benefited. On the consumption side, the breeding of varieties for high grain weight not only raised grain supply and hence restrained price, but produced types of grain that were mainly ‘inferior goods’, standing at a 10 to 20 per cent price discount below the generally finer traditional varieties, and thus

46

Does Aid Work in India?

being especially attractive to poor buyers and to producers who aimed mainly at low-cost, household-level consumption. For both production and consumption, the HYVs reduced risk: for consumers, by enabling a build-up of grain stocks, and by concentrating output growth in relatively low-risk and water-controlled regions; for producers, by a steady stream of varietal adaptations to reduce crop vulnerability. All these pro-poor elements were, more or less, undermined by the dilemma outlined on p. 47, as applied to HYV-based agricultural policy in India. Urban elites and the larger farmers who supplied food to the cities were largely able to divert many of its gains to themselves. The production benefits of increased yield for farm labourers have been greatly reduced as ‘bigger’ farmers (not huge kulaks, but operators of 10 to 25 acres) lobbied successfully for subsidies for tractors and other machinery (and fuel for them), so as to replace workers as they became costlier. (Higher yields made it feasible to pay for such new inputs even with quite small subsidies.) There is nothing about the HYVs that requires labour-replacing complementary inputs8 but the political process—weakly organized rural workers in growing numbers; and larger farmers on whom governments depended for urbanized surpluses of production and savings—has linked the two. Aid donors have often been insufficiently careful—as late as 1973, World Bank-supported combine harvesters, presumably supplied to India on IDA terms, were on sale in Ludhiana—and may sometimes have made things worse. Such processes were at work all over Asia; by the mid- 1980s, each 10 per cent yield increase due to HYVs brought only a 1 to 1.5 per cent rise in labour use, as against the 4 per cent typical of the late 1960s (Jayasuriya and Shand, 1986). But the real lesson, for governments and donors alike, is that the potential ‘pro-poor’ effects of HYVs via extra demand for labour are safe only if (i) the great majority of poor farm workers are self-employed most of the time, or if (ii) (near-) landless farm employees take most of the gains from increased labour demand as higher levels of employment rather than as higher wage-rates. These, plus big-farmer political lobbying mean labour displacement, partly or wholly offsetting the initial gains to poor workers from higher demand for labour. If there are to be subsidies, perhaps they should be to labour migration, not to labour-displacing inputs or the credit to buy them. As regards reduced risk and instability, the favourable effect on the poor of HYVs (and aid to support them) has been undermined by their extreme regional concentration upon areas providing surpluses—of food and savings—for urban use. Thus a small shortfall of rain (or a pest problem) in a few critical—and co-variant—irrigated districts now makes a disproportionate difference to grain output, and even more to marketed surplus and hence urban food supply, at all-India level. Thus, despite the fact that there is usually a reduction in output instability for each individual farmer who adopts up-to-date, more robust seeds, the heavy concentration of such seeds in a few areas, in order to provide a more easily procured surplus, has probably increased co-variance among farmers, and the instability of total output (Hazell, 1982). This can destabilize consumption (i) if stocks do not vary much, and if urban availability is no longer so much affected by import levels;9 (ii) if, as is likely, cutbacks in output mean even bigger proportionate cutbacks in employment. Betteroff middle farmers, when harvests are bad, increase the proportion of farmwork performed by family labour; thus the (poorer) labourers face reduced job prospects just when food prices are pulled upwards.

Aid and Poverty in India 47 Cereal imports—in the context of the strategic food problem, as seen by donors and planners alike in a period when economists concentrated more on import constraints and macro-policy than on poverty reduction or micro-markets—are also critical in determining the consumption benefits from HYV cereals. The very large production increases from biochemical innovation have been used almost entirely to replace foodgrain imports, and to build central grain stocks and replace deterioration in them. Hence calorie availability per person is not significantly higher in India today than in 1950–3 (FAO, Food Balance Sheets, and 1984, 179). Calories per consumer unit among the poor have probably not declined or increased much—though there have been offsetting regional trends, themselves reflecting the patchy spread of HYVs (see Subbarao, 1987, 4). This very pattern suggests that without the biochemical strategy, and especially the new cereal varieties, many more poor people in India in the 1970s would have died. Given oil price rises, the stresses affecting Indian imports would, in the absence of cereal HYVs, have worsened poor people’s consumption levels, probably substantially. But HYVs did not suffice, in face of demographic and politico-allocative pressures, actually to raise most poor people’s food consumption levels. However, even if the extra yield had added to cereal imports rather than displaced them, without structural change to raise demand for grains, prices of cereals to producers would have fallen. Suppose the price of aggregate farm output (in terms of farmers’total purchases) had fallen 20 per cent; probably that output would have fallen by about 3 to 4 per cent, and farm income by about 23 per cent. Employment in rural areas would have fallen—perhaps half as fast on the farms as output, but also off the farms, as declining farmers’ income reduced their local expenditure. Given the distributional factors that allocated the fruits of the HYVs to the not-so-poor, such innovations—even though potentially pro-poor—had to fail the poor, either on the side of production (if imports were maintained, and farm prices and hence output and jobs were thereby restrained), or on the side of consumption (if imports fell, and food availability per head was therefore not pushed up by the HYVs). Crudely, the former was the effect of the policy for rice, the latter for wheat. Sorghum and finger millet HYVs did better for the poor. Hence the GoI and aid donors, if they were to embody HYVs in projects providing long term gains for the rural poor, had to select paths of growth that steered assets or jobs to them. Otherwise, the poor could not afford more food, even if it were produced; and maintenance of producers’ incentives would depend on the use of any extra food to displace imports.10 However, donors have not ‘grasped the nettle’ of how aid to India might be used to get assets or jobs directly to the rural poor. Indirect means have been preferred, principally through support of irrigation, credit, and extension; these have indeed complemented HYVs, and have thereby helped to prevent absolute falls in food availability for the poor. However, such indirect methods have not avoided the biases that, in any highly unequal system, tend to steer the gains largely to the rich. The lesson of donor involvement in Indian irrigation, credit and extension is that ‘focus on poverty’ depends, not only on sectoral involvement in producing poor people’s foods, but on explicit action to strengthen poor people’s asset base or job opportunities, and thus their bargaining power, Such action is mainly the responsibility of governments in developing countries, not of aid donors. However, within a large and democratic federal polity such as India, there are usually many instruments and intermediaries that donors—if they wish—can select to get assets to the poor.

48

Does Aid Work in India? Irrigation

Aid flow data do not show irrigation separately; however, it was a large component of aid flows, The World Bank and other donors showed insight within the irrigation ‘policy dialogue’ (see Chapters 3 and 5), but the genuine and important process of mutual learning, while it improved system efficiency, has failed the poor in several ways. First, the evidence that simple wells (not tubewells) had much the best incomedistribution effects (Narain and Roy, 1980) was not translated into aid. Second, main-system design has until recently understressed both technical devices and irrigation-management choices that might improve crop mixes, reduce corruption (and overuse by persons nearer the source), and hence improve the access by poor users, especially tailenders (Wade and Chambers, 1980); donors spent much energy in persuading water-users to co-operate, but too little, until recently, in creating main systems that made co-operation attractive. Third, some donors’ bias in favour of major new structures has slighted not only local construction methods, but, more important, water management and maintenance. Fourth, big irrigation has often driven out small. Deep tubewells in Northwest India lower the water-table and poor people’s shallow wells run dry. Some donors are correcting these errors, which led to equity losses and gross under-performance of irrigation systems (Chambers, 1988). Despite the failures of irrigation expansion to reach the poorer regions, or even many poorer farmers in richer regions—and despite shortfalls in actual behind stated, and stated behind planned, irrigated area—irrigation expansion, especially of groundwater, has been essential to attain the huge expansion of grain output associated with the HYVs. It is impossible to say how much of the new irrigation capital paid for out of aid would have happened any way, using GoI money, without aid, but some would not, Indian total investment was constrained by shortages of foreign exchange for most of the 1960s and 1970s. Moreover, in India—unlike sub-Saharan Africa (FAO, 1978, 1979)—the expansion of aid to agriculture since the mid-1970s has not been offset by a corresponding fall in the sector’s share in domestically financed investment, in which irrigation plays the major role. Rural credit Donors have provided substantial support for institutional rural credit expansion in India (OED, 1981)—both in the ‘co-operative period’ and in the more recent period of commercial-bank emphasis. Many donor inputs have flowed through apex bodies like the Agricultural Refinance Corporation and its successor, the National Bank for Agriculture and Rural Development (NABARD). The growing emphasis of the USA and other donors upon the role of private enterprise in economic development should make them even more favourably disposed to aid rural credit (which each farmer applies to support his own productive activities) rather than state investment Yet aid for rural credit is a major source of controversy in regard to overall usefulness, as well as to effectiveness in reaching the poor. The controversy centres on two issues. First, primary organizations (village-level co-operative societies, branch banks) generally show dismal, non-improving repayment and default records. The appearance is somewhat excessive; e.g. in Southern Tamil Nadu, where 40 per cent of loans are recorded as overdue, only 4 per cent of lending is seriously enough in default to warrant court action or writing off. The rest may be no more than the results of delayed repayment. Yet default is reflected in growing government subsidies, largely as de facto grants or write-offs to allow apex

Aid and Poverty in India 49 organizations (such as NAB ARD)—themselves apparently very efficient—to make up loans not repaid by farmers to local branches, and hence by branches, via national bodies like commercial banks, to the apex. Such chronic deficits, apart from impairing financial management and diverting public savings away from planned investment uses, help mainly non-poor farmers. Yet, despite the bad financial and social record, the economic performance of lending in support of rural credit in India—as reflected, for example, in the economic rate of return recorded in World Bank post-evaluations—generally looks much better than that of either farm or non-farm projects as a whole (Lipton, 1987a, 15; OED, 1981). This is perhaps due in part to the attribution to private investment out of rural credit of some benefits in fact accruing because of public investments; but, even when this is allowed for, the economic performance of aid to rural credit in India looks good. Yet the financial performance does not; institutional lenders, such as rural branch banks, have great difficulty in recovering their loans, especially from powerful large farmers. This remains the case even when we allow for the fact that much alleged ‘default’ is merely roll-over from one year to the next. Second, major efforts are made, in some aid-financed credit schemes, to provide lowinterest loans confined to poorer farmers and areas; or to link credit to the purchase of assets in small amounts by poor households, as with IRDP. Yet substantial parts of such subsidized institutional credit are collared by influential, articulate, and normally richer farmers. This is not only an inequitable use of aid; the cheaper credit also encourages the non-poor rural recipients to adopt longer-maturing, i.e. more capital-intensive, sets of inputs and outputs. Credit projects are analysed in Chapter 5; here we address the issue of whether aid to rural credit can help the poor. Donors’growing scepticism about aid for rural credit has been fuelled mainly by experience of Latin American rural banks and cooperatives, which—first at branch level, eventually at the centre—have been bankrupted by subsidizing on-lending and permitting defaults, while rapid inflation eroded what did get repaid.11 Even when Indian formal rural credit depended almost entirely on co-operatives, it was never typified by such indiscipline as in much of Latin America, nor surrounded by such rapid inflation. Yet default rates were high. This led in 1982–3 to the cessation of aid to NABARD by the USAID (despite an expanding overall programme of US aid to India) and to a threat of suspension, until overdues were reduced, by the Bank group. Within this group, a major attack on subsidized credit was launched (e.g. Adams et al, 1984). In 1989, the Bank group stopped new support. Such reactions are not motivated only by justified concern for financial stability—or even by less-justified doubts about economic returns to aid for rural credit—but also by concerns about equity and poverty. Since loans (for purchased inputs) form a much higher proportion of production costs among better-off farmers than among poor ones, aid for rural credit, especially but not only12 if on-lending is subsidized, appears to involve transfer to the better off. This is even clearer when, as is usually the case, richer farmers’ default/ loan ratios are higher (Lipton, 1976). The effect on the poor is still worse if—as in the case of major World Bank credit projects—the credit permits the purchase of tractors, profitable allegedly because they increase output, but in reality because they displace labour.13 It is, nevertheless, mistaken to conclude that aid for credit does not help the poor in absolute terms (though it may raise rich people’s incomes by proportionately more). In India, such aid has almost certainly raised the supply of both institutional and total rural

50

Does Aid Work in India?

credit (i.e. fungibility for the state, and to intermediary public-sector lenders, of aid for rural credit has been well below 100 per cent). This increased and more competitive supply of credit has reduced its price to the poor, not only to the rich who initially benefited. At farm level, certainly, some production loans have been, or have permitted farmers’ own resources to be, diverted to consumption (i.e. fungibility for end-use borrowers of rural credit is significant); but such diversion is less likely for larger farmers, well below 100 per cent even for small ones, and not obviously damaging anyway.14 Low or subsidized rural interest-rates in India did little to discourage the supply of savings, but much to raise the demand for cash to invest (Iqbal, 1983). This must have raised investment in, and hence availability and stability for poor people, of food production. Moreover, extra rural credit usually increased the use of current inputs such as fertilizers, whose expansion makes so much difference to HYVs. These inputs absorb the great bulk of credit for Indian rural production, and are highly complementary with extra work from poor hired labourers.15 If some credit has been diverted to labour-displacing investments, the fault lies much more with overt subsidies to capital (and fuel) than with credit or credit subsidies. These, in any event, have been dwindling in India. This does not justify complacency about the effects on poverty (or efficiency) of aid to a default-ridden, and in places still subsidized, rural credit system. However, in village economies that are credit-constrained even at existing production levels, the increases in input use needed to make high-yielding varieties profitable for farmers (and hence to achieve extra local employment and food security for the poor) depend on extra credit supply. Initially, except where there has been a really dramatic breakthrough in agricultural productivity, this may well be acceptable to poor and risk-averse farmers only at rather lower rates than the informal credit market will provide. The leakage of formal credit into ‘rich farmers’ subsidies’ can best be tackled by a gradual shift to more realistic interest rates, firm action against defaulters, and avoidance of price or credit biases towards labourintensive inputs. Donors increasingly find Indian credit authorities supporting such aims, albeit in the teeth of local élites and their ‘official’ supporters. To withdraw from aid to rural credit would be to retard, not to assist, its effective reform—and to damage the prospects of spreading, to smallholders in less promising and safe areas, capacity to buy modern inputs and hence to provide poor people, as workers and consumers, with more local work and securer food. In the longer term, however, direct producer benefits to the rural poor, linked to aid for credit, will require access to productive assets; IRDP, for all its faults, is a step in this direction, but major long-run progress is likely to require land reform. Other agricultural aid and poverty Despite the world-wide tendency for extension workers to reach more ‘acres per day’ (and to satisfy patrons) by concentrating on big farmers, the extension services also remain an important determinant of the speed at which many innovations reach poorer and less wellinformed people. Major aid support—amounting to about half the total cost of extension, which was 0.22 per cent of annual GNP from Indian agriculture between 1977 to 1982 (Moore, 1984)—has gone, chiefly from the World Bank, to the ‘training and visit’ system of agricultural extension in India. ‘T and V may have set the scene for major improvement in Indian extension by removing from extension workers the panoply of tasks in input

Aid and Poverty in India 51 distribution (and paperwork) that had encumbered them. It has replaced such tasks with clear messages, initially to identified ‘contact farmers’. It has defined duties, training and overview in the framework of farm needs, instead of crop-specific traditions. However, if T and V is to help poor farmers, extension workers have to be trained and motivated to this end; and critics complain that career structures of village-level workers remain largely neglected, yet that T and V has diverted key senior personnel out of the overall rural administrative system that is needed to make extension useful. As for training, by early 1983, only 11 per cent of the Indian T and V vote for training had been used. The ‘contact farmer’ method had degenerated, via a search for ‘progressive farmers’, into a widespread misperception that wealth and literacy were, as such, desirable in the initial audience for extension messages. Part of the problem may have been the familiar ‘pilot illusion’: the T and V ‘pilot project’ (Benor and Harrison, 1977) had gained greatly not only from high ‘pilot’ inputs, but from the progressiveness of agriculture in Chambal district, Rajasthan, and from the suitability of centralized ‘fortnightly T and V messages’ to a rather homogeneous canal area. As T and V was rapidly spread across successive states—partly because state governments saw it as a source of cash and jobs—it had to handle less promising, more diverse, and more polarized farming conditions. The gains from T and V (as indeed from any form of extension) in such circumstances—and especially its effects on poor, illiterate, non-contact farmers—are more dubious (Moore, 1984), though some increase in adoption of recommended practices has been demonstrated (Feder and Slade, 1986). Such a critique, however, may be too short-term, not because the above difficulties are mere teething troubles; rather, the long-term prospects of extension–and of its orderly management via T and V–depend upon preparing farm habits for (and readying extension officers to make the best of) subsequent innovations. These may include new biological and chemical inputs that provide significantly more profitable prospects for smallholders– with expanded labour inputs–to grow more food for poor consumers. The districts that achieved the apparently cheap, dramatic green revolution breakthroughs of 1965–72 were those with substantial earlier outlays on traditional research, water control, credit, and above all conventional, in the short term not very high-yielding, extension (Evenson and Kislev, 1975). True, HYVs’ potentially ‘pro-poor’ biases in production have been in part subverted by the ‘pro-rich’ biases of access: to extension information, credit and irrigation. Aid donors have shown too little response to this. However, the evidence on biological and chemical innovations is clearly that, though bigger farmers gain first and most, poor ones follow. The role of ‘anti-poverty’ aid to extension is to help to train workers motivated to help poor farmers to adopt swiftly–and, if reasonably profitable, labour-intensively16–as and when such innovations become available. In many developing countries, donors have sought to support attempts to increase poor people’s bargaining power by means of ‘integrated rural development projects’.17 Cutting across ministerial boundaries, such projects seek to tackle a whole range of interlocking barriers to poor people’s advance: in agriculture, credit, health, education, etc. In practice, many such projects have depended heavily on non-replicable levels of local administrative input or, more usually, on project personnel from the aid agencies themselves. Such personnel, even if very able, are not integrated into national administrative systems. When they leave, those systems, inevitably divided into departmental areas of responsibility, must take over the project, as merely one of many responsibilities. Too often, coherence

52

Does Aid Work in India?

and purpose dwindle, and buildings alone remain to impress, as ‘cathedrals in the desert’,18 It must be said, however, that integrated projects in Asia, especially in Sri Lanka but also in some parts of India, have a much better record than in Africa (Lipton, 1987, 8–10). There is a compelling logic, despite its current unpopularity, in the view that a given public outlay in a rural area (or elsewhere: see p.178) will do best if divided among complementary activities—health and food production; irrigation, crop development and extension—than if concentrated on just one. The problem is: how to exploit the multisectoral complementarities, without risking either administrative incoherence or a sort of area-level donor neo-colonialism? Rural assets for the poor? Both the central and state governments have, sometimes with donor support, engaged in multi-sector activity. Comprehensive watershed management, the Drought-prone Areas Programme, and other central schemes have aroused opposition from line ministries, and have recently been played down. West Bengal’s ‘Comprehensive Area Development Programme’—more orientated to the village level and linked to land rights—remains a significant effort. Both centre and states, however, have largely avoided ‘integrated’ projects of the variety that, in the name of fighting poverty in the short run, disintegrate the long-run administrative capacity to do so. Unfortunately, donors, faced by this restriction on error, have not been presented by GoI with (and have not themselves adopted) any clear alternative anti-poverty strategy (at least rurally; urban housing aid and health aid nationwide, are partial exceptions). India’s Seventh Plan, for the first time, contains such a strategy, with the ‘twin pillars’ of the ‘integrated rural development programme (IRDP) and the rural employment programmes; but it is not a strategy for which donor inputs—increasingly focused by GoI on higher-technology activity and increasingly ‘tied’ by bilateral donors—are obviously appropriate, or sought; and it comes when India’s real aid receipts are lower than ever, and when the salience of the Plan in government thinking (and of poverty in some donors’ thinking) is not clear. Moreover, to ensure that the benefits of rural change are steered towards India’s poor, those poor must gain local power, with a base in the local economy. Political power in a state capital has, on its own, severely limited scope to help the really poor, because ‘poor people’s parties’ in India are weak; torn between hostile constituencies (Herring, 1983) and limited by the local power, over law as well as economy, of the better-off. Could the poor gain local economic power from an emerging labour shortage? The rapid growth of the Indian workforce, and the world context of technical progress that saves on high labour-costs both militate against that. The alternative is for poor people to gain more ownership, privately or jointly, of the means of production. There is nothing statist or centralizing about this proposal. Income and asset distribution in rural India are unequal by Western (or African) standards. Yet abundant labour—and persistent scarcity of savings, domestic or foreign—should dictate labour-intensive production, such as is associated with small, fairly equal family enterprises. At issue is not ‘state ownership’—indeed, the state is bound to reflect existing power—but diffusion of private control over assets. Suppose, however, that the state reflects the dominant, asset-owning groups within existing power structures; that donors see their own interests as lying in the use of aid to help the efficient poor; and that this ultimately requires asset redistribution. The scope of

Aid and Poverty in India 53 aid to influence a strong, sovereign, and in India elected, state in these matters should not be exaggerated. Nevertheless, many parts of the Indian state would welcome an aid emphasis upon asset redistribution to reduce poverty. Donors have done rather little to respond with prompt offers of cash help to support the many official Indian policy statements favouring redistribution of rights in land and water. Land reform—where efficient, and where planned in conjunction with other input supplies—is formally supported, as an object of aid, not only by such countries as Sweden and Holland but by the leader of the Indian aid consortium (World Bank, 1975–82). The widespread belief that there is, in India, no land reform to support is quite wrong; not only have several million hectares been officially redistributed since 1960, but much more has been privately ‘redistributed’, via market sale or via transfer to poor family members, by big owners, fearful of the implementation of official land reform and its ceilings legislation (Vyas, 1979; Lipton, 1985). Despite the avowed policies of most donors and GoI in favour of land reform, aid to India has done very little of significance to assist it—not even on the modest scale achieved, in the much less auspicious political circumstances of North-East Brazil, through World Bank support. Moreover, where land reform appears unacceptable, there are other approaches that get assets to the poor. For example, Oxfam in Western Orissa and Proshika in Bangladesh (Wood, 1984) both provide landless workers with sources of, and rights to, irrigation water, which is then sold to farmers. The IRDP, which is being massively expanded in India’s Seventh Plan, seeks to identify the poorest households in selected villages in each block, and then to provide them with loans, tools, training, and supervisory help in the acquisition and use of income-generating non-land assets. Assets’, inputs’, and HYVs’ impact on the poor Research and development of HYVs of food staples, with necessary support from expanded irrigation, credit, and extension,19 have done much to prevent decline in the food consumption of India’s growing numbers of poor. Huge public outlays have been required for that support, and significant though much smaller outlays for the HYVs themselves. Aid has helped to finance these outlays, and to provide useful, two-way policy dialogue. Aid outside India, to international research into cereals, has also made a major impact, in conjunction with sophisticated adaptive research in India, upon food availability (and emergency stocks). However, outside the green revolution lead areas, the poor as producers have gained little from aid. The success in ‘holding the line’ as regards poor people’s consumption—perhaps a bit better in the 1980s—despite populalion growth, alongside the failure in most respects to strengthen them as producers, is mainly a reflection of Indian efforts, successful in growing more food, but disappointingly inequitable. North Indian HYV wheat—and probably HYV rice in South India, and hybrid sorghum in the West-Centre—is now adopted as widely by poor farmers as by rich ones, and perhaps farmed more intensively. However, despite the pro-poor nature of the HYV-fertilizer technology and its recent spread, the gains in adopting areas have accrued mostly to the middle, not the poor, farmer; landless and near-landless labourers, who found employment rising by about 40 per cent when HYVs doubled yields (as in several districts they did in ten years), now find the employment rise barely 10–15 per cent.20 While no state is without major HYV areas, there are huge blanks (especially in the East-Central Indian ‘poverty square’) where poor wheat farmers in small surplus (and their employees) have actually

54

Does Aid Work in India?

lost from lower prices due to HYVs. The most striking symptom of the triumph and tragedy of poverty-orientated food efforts—and aid—in India is the build-up of a 20 to 30 million ton public grain reserve, alongside an almost unaltered incidence of under-nutrition. The ultimate reason, about which aid can do little, is lack of ‘poor power’. But the immediate reasons is that since the very poor are increasingly near-landless labourers (Lipton with Longhurst, 1989), the effect of agro-technical change on rural poverty depends more and more on its employment impact, not on its impact on ‘big’ or ‘small’ farmers.21 However, farms with less land per family member are more labour-intensive than other farms; also, by absorbing the labour of their moderately poor operators, small farms keep their owners from seeking hired work, which is left to the very poor as an income source. Therefore, the growth of small farming is a crucial weapon in the struggle for high-employment agricultural growth patterns. Why, despite their potential techno-economic merits for small and labour-intensive farms, have not HYVs in India helped them more? The first immediate reason is the adoption sequence. Whatever the objective risks,22 a major innovation is always subjectively risky, until farmers’ observation of nearby successes and techniques has created widespread confidence. Hence poor, risk-averse farmers are seldom early innovators–especially because bigger farmers have better access to early information. So the bigger farmers adopt a new HYV first, and can sell more in the early stages of its spread–before its wide-scale adoption has greatly raised the cereal’s supply and brought down its relative price. Anything that helps smaller farms to adopt innovations sooner will also help them to share more in the gains–and to deliver employment benefits as a result. Such projects as the UK-aided ‘fertilizer education scheme’ in Orissa are clearly relevant here. The second reason why smaller farmers do less well than the HYV technology would suggest lies in its delivery system. Timely and adequate credit tends to go to bigger farmers, often through subsidized official channels. Inputs, especially appropriate fertilizer-mixes, possess economies of scale in delivery, again favouring larger users. Yet extra inputs (and credit to buy them) are often needed to justify outlay on better seeds. The third reason lies in regional imbalances in research and irrigation. HYVs were at first heavily concentrated in a few reliably-irrigated areas. By increasing domestic marketed surplus, this sometimes actually reduced incomes in competing areas, for example parts of Madhya Pradesh. This has happened even in well-favoured areas, to some farmers unable to secure reliable water. Of course, that is not a case for slowing down HYVs in suitable areas, or for forcing them on unsuitable ones. It does make a case for research (and sometimes irrigation) to shrink the pale of unsuitable areas—and for using some of the ‘gains of the green revolution’ to steer new, potentially economic, off-farm activity to poor people in those regions. Aid for research in semi-arid areas has helped with the first task; but donors have not yet contributed much support for non-farm rural activity in ‘backward’ areas. Fourth, we have referred to adverse price effects on some of the rural poor. As big farmers in favoured areas, helped by early innovation, raised output, prices were pressed downwards. This reduced the value of real outputs—whether sold or eaten at home—from generally poorer farmers, from the later innovators, from those with less access to delivery systems, and from areas less responsive to HYVs. Aid, both for HYVs and for irrigation and other supportive inputs, by its regional concentration tended to increase these disparities. However, by concentrating on extra output in rural areas most readily amenable to rapid HYV-based growth, aid helped to employ the many near-landless in such areas. Such aid,

Aid and Poverty in India 55 by raising food supply and thus restraining food prices, also helped the urban poor. In the short term, this effect also helped the rural poorest—net buyers of food—by restraining food prices. However, the longer-term impact on these groups via rural employment, especially in some non-HYV areas, may be damaging. Fifth, underlying all these sequences of frustration—these slips between the cup of poverty-reducing HY Vs, and the parched lips of the poor—is the motivation of Indian food policy. Essentially, and despite the genuine concern (predating that of most donors) of large parts of the GoI machine to reduce poverty, the central aim of Indian food policy has been to obtain an adequate domestic supply of food, so as to ensure that no major foreign-exchange shortage would be threatened by a need to import food. Planners with this ‘central aim’—however genuine their hatred of poverty—are bound to be at least half pleased when food surpluses, delivered to urban areas, surge ahead. The ‘central aim’ is also almost unavoidable, once GoI accepts that food output in the wake of HYVs increases much faster than poor people’s purchasing power. In that case the extra output at least of the main urban grain (wheat) must be used, first to displace food imports, then to build up stocks—not to add massively to supply; for, if the latter happens, farm-gate prices of wheat could tumble out of control, and future investments in expanded farm output would be discouraged. This ‘central aim’ also explains why Indian (and donor) authorities, in large part genuinely concerned for land reform and employment-intensive technology, have—at least—tolerated dynamic damage to the rural poor, in the wake of HYVs even if not necessarily as a result of them. This dynamic damage is probably at least as important as the static maldistributions we have discussed. First, in the Punjab and Haryana—in marked contrast to other Indian states less ‘overrun’ by HYVs—the process of tenancy resumption, well under way before the HYVs arrived, has accelerated in their wake. Since the early 1960s, owners of 10 to 20 acres have resumed tenancies and turned tenant farmers into landless employees, while owners of 0 to 1 acres have ceased to farm the land themselves and have rented it out to medium-to-large operators. The tiny ex-owners are probably better off as landless employees in the Punjab’s post-HYV boom—certainly the incidence of poverty sharply declined in that state in the 1970s (Chadha, 1983)—but ‘poor power’ (and security) suffer in the process. Second, again especially in these areas but also in the HYV rice areas of Tamil Nadu and coastal Andhra, labour-replacing innovation has followed in the wake of HYVs.23 Combined with the increasing reliance of India’s rural poor—as population growth has reduced holding size—on employment for income, this too has reduced the benefits of HYVs to them. It is demographics and politics, not technology or economies, that have shifted the benefits of HYVs—and hence of the aid that supports them—to some extent away from the poor. HYVs do not render profitable, in most cases, the tractors, weedicides, and modem rice mills that ‘unemploy’ poor workers, ‘save’ labour-costs, and add little or nothing to consumable output (Agarwal, 1980; Binswanger, 1978; Greeley, 1987). What happens is that HYVs first intensify seasonal labour peaks; then induce bigger farmers to press for subsidies to inputs that substitute for increasingly organized, or temporarily costly, labour; and finally induce urban groups to concede such subsidies. The new labour-displacing equipment is then used not only in peak seasons, but for much of the year. It displaces workers and is less costly for larger farmers. The process further concentrates HYV output (and benefits) with surplus farmers, who restrain food prices via their sales to urban groups.

56

Does Aid Work in India?

HYVs may well strengthen both urban food buyers and surplus farmers politically—but HYVs do not create an economic case for labour-replacing innovations, or for the exclusion of poor people from operating high-grade land. Nor, of course, do HYVs cause the steady demographic pressure that renders such innovation and exclusion so harmful to the poor. Given the politico-demographic framework—are HYVs nevertheless a cheap way for projects (and aid donors) to benefit the poor? Certainly, with GoI cooperation and a supportive local frame of research and extension, aid to cereal production does more for employment, food-price restraint, and hence poverty alleviation, than the heavy-industry patterns of aid that prevailed before the 1970s (see p.49). Since inadequate supply of wage-goods from agriculture in the 1950s and 1960s have regularly constrained nonfarm growth, but not vice versa—and since investment in agriculture has been regularly, and probably since HYVs increasingly, associated with much more GNP gain than equal amounts of off-farm investment (Lipton, 1968)—it is not plausible to deny the relative efficiency of farm investment and aid.24 Moreover, there does seem to have been a (small) real rise in agriculture’s share of total Indian investment since about 1970; unlike most of Africa, in India the ruralization of aid has not in practice been wholly fungible, i.e. has not led to a wholly offsetting fall in the rural share of domestic investment efforts. It is admittedly difficult to steer farm support to really small holdings; however, several states, and several national institutions of agricultural development, try hard, and with some success. Smallholder development tends to mean labour-intensive farming, and thus to encourage smallholder families to work on the home farm instead of competing with landless labourers for ‘commercial’ work. Also, it usually enriches people with a higher propensity to buy labour-intensive inputs and consumer-goods than does big-farm development.25 So there is some spin-off even to the landless. This is a specific economic process, not a vague hope of ‘trickle-down’. Nevertheless, the poverty of those with only labour-power to sell has, ultimately, to be tackled by off-farm employment and/or generation or distribution, of productive assets. Probably both off-farm employment and non-farm assets will not suffice; a combination of the HYV approach (spread to rain-fed areas and to ‘inferior’ cereals such as millets and sorghum) with land redistribution (so that extra HYV output accrues to poor growers who will eat their extra output, instead of selling to to swell central grain stocks) may well be necessary.26 In well-watered places, HYVs plus redistribution of access or demand or entitlements to food, can greatly ease the problems of rural poverty; and aid can help. What of the Eastern, Central and other areas of unreliable water? There has, since the mid-1970s (and due to combined efforts by Indian scientists and the aid-supported International Crops Research Institute for the Semi-arid Tropics, Hyderabad), been major progress with hybrid sorghum and finger-millet in some rain-fed areas (Rao, Rajpurohit, Sawant). However, the spread of HYVs to rain-fed areas is as yet small scale, slow, costly, and uncertain. ‘Fast food’ from HYVs will usually require, for many years, some degree of water assurance. But in most of India the inexpensive options in irrigation have been used up, and it is increasingly expensive as it spreads to new areas. Such rising costs—and management problems—are smaller in the case of farmer-controlled micro-irrigation, but many forms of this depend upon the costly use of fossil-fuel energy to lift water to the cropland. Also, while generally not more so than land (Narain and Roy), irrigation water can readily be maldistributed.

Aid and Poverty in India 57 Nevertheless, the spread of water control to poorer farmers and regions remains a key element in their food security. Irrigation—which has been supported by major aid inputs to Indian (and other Asian) agriculture—constitutes a major element in an effective shortrun attack on poverty and hunger, even if there are plenty of cereals in government stocks and the poor lack the money to buy it. Reliable irrigation can provide reliable jobs to the landless; can ease the process of agricultural diversification into labour-intensive and income-elastic crops, such as vegetables; and can stabilize food supply from poor farmers’ own land. Its success in these areas depends on an outlay on well-designed irrigation; land developed per rupee spent; irrigated area as a proportion of developed land; and access of poor people to such areas or to jobs (or cheap food) from them. (i) In discussing policy dialogue, we consider the World Bank’s role, learning as much as teaching, in improvement of strategies for Indian irrigation design. (ii) However, there has been less success in improving the rate of land development per rupee spent on irrigation construction, however good its design, (iii) Despite the new concern of the agricultural and irrigation experts with inadequate ‘water management’, shortfalls in irrigated area per unit of land developed below design expectations have, especially in South India, massively reduced the efficiency of outlays on irrigation, not least in producing extra food or work for the poor. (iv) Finally, the choice of systems that might raise the share, in irrigated area, of tail-enders and other poor users remains neglected, especially in the bigger aided projects. Although some small-scale irrigation systems are better at blending efficiency with equity, preference at the construction stage is often given to big gravity-flow systems. Such systems ‘build in’ preferences for larger farmers, head-enders or co-operatives of several adjacent farmers (usually in practice dominated by larger farmers or head-enders). Poor users are often harmed by neglect of negative externalities from richer or stronger water users: upstream growers plant ‘thirsty’ paddy, depriving tail-enders of water for a second crop; or deep tube- wells lower the water table and render hand pumps useless (Wade and Chambers, 1980; Howes, 1982). Improved design can reach the poor only if it anticipates problems and provides technical structures and/or incentives to irrigation engineers and managers to facilitate equitable solutions. In summary, aid has successfully supported the outlays on HYVs research, irrigation, extension, and credit in raising food output. This has prevented the rise in oil and industrial import costs (and the associated fall in food imports) from making the poor even poorer, as consumers. But such aid—despite the fact that HYVs are technically suited to employmentintensive production on small farms—must do more for the poor as producers. If poor people’s access to income from production does not rise, then in the long run their food consumption cannot rise either, unless real food prices fall. Higher food supply may restrain food prices—but, with population and workforce growing, employers can respond by correspondingly restraining money wages.27 Poor people, increasingly rural labourers rather than even ‘small’ farmers, did initially get much more work due to HYVs—but that gain was eroded as labour-displacing methods spread. HYVs are, as such, a ‘poor-specific’ technology; but they enter a power context in which their gains are largely diverted to help the rich. The answer has to be redistribution of agriculture-linked assets, and of work chances, towards the poor. Both donor and recipient rhetoric in India supports this. So do many specific, substantial outlays, actions, and agencies. But huge problems remain; and progress has so far been very modest. Since India—unlike most LDCs, and surprisingly in

58

Does Aid Work in India?

view of its fairly steady growth—has not significantly reduced the share of its workforce in agriculture (of about 70 per cent), it is plainly central to get assets and/or labour income to its agricultural poor. At a time and by methods agreed after consultation with the Indian Government, leading donors should clarify their readiness to lend in support of wellconceived programmes for redistributive land reform. Aid other than to food and agriculture: poverty impact Apart from raising private consumption of the poor via increases in food availability that restrain its price, or increasing poor people’s incomes from work or assets, aid (or domestic outlays) can take a third path towards reducing poverty. Several types of product are already purchased privately by the rich, yet are normally unavailable to the poor unless supplied socially, so that any expansion of public provision is by definition poverty-focused. At least three types of such service–primary-preventive health and primary education (World Bank, 1980) and contraception (World Bank, 1984)–have high economic rates of return, and major positive externalities, even when supplied free to people who cannot afford them. Health Health aid comprised 8 to 10 per cent of total public health outlay, capital, and current, in the 1950s, when PL 480 counterpart funds from the US paid for most of the attempts— initially very successful—at malaria eradication and control. The proportion fell steadily to 2 to 3 per cent in the 1970s (Jeffrey, 1985). However, health aid—and technical assistance from WHO and UNICEF—proved effective in supporting the ‘socio-economic’ wing of the Indian medical establishment against its ‘medical purist’ wing. The socio-economic wing has sought to redirect effort, away from high-technology teaching hospitals, towards preventive medicine, rurally focused, and making much more use of para-medical personnel. The medical purists aimed at exemplary standards of treatment, rather than at high costeffectiveness in use of medical resources. In the early 1970s, the socio-economic wing was in a minority, and the medical purists were dominant. By 1980, that had been reversed, substantially because of policy dialogue—mutual learning and exchange of knowledge, not arm-twisting—associated with health aid. Education Accompanying this reversal came a shift in educational outlays towards primary schooling. Aid, since it finances almost entirely capital outlays in this sector, pays for only a small part of total public-sector educational costs, which comprise mainly teachers’ salaries. However, IDA educational aid shifted heavily towards the primary sub-sector (World Bank, 1982) in India as elsewhere. This helped those forces in the Indian educational sector which sought parallel changes, shifting educational outlays towards the poor majority, who seldom proceed beyond primary school. Family planning In family planning, major Indian programmes began during the 1951–6 First Plan, long before any substantial donor interest. But ‘new model projects’ were aid-backed after

Aid and Poverty in India 59 1972—by Sweden in Uttar Pradesh, by Denmark in Tamil Nadu and Madhya Pradesh, by the UK in Orissa, by the UN Fund for Population Activities in Bihar and Rajasthan, and by USAID in five states. These projects all sought effective integration of family planning and child care. They were ‘new’, in that they at once sought to set contraception into the context of reducing infant and child mortality, thus persuading parents to reduce their family size norms; and sought to provide a wide range of subsidized services locally. It is poor people whose households and child/adult ratios are much the highest (Lipton, 1983a), who have least access to contraception and highest family size norms (World Bank, 1984), and who lose income if their workforces grow fast relatively to the demand for labour. Therefore, the poor gain most from aid directed towards family planning programmes. In 1951 an average Indian woman bore six children in her life; today, only four. But only Kerala—which has no special programme but has wider health access and more equal income distribution than other states, both reducing fertility—has significantly accelerated fertility decline among the poor (Zachariah and Kurup, 1984). Elsewhere fertility reduction is more successful if gains from growth are widely distributed among the poor (Repetto, 1979). Donors may need to seek out programmes that operate alongside effective action to raise the income of poor people, thus reducing their expected dependence on child labour (and on incomes from grown-up children, for security in old age). Urban housing Aid to urban housing has seen major reversals of donor strategy, pioneered by Canada and the Bank group. Unusually, such reversals were clearly more cause than effect of an improved poverty-focus of programmes in recipient countries. Until the mid-1970s, aid supported large, so-called ‘low-cost’, public housing programmes, which in fact were affordable only to the top 30 to 50 per cent of townspeople. Donors became uneasy about such patterns. The first move was towards loans to help poor people buy ‘site-and-service’ plots, with loans advanced to borrowers on the understanding that they would use simple materials (supplied) plus their own labour to construct owner-occupied homes. Next came a further move ‘downmarket’, towards loans for ‘slum upgrading’, again assuming owner-occupancy and dwellers’ labour. Most recently, donors and lenders have shifted from prohibition to positive encouragement of hiring-in of construction labour and rentingout of parts of improved homes. World Bank (1982) and other assessments indicate that these three steps have provided progressively increasing, and recently significant, shares of benefit, as far down the per-person income-distribution as the fourth, third, and even second-poorest decile of households. While the main examples of this success are not in South Asia, Calcutta has benefited substantially (see Chapter 5). In these urban schemes, one major drawback is the lack of opportunity provided, on most sites, for productive work and/or asset ownership. Such sources of ‘poor power’ are usually essential, in city as in village, if aid is to improve the sustainable growth rate of poor people’s incomes. An ongoing UK-aided project in Hyderabad is one of the few to combine slum upgrading (and site-and-service building) with support for productive on-site micro-assets—repair work, family retailing, etc. Using the Hyderabad City Department of Community Development as a focus, it seeks to link urban housing, education, family industry, and housing, in a sort of Integrated Urban Development Programme, directly

60

Does Aid Work in India?

responsive to highly localized committees of slum dwellers—initially, perhaps, subverted or led by ‘big men’ or slum landlords, but often, it has been found, reverting to more genuine participation later. Whether this will prove replicable or succumb to ‘pilot syndrome’ remains to be seen. Learning from India There is now widespread concern about the falling trend of food output per person in sub-Saharan Africa. There has also been considerable redirection of aid towards SSA with the aim of remedying this.28 Much of this extra aid has rightly gone in general balance-ofpayments support; however, such support—both sectoral (and maintenance) aid and ‘policy dialogue’ has increasingly emphasized agriculture. African governments have undertaken some policy reforms affecting food prices and marketing (though few have shifted significant real resources towards support of food production). It is increasingly obvious in almost every country in SSA (as in South Asia) that—with 70 to 80 per cent of poor people’s outlay spent on food, and 55 to 75 per cent of their working time used to produce it—‘poor people’ and ‘food adequacy’ tend to be helped or harmed by almost identical policy sets, including aid policy sets. Food adequacy, of course, depends on entitlements to food, as well as its physical availability. However, in much of SS A most of the poor are still small farmers, and so widespread rises in food output and availability in SSA—more clearly than in India—would directly improve poor people’s ‘entitlements’ to food (Sen, 1981) in most cases.29 If more rapid increases in food availability via extra output from smallholders would contribute to solving a major part of SSA’s poverty problem, how can aid help? In that context, India has more to teach donors about ‘aid effectiveness against poverty’ than to learn from them. First, India teaches that big rises in food output are necessary but not sufficient to roll back poverty in face of a rapidly growing rural labour supply. In the sense now being given to the expression ‘solving the food problem’ by most African countries, and by most recommended food strategies, India has solved its food problem: it has achieved self-sufficiency in food staples, given the effective demand for them in an average year, and has accumulated enough stocks to reduce acute food insecurity greatly. Yet it has not, at national level, substantially reduced the proportionate incidence or average intensity of poverty, under-nutrition or hunger. Aid to African agricultures could fall into a similar trap if it ‘merely’ helped to solve food production problems by raising output of food staples on larger farms—typically without greatly raising demand for labour. A typical producer of extra food staples under such a strategy, a big farmer or a tractor-driver, being decently fed already, would not greatly raise demand for staples as income rose. In such circumstances, either food prices fall, or extra output of food staples is offset by reduced food imports. If food prices fall, future investment in food production is discouraged, and employers can restrain poor workers’ money-wages to match the restraint in living costs (because supply of unskilled labour is elastic to the real wage). If net food imports fall to offset rising domestic output, food availability does not increase. In either case, extra food output alone—unless on poor people’s farms, or labour-intensive—need not create extra real incomes for the poor. Extra food supply, unless the poor also attain higher demand capacity, can leave them, as in India, almost as under-nourished as before in a normal year (though a bit safer in a bad

Aid and Poverty in India 61 year). Although, in most of SSA, greater access to new land enables the poor to gain more as producers from expanded food production than in most of India, the inexorably rising rural labour supply is eroding these options in more and more of Africa, as it has already done in most of Asia. The second ‘Indian lesson for SSA’ is more hopeful. India, despite national failure to reduce average poverty, has—with the help of very modest aid inflows—scored two sorts of success. (i) Nationally, the extra food output, centrally stored, plus improved information and distribution, mean that the terrible downward fluctuations in poverty, the onceregular famines and mass deaths, have been largely eliminated since 1943. Even extreme climatic misfortunes (e.g. 1965 and 1966; 1987; and the long Western Indian drought of the early 1970s) were met without major rises in death-rates (Drèze, 1987). (ii) As for poverty in average years, there is a plethora of local and scheme-specific successes. Some general conclusions are already being drawn from these successes (Subbarao, 1987). The several hundred formal evaluations of micro-projects in the GoI’s ‘Concurrent Evaluation’ of IRDP—while of mixed quality, and usually damaged by absence of proper ‘before and after’ comparisons or control-groups—provide much further information. Donors should, perhaps via the Consortium, request GoI permission to finance and coordinate a detailed review, to be undertaken by one or more independent Indian institutions, of the national, village, and project evidence. Such a review should cover not only ‘successes and failures’ in cost-effectively (and apparently lastingly) raising the level or stability of poor people’s productivity and/or well-being, but also the usefulness or otherwise of different sorts of aid and TC in different sorts of success. It is hard to envisage a piece of research more useful in the short-to-medium term for poverty alleviation in SSA, especially if African institutions of scholarship and administration can be involved. For India, there has also been a chronic training problem, which—in a proper context—such a review could help to solve. After the review had helped to evaluate the successive poverty alleviation programmes (local, state, and national; public, voluntary, and private), its results should be used to train personnel to staff such programmes. A review of Indian experience (and data) on poverty alleviation programmes, apart from greatly helping donors—and African administrators—could also catalyse one of the larger Indian research centres or universities, already involved in some work of this sort, to formulate a standing programme of ‘Poverty Action, Training, and Research’, possibly networked across several state-level research institutions. Such a programme should help central and state governments to select, design, implement, staff, and evaluate future DPAPs, EGSs or IRDPs.30 Finally, although there is some conflict of evidence, we believe that most donors have not done enough to render their own procedures conducive to cost-effective anti-poverty aid. (i) Many Indian officials believe that aid agencies’ inability or reluctance to pay local or recurrent costs pushed their aid away from poverty-oriented programmes, which often have a small direct foreign-exchange content.31 Many—indeed most—donor agency representatives, on the other hand, deny that this inability exists, or assert that it is avoidable. In respect of multilateral—or untied bilateral—aid, we see no reason why donors should refuse support of local capital costs. Recurrent costs present real problems, however. A project may die unless the aid agency takes some initial risks by supporting recurrent costs. Yet, if it is too ready to bear too many such costs for too long, the recipient government is not adequately involved in, or committed to, the project.32

62

Does Aid Work in India?

(ii) Aid-tying—which was certainly linked to the extreme concentration of early aid to India on industry—plainly and seriously impedes poverty impact. It tends to impose capital-intensity and import-intensity, at the cost of employment And it is difficult to identify imports to tie, in direct support of most rural poverty programmes. (iii) Consortium procedures insufficiently review poverty alleviation, let alone to encouraging donors (or recipients) to spend on it. The Bank group, as chairman, does prepare several excellent papers on basic needs, poverty, health, etc., for Consortium meetings. These papers appear to be tabled, but not incorporated into main documents, and, according to the minutes, not discussed when pledges are being gathered (see World Bank, 1978). The decline from the mid-1970s in the regularity or care with which poverty and unemployment are tracked in India may well, in part, be a response to the weak follow-up by the international aid community of its stated concern for poverty reduction in India. (iv) Internal relations, inside aid agencies, between central and line (project and loan) staff may militate against effective anti-poverty aid. Line staff are motivated to move money quickly, safely, and in big amounts. Poverty screening during the project cycle, however attractive to central staff responding to international pressures, is far less so to time-constrained line officials. Quick, simple, reasonably reliable guidelines are needed— even on such apparently commonsense matters as ensuring that poor persons displaced by irrigation schemes, often tribals without formalized rights, are not harmed. References Adams, D.W., Graham, D.H., and von Pischke, J.D. (1984) Undermining Rural Development with Cheap Credit, Colorado, Westview Press. Asian Development Bank (1977) Asian Agricultural Survey 1976, Manila. Agarwal, B. (1980) ‘Tractorization, productivity, and employment: a reassessment’, Journal of Development Studies 16:3, April. Ahluwalia, M.S. (1978) ‘Rural poverty and agricultural performance in India’, Journal of Development Studies 14:3, April. Ahluwalia, M.S. (1986) ‘Rural poverty, agricultural production and prices: a re-examination’ in J.Mellor and G.Desai (eds) Agricultural Change and Rural Poverty, Delhi, Oxford University Press. Alderman, H. and Sahn, D. (1987) ‘The effects of human capital on wages and the determinants of labour supply in a developing country’, Journal of Development Economics 29:2. Alexander, K. (1981) Peasant Organisation in South India, New Delhi, Indian Social Research Institute. Becker, G. (1981) A Treatise on the Family, Cambridge, Mass., Harvard University Press. Benor, D. and Harrison, J. (1977) Agricultural Extension: the Training and Visit System, Washington DC, World Bank. Bhalla, S. (1979) ‘Real wage rates of agricultural labourers in the Punjab’, Economic and Political Weekly XIV, 26, 30 June. Binswanger, H. (1978) The Economics of Tractorization in South Asia, New York, Agricultural Development Council. Cassen, R.H. (1978) India: Population, Economy, Society, London and Basingstoke, Macmillan. Chadha, G.K. (1983) Dynamics of Rural Transformation: a Study of Punjab 1950–80, New Delhi, Jawarharlal Nehru University, Centre for Regional Development. Chambers, R. (1988) Managing Canal Irrigation: Practical Analysis from South Asia, New Delhi, Oxford and IBH Publishing Co.

Aid and Poverty in India 63 Chaudhri, D.P. (1979) Education, Innovations, and Agricultural Development, London, Croom Helm. Dasgupta, B. (1977) Village Society and Labour Use, New Delhi, Oxford University Press. Drèze, J.P. (1988) Famine Prevention in India, London, LSE Development Economics Research Programme 3. Evenson, R. and Kislev, Y. (1975) Agricultural Research and Productivity, New Haven, Yale University Press. FAO (1978, 1979, 1984) State of Food and Agriculture, Rome. Feder, G. and Slade, R. (1986) ‘A comparative analysis of some aspects of the training and visit system of agricultural extension in India’, Journal of Development Studies, 22:2, January. Ford Foundation (1959) Report on India’s Foodgrain Crisis and Steps to Meet It, New Delhi, Government of India. Greeley, M. (1987) Postharvest Losses, Technology and Employment: the case of rice in Bangladesh, Boulder, Colorado, Westview Press. Hazell, P. (1982) Instability in Indian Foodgrain Production, IFPRI Research Report 30, Washington DC, International Food Policy Research Institute. Hazell, P. and Roell, A. (1983) Rural Growth Linkages, IFPRI Research Report 33, Washington DC, International Food Policy Research Institute. Herring, R. (1983) Land to the Tiller, New Haven, Yale University Press. Howes, M. (1982) The creation and appropriation of value in irrigated agriculture’ in M.Greeley and M.Howes (eds) (1982) Rural Technology, Rural Institutions and the Rural Poorest, Comilla, Bangladesh, CIRDAP/IDS. Iqbal, F. (1983) ‘The demand for funds by agricultural householders: evidence from rural India’, Journal of Development Studies, 20:1, October. IRRI/ADC (1983) Consequences of Small-farm Mechanization. Los Banos, Philippines, International Rice Research Institute. Ishikawa, S. (1918) Labour Absorption in Asian Agriculture, Bangkok, International Labour Organization. Jamison, D. and Lau, L. (1982) Farmer Education and Farm Efficiency, Baltimore, Maryland, Johns Hopkins University Press. Jayasuriya, S.K. and Shand, R.T. (1986) Technical change and labour absorption in Asian agriculture: some emerging trends’, World Development 14:3, March. Jeffrey, R. (1985) Health and the State in India, Edinburgh, University of Edinburgh PhD thesis (unpublished). Kumar, D. (1974) ‘Changes in income distribution and poverty in India: a review of the literature’, World Development, 2:1, January. Lipton, M. (1968) ‘Urban bias and rural planning’, in P.Streeten and M. Lipton, (eds) (1968), The Crisis of Indian, Planning, Oxford, Oxford University Press for RIIA. Lipton, M. (1976) ‘Agricultural finance and rural credit in developing countries’, World Development 4:7, July. Lipton, M. (1978) ‘Interfarm, inter-regional and farm-nonfarm distribution: the impact of the new cereal varieties’, World Development, 6:3, March. Lipton, M. (1983) Poverty, Undernutrition and Hunger, Staff Working Paper 597, Washington DC, World Bank. Lipton, M. (1983a) Demography and Poverty, Staff Working Paper 623, Washington DC, World Bank. Lipton, M. (1983b) Labour and Poverty, Staff Working Paper 616, Washington DC, World Bank. Lipton, M. (1985) Land Assets and Rural Poverty, Staff Working Paper 744, Washington DC, World Bank.

64

Does Aid Work in India?

Lipton, M. (1987) Improving the Impact of Aid for Rural Development, IDS Discussion Paper 233, Brighton, IDS. Lipton, M. (1987a) Improving Agricultural Aid Impact on Low-Income Countries, IDS Discussion Paper 234, Brighton, Institute of Development Studies. Lipton, M., with Longhurst, R. (1989) New Seeds and Poor People, London, Unwin Hyman. Mellor, J.W. (1976) The New Economics of Growth: a Strategy for India and the Developing World, Ithaca, New York and London, Cornell University Press., Minhas, B.S., Jain, L.R., Kansal, S.M. and Saluja, M.R. (1987) ‘On the choice of appropriate consumer price indices and data sets for estimating the incidence of poverty in India’, Indian Economic Review, 22:1, January-June. Mitra, A. (1978) India’s Population; Aspects of Quality and Control, vol. 1, New Delhi, Abhinav. Moore, M. (1984) ‘Institutional development, the World Bank, and India’s new agricultural extension programme’, Journal of Development Studies, 20:4, July. Narain, D. and Roy, S. (1980) Impact of Irrigation and Labour Availability on Multiple Cropping (India), IFPRI Research Report 20, Washington DC, International Food Policy Research Institute. OED (Operations Evaluation Division) (1981) Agricultural Credit Projects: a Review of Recent Experience in India, Report 3415, Washington DC, World Bank. Plocki, J. and Blandford, D. (1977) Evaluating the Disincentive Effect of PL 480 Food Aid: the Indian Case Reconsidered, Ithaca, New York, Cornell University Press. Rath, N. (1985) ‘Garibi hatao: can IRDP do it?’, Economic and Political Weekly 20:6, 9 February. RBI (Reserve Bank of India) (various years), Report on Currency and Finance, Bombay. Repetto, R. (1979) Economic Inequality and Fertility in Developing Countries, Baltimore, Maryland, Johns Hopkins University Press. Saith, A, (1981) ‘Production, prices and poverty in rural India’, Journal of Development Studies, 17:2, January. Sen, A.K. (1981) Poverty and Famines: an Essay on Entitlement and Deprivation, Oxford, Clarendon Press. Sen, A.K. (1983) ‘Development: which way now?’ Economic Journal, 93: 372 December. Subbarao, K. (1987) ‘Interventions to fill nutrition gaps at the household-level: a Review of India’s experience’, mimeo, Delhi, World Bank. Sukhatme, P. (1978) ‘Assessment of adequacy of diets at different income levels’, Economic and Political Weekly, 12:31, 3 August. Vyas, V. (1979) ‘Structural change in Indian agriculture’, Indian Journal of Agricultural Economics 1. Wade, R. and Chambers, R. (1980) ‘Managing the main system: canal irrigation’s blind spot’, Economic and Political Weekly XV: 39 September. Wood, G. (1984) ‘Provision of irrigation assets by the landless: a new approach to agrarian reform’, Agricultural Administration, October. World Bank, (1975,1982) Land Reform, (policy paper), Washington DC. World Bank (1978) Report on Consortia and Consultative Groups, Washington DC. World Bank (1980,1984,1985,1987) World Development Report, Washington DC. World Bank (1982) Focus on Poverty, Washington DC. Zachariah, K, and Kurup, R. (1984) ‘Determinants of fertility decline in Kerala’ in T.Dyson and N.Crook (eds) India’s Demography: Essays on the Contemporary Population, New Delhi, S.Asian Publishers.

Chapter 3 Policy Dialogue

Surprising facts and aid processes The facts about aid and India’s economic performance since Independence, as we have shown, are fairly clear. But they are also counter-intuitive. Most ‘intelligent laymen’— even most economists not specializing in development—probably believe: (i) that aid to India has been large; (ii) that India’s growth performance has been bad; (iii) that this is partly due to substantial, and rather successful, policy emphasis on equity and poverty alleviation at the cost of growth; and (iv) that aid played a major role in both the growth outcome and the equity outcome. The truth on the first three propositions is almost exactly the opposite: (i) aid to India was smallish relative to GNP, investment, or population—especially by comparison with most other LDCs; (ii) India’s growth performance, by comparison with its own past or with most other developing countries and regions, has been acceptable, consistent, and above all improving, despite the increasingly difficult circumstances of 1973–86; (iii) despite real policy efforts, however—and although equity (e.g. in access to schools or land) normally improves growth—income distribution has worsened, and poverty did not become less prevalent in 1961–86. The truth in regard to proposition (iv) above is more complex. Aid has usually been too small and diffuse to be the main factor underlying India’s growth or equity performance. Yet, as Chapter 2 suggests, its impact on policies for some sectors, and on some macropolicies also, has been crucial. While part of this impact can be related to numbers— gross aid was and is not small compared to the Indian Centre’s public-sector development investment, or to its foreign-exchange requirements—most cannot The story of the 1965 devaluation, or more happily of the policy adjustments (described on pp. 93–4) cannot be told without explaining the changing processes by which the Government of India and the donors jointly determined both the uses of aid and its policy surround. From bilateral leverage to multilateral sector dialogue Until the mid-1960s, the relationship between India and aid donors seemed to be smooth and successful. The Planning Commission was using a small but significant volume of aid resources, together with public savings, to negotiate, with states and ministries, more or less binding assignments of public investment. This accompanied fairly steady, albeit slow, real growth of GNP per person. India was able to run down the sterling balances, and, together with the first major expansion of aid, these cushioned the shock of the midterm adjustment required to the Second Five Year Plan. The alleviation of poverty, already spelled out as an objective (alongside growth) in the First Plan, was not yet demonstrably DOI: 10.4324/9780203840153-4

66

Does Aid Work in India?

under-fulfilled. With spare land to keep food output growing ahead of population, and with some reserves of foreign exchange, India’s planning problems at this early stage presented an illusion of being readily manageable; bad outcomes could to some extent be blamed on bad monsoons. The model of state-led, industrializing investment was accepted by India and tolerated by the donors. They therefore saw little need to use aid flows to India to exert influence on economic policies there, either by direct leverage involving threats to withdraw aid or promises to increase it, or by give-and-take policy discussions. Moreover, while it is now unfashionable to say so, formal economic planning—which need not, though in India it increasingly did, supersede market mechanisms—played a vital role in this smooth relationship. However, such planning was to be disrupted. This happened largely because of the increasing sharpness of inter-ministerial and inter-state conflicts about scarce foreign exchange. These conflicts exposed the lack of political clout—of interest-group base—of the Planning Commission, especially after Nehru’s death in 1963 was followed by a gradual reduction of the real involvement of the prime minister (who remained, and still remains, formally the chairman). In 1960–5 the balance of payments constraint on growth was tightened by a series of events: exhaustion of the sterling balances; hostilities with China and Pakistan; drought; and the using up of spare land. Also the constraint was made more harmful, and harder to relax, by policy weaknesses, including trade policies hampering exports, and neglect of agriculture. Events plus policy weaknesses undermined the low-key approaches by donors to Indian economic policy, and also, unfortunately, their acceptance that this was made by Indians. Donors no longer had confidence that a policy régime, which they were increasingly supporting with aid, was working. Moreover, the way in which it seemed to be not working—deepening, successive foreign-exchange shortages—meant that donors ceased to be able to ‘send messages’ in relatively noninterventionist ways by selecting projects from the plan to support. Such plainly legitimate influence became no longer feasible because foreign-exchange (and other) constraints made the plan an increasingly unreliable predictor of what GoI would in fact be able to do, and when. A major lesson of the Indian experience is that satisfactory donor-recipient policy interchange—whether as ‘policy dialogue’ or as simple mutual trust—may depend on at least a medium-term public investment programme, if not necessarily on a full-blown ‘Plan’. This suggests that donor pressure in Africa in the 1980s, to ‘wind down’ formal state activity and planning, may militate against donors’ capacity to relate effectively to public policies. It was not just India’s growing balance of payments constraints that doomed the easy aid relationships of 1951–64.1 The growing role of such aid, and the dominance of a single donor, also militated against a low profile. In 1963–4, gross aid utilized was US$1.25bn, or 29.1 per cent of gross public investment in that year (Chaudhuri, 1978, 98–9; Mellor, 1976,152). Over half this aid, and a much larger proportion of the crucial food aid component, came from a single donor, the USA (Singh, 1973,68). If one donor (as a source of finance or as a respected co-ordinator of other donors) becomes so dominant, it is always on the cards that it may seek to turn its views of—or perceived interests in—the recipient’s economic, or foreign, policy into a major determinant of the recipient’s scale or pattern of public investment. Such attempted ‘leverage’ was a major feature of India’s aid scene from 1965 to 1971, when US bilateral aid was suspended (until 1978) following the hostilities between India and Pakistan.

Policy Dialogue 67 India’s Plans by the mid-1960s assumed smooth access to foreign exchange, and hence a smooth inflow of pledged aid,2 especially US aid. If donors were to provide such smooth flows with minimal ‘strings’, they had to be confident of ‘acceptable’ Indian policies, foreign as well as economic. This was especially true of the US, where the Executive faced an annual, and increasingly critical, congressional process of authorization before it could release its aid. The whole process was consistent with respect for India’s sovereignty only so long as the set of policies ‘acceptable’ to donors overlapped substantially with the set ‘acceptable’ to the GoI. Both Indian and donor assumptions were shattered, never fully to recover, by the events of 1965–6, and the subsequent combination of dwindling aid and increased leverage. These events showed that bilateral macro-leverage, whether desirable or not, is unlikely to prove feasible except in conditions of acute distress. Even then, the use of such leverage produces reactions and resentments that poison future aid relationships. The experience has produced three realignments in the 1970s and 1980s. First, leverage has (supposedly) given way to policy dialogue. Second, the pressure of a predominant single donor, the US A, has given way to the concerted action of donors and recipient together within the World Bank Consortium (often supported, as in 1974–5, by IMF approaches). Third, the emphasis by donors on macroeconomic policy has given way to concentration on particular sectors. All three changes have improved the capacity of donors to raise the benefits of aid to India, but have probably not gone far enough. Moreover, a sectoral policy dialogue mediated by a ‘key donor’ such as the World Bank—while it has had major successes—is under some threat. It requires, from the key donor, deep knowledge of India; continuity of personnel both in Delhi and at the aid headquarters; and substantial decentralization of staff to local and project level. These requirements, especially the last, have not been fully achieved, even by the World Bank in India. A Consortium-based sectoral dialogue could also be subverted by the very errors that helped to destroy—in our view fortunately—the older attempts at one-donor macroleverage: by an unduly ideological approach; by excessive rigidity; by shifting fashions; by mistaken selection of topics, or sub-sectors, for dialogue; or, above all, by failures to ‘deliver the goods’. All these problems can affect either donors or recipient. They are detailed later in this chapter (p. 94 top. 113). Finally, donors that seek ‘lessons from India’ for other policy dialogues—notably with African countries—need to recall four special features of the Indian aid relationship. First, clear and project-specific GoI priorities for foreign exchange allocation—initially published in the successive Five Year Plans (Rao and Narain, 1963, 16)—provided until the mid1960s, and to a lesser extent still provide, a framework in which, without interfering in Indian policy-making, donors could signal their own preferences and request responses. Second, India has many trained, skilled public-sector personnel at various levels, who can interact with donor personnel—teaching at least as much as learning—in sectoral policy dialogues. Third, both India and the key donor (the World Bank) have access in most activities to a supportive, open and critical network of professional and academic analysis, operational review, and basic research, both inside and outside official institutions.3 Where these conditions are not met, even good donor procedures are unlikely to produce a genuine and sustainable policy dialogue. Creating these necessary conditions is obstructed by some of the reductions in state planning, power, and spending that in the 1980s have been called for by donors in their dealings with some African countries (World Bank, 1981).

68

Does Aid Work in India?

The donors’ current approach to Africa—more, and more orderly, aid, to be funnelled to or by smaller and less powerful governments—risks inconsistency. Without the proper environment for aid management and policy, and specifically without the three conditions listed above, more aid to, and more effective policy dialogue with, smaller governments is certainly inconsistent, and a wrong lesson to draw from the Indian experience. A fourth special circumstance of the Indian aid scene, affecting the scope for policy dialogue, is the sheer volume of aid. If gross aid utilization is financing very large parts of gross public investment—as in India in the mid-1960s, and even more in sub-Saharan Africa today4—the performance of aided projects as a whole becomes critically dependent on ‘good’ domestic policies.5 Hence the inducement to donors to seek ‘leverage’ on macropolicies becomes extremely strong (although the responsive capacity of governments is often weakened at macro-policy level by very large aid involvements, which can involve the diversion of domestic personnel into managing aid projects) (Lele, 1979, 1987). If the proportion of gross public investment financed by gross aid utilization drops below about 25 per cent, as has been the case in recent years in India, donors lose both power over and interest in macro-policies, and must shift from macro-level policy dialogue. But to what should they shift: projects, areas, or sectors? (i) In many countries donors have identified large ‘island projects’ for aid. These have been outside the purview of general administration, yet have sucked resources from it. Wisely, GoI has seldom allowed this. (ii) Also, perhaps less wisely, GoI has left little scope for donor dialogue with particular states. It has instead centralised aid negotiations at Central Government level. (iii) Thus a donor to India, seeking to concentrate policy dialogues associated with scarce aid, has had little choice but to conduct these dialogues about particular sectors. Such sector dialogue is made much easier by a fairly long-term relationship, in which mutual respect is built up between specific, experienced sector experts from the donor and the recipient sides. Where the volume of aid is large (e.g. from the Bank group), or where the donor concentrates heavily on a sub-sector and/or allies itself with other ‘like-minded donors’ (e.g. the Scandinavians on health aid), donors are likely to make the necessary long-term full-time appointments. Such long-term personnel, especially in conjunction with corresponding Indian experts, can achieve a lot in a defined sector—but the process carries risks (see pp. 98–9). Such sectoral concentration is a sensible outcome of the search for effective aid, and in India is almost unavoidable now that aid has shrunk, given the GoI’s dislike of island or regional concentration. However, donors tend to press for macro-policy conditions if aid is large and outcomes seem unsatisfactory (see note 5). This is so especially when aid goes not to projects but to general balance of payments support (Examples are food aid to India in 1955–70 or to much of Africa since the mid-1970s; and general-purpose, non-project aid to permit the maintenance of current imports in much of Africa since 1980). How is a donor, in these circumstances, to avoid improper—and often politically naïve—leverage, while safeguarding a proper interest in the good use of non-project aid?6 Unless one or two key variables are so seriously out of alignment as totally to subvert development in a sector—quite a rare occurrence7—sectoral dialogues are probably the most fruitful approach developmentally. In Africa, the World Bank’s switch since the mid1980s, from macro-policy conditions (in ‘structural adjustment lending’) towards ‘sectoral adjustment loans’ (Rose, 1985) renders the approach even more plausible. However, sector

Policy Dialogue 69 dialogue may be less appealing politically, especially for bilateral donors—who, after all, fund the multilateral agencies. Such donors are tempted to use aid to impose overall political ideology (e.g. market capitalism or centrally planned socialism), or to buy political or commercial favours by a high-level project presence. In the absence of ‘India-style’ constraints of aid size (gross aid utilization below, say, 25 per cent of public investment), will big donors yield to the temptation to switch back, from genuine economic policy dialogue, to macro-leverage with overtly politicoideological components? In the absence of ‘India-style’ aid management by central Government, will donors concentrate even genuine policy dialogue not on sectors, but (perhaps less desirably) on support for largescale ‘island projects’, e.g. in areas controlled by local authorities—what has been bitterly termed ‘rent-a-district’—as has happened elsewhere in South Asia, and as is widespread in Africa? Before we return to the sectoral ‘middle path’, we review the history of the alternative approaches to aid policy dialogue in India. Before leverage The coming in of foreign aid has coincided with the starting of economic planning in India; and this has been most fortunate from the point of view of evolving some kind of rationality and order in formulating requests for foreign aid…the Plan provides a measure of the total foreign exchange required and the balance [needed, after expected] export earnings. And it is this gap that…determines the country’s need for foreign aid. (Rao and Narain, 1963, 16).

Each Five Year Plan lists projects, priorities, and expected foreignexchange sources, including donors, who in some cases are yet ‘to be identified’ when the Plan is published. On top of this, when India was in food deficit, expected food aid inflows released GoIcontrolled foreign exchange for Plan projects. Each donor country could select projects to support, and could assess how much support to give to India’s total programme, by project aid or by general programme support (including food aid). In 1963, an authoritative Indian view stated: there is a high degree of mutual discussion…regarding the terms of aid, including projects selected… One does not get the impression that the donor country dominates…nor does there appear to be any atmosphere of tension and strain… While all this is no doubt due in part to the good sense and goodwill of the donor countries, it is mainly due to the fact that India has a systematic and wellformulated programme of economic development through its Five-year Plans…into which fall all the projects for which foreign aid is either offered or asked… Coordination…is secured on a national basis by the Central Ministry of Finance, Department of Economic Affairs, [although] in choosing the projects which they will aid, donor countries are influenced by their own ideological or economic outlook. (Rao and Narain, 1963, 36)

A corresponding view from a donor expert8 was equally sanguine. Rosen (1967, 262–3, 268–9), after describing the US interest in political democracy and stability in India, continues: The US has given the aid on the assumption that Indian planning and policies were of sufficiently high quality and were adequate for the requirements of development, and

70

Does Aid Work in India? therefore American interests in the structure of the plans and in Indian policies for their implementation were relatively slight. (Rosen, 1967)

Rosen, indeed, recommended more active, though modest and diplomatic, policy involvement by the US in India. The best testimony to the non-salience of ‘leverage’ and ‘policy dialogue’—during the period when aid was building up to its greatest proportionate contribution to India’s public expenditures (Mellor, 1976, 152; Chaudhuri, 1978, 99)—comes from Tarlok Singh (1974, 310–40). In his persuasive discussion of ‘External Resources and Self-reliance’, the Deputy Secretary of the Planning Commission (1950–66) and a major author of the first four Five Year Plans made no mention of leverage or policy dialogue whatever. Later he makes rather cryptic remarks about the events of 1965–6, to which we return below. But plainly the interesting fact about leverage and macro-policy conditionality, from 1960 until late 1964, as with Conan Doyle’s Silver Blaze, is that the dog did not bark in the night. A justly proud recipient, India, and a justly concerned (and increasingly predominant) single donor, the USA, managed, in the 1950s and early 1960s, without overt and explicit policy conditions. Yet these proved deeply offensive in India after 1965 (so that explicit sectoral substitutes were evolved), and for most of Africa in the 1980s have been supposed by major donors to be almost essential. Why the peaceful times in India before 1965? Part of the reason, as Rao and Narain state and as Rosen implies, lay in the integration of aid into a functioning series of Plans; the loss of this mechanism certainly sharpened conflicts and concerns about leverage. The formal arrangement remained—and still remains—as described by Rao and Narain (1963, 16), but the substance is much less. The Draft Fourth Plan (1968) represented an assertion of older methods and priorities against the ‘deal’ negotiated with the donors in 1965–6, but the Commission could not carry the chief ministers of the states. Three years of ‘plan holiday’ (Annual Plans) followed. The Fifth Plan was knocked sideways by the 1973 oil price explosion; the Commission’s refusal to recognise the implications (more taxes or a smaller Plan) led to a public row, to the resignation from the Commission of a disinterested and perceptive economist (Dr B.S.Minhas), and to considerable loss of public confidence in the Commission. Meanwhile—although the prime minister retained the formal chairmanship—Mrs Gandhi appeared to downplay the Commission’s importance, and seldom attended its meetings (as her predecessors had done frequently). Donors, too, weakened the Commission by public utterances that increasingly appeared to equate planning with misplaced economic interventionism. When, as often happened, Plans had to be pared to the core in order to meet foreign-exchange crises, that core was determined largely by pressures from chief ministers (of states), secondly by line ministries (at the centre), and least by Planning Commission priorities. During the 1970s the growth of a ‘parallel economy’, fuelled by money from tax evasion and other illegal activities, reduced the role of planners (Jha, 1980, 45–60). During Rajiv Gandhi’s premiership, serious attempts have been made to reduce the political factors (associated with the financing of the Congress Party) leading to this last problem; however, the Commission’s authority has been further damaged by the government’s felt need to make large investment commitments, not included in the Plan, in public speeches not previously discussed with the Commission.

Policy Dialogue 71 The framework of 1950–64, in which the Indian Plans enabled policy dialogue to take place as an informal and relatively uncontentious contraposition of donor preferences with a clearly stated set of GoI priorities, has therefore never been fully restored. There remained a set of projects, foreign-exchange assignments, and priorities, formulated in a Five Year Plan. But the authority of that Plan had dwindled.9 Hence neither donors nor recipients knew whether Plan projects—to which donor money might have been assigned—would, in time of pressure, retain the support of complementary domestically-funded outlays. Since neither individual donors nor the Consortium could rely alone, or even mainly, on the Plan, a less project-related and more apparently interventionist dialogue was bound gradually to supervene, even apart from the traumatic events of 1965–71. This suggests that formal planning has clear advantages for donors even if they are highly market-oriented. In the acrimonious debates of 1965–71, US critics were too ready to identify India’s ‘big’ Plans, a big state, capital-intensive industrialization, inwardlookingness, a preference for ‘material balances’, and sympathy for Soviet-type planning. Yet all these are separable, even if they overlapped in some key Indian persons of influence. A Plan, as a modality for ordering public (and predicting prioritized private) activity, always eases the task of reconciling donor and recipient priorities. Such a Plan can involve a small or big, shrinking or growing, private sector. However, the decline of India’s Plans was not the only reason why the ‘golden age’ of relaxed, yet clear and specific, policy dialogue gave way to the mutual abrasiveness of 1965–72. The underlying reason was that India, lacking much scope for bringing new land into cultivation, or for low-cost irrigation, was faced with severe wage-goods (food) constraints on growth,10 and with near-crisis when, as in 1965–6 and 1966–7, monsoons were bad. On top of the growing foreign-exchange costs of food imports, India’s hostilities with China (1962) and Pakistan (1965, 1971) had major foreign-exchange costs, immediate and long-term, further restraining the planned rate and path of industrial growth. Meanwhile, some of GoI’s methods of licensing, allocating and rationing, especially for foreign exchange, induced inefficiencies, actual and incipient corruption, and diversion of GNP into burgeoning ‘economic rents’ for those lucky enough to own the capacity to make something in heavy demand but legally limited supply. This diverted GNP, from risk-takers towards possessors; business energy, from seeking profits towards obtaining licences to corner protected markets (Krueger, 1974; Papanek, 1971); and (by encouraging the production of non-tradeables, and secondly of protected industrial products) resources, away from the agricultural sector, already so crucial a constraint upon growth, and increasingly constrained by land shortage. In short, the period of cheap land, cheap post-colonial catching up, and steady growth was by 1965 over. Its ending cruelly exposed the extent to which GoI policies rationed the newly-scarce resources inefficiently. For the donors (increasingly dominated by the USA), it was less and less possible to feel that India’s democracy and firm government sufficed to justify its ever-growing aid. The honeymoon period, with informal—or no—policy dialogue, was over. Towards the crisis of bilateral macro-leverage Early signs of danger, for an aid budget increasingly dominated by the bilateral Indio-US aid relationship, are recorded by Rao and Narain

72

Does Aid Work in India? Since this study was completed, a series of events…between aid-giving and aid-receiving countries [created] tensions… Certain aspects of Indian action and policy…—the Goa action, the Kashmir dispute, India’s defence expenditure, the [purchase of Mig fighters from the USSR, and statements by Foreign Minister] Krishna Menon…—led to some strong speeches in the [US] legislature which appear to encroach on India’s sovereignty and freedom of action…the Senate Foreign Relations Committee voted a specific cut in the proposed US aid to India… [Although] the Kennedy administration eventually succeeded in getting…the bulk of what it had [requested for India], initial reactions were somewhat intemperate in character. [But] subsequent action both in the US and in India showed considerable maturity which nipped in the bud the growth…of the tension. (Rao and Narain, 1963, Epilogue, 97)

Unfortunately, a better and gloomier guide to the future of the Indo-US aid relationship (and of bilateral macro-dialogue in aid generally?) was Eugene Black’s farewell speech as President of the World Bank (cited in Rao and Narain, 1963, 98): ‘Aid which is at the mercy of the variable winds of diplomacy offers a poor basis for the rational programming of economic development’. The existence of alternative donors, with different ideologies, did not suffice—despite the hopes of Rao and Narain (1963, 73)—to keep post-1964 bilateral, above all Indo-US, aid dialogues respectful of national sovereignty, and concerned mainly with national economic policy rather than with international politics. This is the main theme of the ‘crisis of macroleverage’ in the mid-1960s. Before we return to it, a few words are needed on a parallel source of tension: the use of aid-linked political relationships to divert economic aid to the commercial advantage of the donor. Bilateral aid, tied by the donor to the supply of otherwise uncompetitive products, seldom revives those product lines in the donor country, (Mosley, 1986; Toye and Clark, 1986)11 and cuts across the effective use of policy dialogue to increase the economic returns of aid to the recipient. The Bokaro steel plant typified: the conflicts between the Indian need to develop and utilise indigenous design and the interests of potential foreign financiers of the plant, often paralleled by the influence of controlling politicians and diplomats. By 1962, India [could] design steel mills, particularly in Dastur and Company. But no major aid provider, whether the USA or the USSR, would fail to insist that the design be made by its own nationals. This insistence was justified, not by the technical assistance needs of India, but by the political and economic requirements of the donor. (Mellor, 1976, 229; cf. Desai, 1972).

Not all tied bilateral aid is wrong;12 the commercial demands of ‘tied’ exporters may even deter donor governments from otherwise distorting policy dialogue, via ‘ideological’ or foreign-policy conditions for aid. However—especially in the area of consultancy, technical assistance, and choice of project techniques—bilateral tied aid can induce donor governments to apply criteria inconsistent with genuine policy dialogue.13 There are, then, two underlying (if, happily, conflicting) dangers of bilateral leverage: its use for commercial advantage and for international diplomacy. These are dangers, not because it is wrong to use aid to secure gains for the donor, but because these aims (i) do not work, and (ii) subvert the usefulness of aid to the recipient.14 Even if the donor genuinely believes that its leverage seeks only to increase recipient benefits from aid, it must

Policy Dialogue 73 take extreme care to respect recipient sovereignty. Recipients may define ‘effectiveness’ differently from the donor, or may have different views of the politically feasible options.15 Even where the topic for bilateral pressure is well chosen, there are serious risks in concentrating on a narrowly-focused ‘ideological’ approach (see pp. 99–102). The outcome was well described by Tarlok Singh: [We needed in 1965–6] a careful analysis of the basic weaknesses of the economy, the assumptions which had gone wrong…reorientation of…policies, and economic discipline. A democratic political system could have posed the issues openly in these terms. Instead, influential criticism from abroad focused on one element, the overvaluation of the rupee, as the prime cause of current difficulties. Encouraged by informal assurances of greater external support, the GoI undertook a hasty and ill-prepared devaluation under exceptionally adverse circumstances. (T.Singh, 1974,350–1)

The aid environment, the policy environment, and the nature and concentration of bilateral leverage combined after 1964 to create the conditions for a ‘crisis of leverage’. The aid environment involved India’s growing reliance on aid, and the growing concentration of that aid on one donor, the USA. The policy environment involved—in the context of the exhaustion of India’s supply of readily cultivable and unused land, and of tightening growth of foreign-exchange bottlenecks associated with food supply and capital- (and import-) intensive industrialization—policies discouraging exports, creating economic rents, and persisting in import-intensive growth paths. Bilateral leverage involves dangers of diplomatic (or undiplomatic) pressures based on international politics, of commercial rather than economic interventions, and of over-simple, insensitive ideologizing. Yet these three conditions might not have produced a crisis of bilateral macro-leverage, but for the events of 1965–6: war with Pakistan, a disastrous monsoon, foreign-exchange exhaustion, and hence absolute dependence by India on relief from a single powerful nation. In Summer of 1965, the US A: as a sign of displeasure at the waste of scarce resources in an unproductive war, had indefinitely suspended all aid to both India and Pakistan [and] refused to sign a fresh long-term agreement under PL 480 when the existing agreement expired in August 1965. Instead, the Johnson administration, egged on by Congressional opposition to concessional food aid for India, adopted a ‘short tether’ policy of doling out stocks sufficient only to meet requirements a few months at a time, and explicitly tying the continuation of food aid to the adoption by India of policies aimed at increasing agricultural production and curbing population growth. (Frankel, 1978, 287).

The then head of USAID in New Delhi, J.P.Lewis, spoke of ‘specific aid offers contingent upon the institution of particular adjustments in indigenous rural policy’ (cited in Chopra, 1968). We thus have two elements of bilateral leverage: tie-up with international politics (the hostilities with Pakistan) and conditionality of aid on specific macro-policy changes. Later in 1965, two further contentious elements were added. With Indian food reserves and foreign exchange almost exhausted, India effectively submitted to foreign macro-economic policy review: in a ‘meeting in Rome between [India Agriculture Minister] Subramaniam and

74

Does Aid Work in India?

US Secretary of Agriculture, Orville Freeman, the specific policy proposals were revised item by item’ and pushed through Cabinet ‘despite the vehement objections of the Finance Minister’. Second, this review involved the inclusion of items commercially favourable to donors. The Freeman-Subramaniam review ‘includ[ed] the plans for incentives to foreign private investment, especially in fertilizers’, and shortly afterwards GoI ‘announced a new policy of concessions to foreign private companies willing to invest in the fertilizer industry in India’ (Chopra, 1968, 287). We are not at all arguing that it is unreasonable for a donor, if even-handed, to refuse aid that finances recipients to wage war on each other; or to insist that, if more food production (as a wage-good) is necessary for progress, the recipient uses resources to that end. All the above pressures (except the commercial one) may well have been perfectly proper, and their outcomes largely desirable for India. But the manner of their effective imposition upon a proud and large country, by a single donor not obviously representing the international community, in the wake of aid suspension in a war and ‘short-tether’ food relief in a drought, led to ‘anger and humiliation’ (Patel, 1986,73) and probably doomed bilateral macro-leverage as a route to policy dialogue for India’s aid donors. The cap was put on this by the devaluation saga of 1966 and its aftermath. The above donor interventions—while in our judgement they improved Indian economic policy, not least in its impact on the poor—involved major concessions of Indian economic sovereignty; yet, as regards aid inflows, these concessions had restored no more than the status quo ante and that only with respect to food aid. ‘A final condition for the resumption of large-scale foreign [i.e. US capital] aid had still to be met, that of devaluation. [T.T.Krishnamachari, the Finance Minister] still refused to consider such a step’. L.B.Shastri’s apparently sudden decision16 to set up a commission of enquiry to investigate the longstanding allegations that Krishnamachari had abused his office to advance his sons, was certain to cause ‘TTK’ to resign in protest, and was widely seen in India as a bridge to devaluation. Despite TTK’s negotiations with the World Bank (pp. 93–4) and the apparent agreement, it soon became clear that the key US ‘aid would be forthcoming only after India had acted on her earlier promise to devalue’. This happened on 6 June, 1966, despite major dissension in Cabinet and AICC. US aid resumed ten days later (Frankel, 1978,228, 297–8). This was bound to look like extreme bilateral leverage, at the borders of disrespect for sovereignty. Had devaluation been seen to work, such leverage might just have proved acceptable in India. However, price-inelastic supplies of key exports (or of inputs to make them) meant that ‘the immediate impact of devaluation proved unfavourable. Although mining and manufacturing exports did increase, stagnant or declining sales of traditional exports produced an overall decline of real foreign exchange earnings of about 8 per cent in 1966–7’ (Frankel, 1978, 322). Nor were the pledges of aid, especially those by the USA, made to secure the devaluation/policy reform packages, in fact met (see p.112). Thus GoI had been subjected, in crisis, to extreme bilateral pressures—albeit, in many cases, pressures to do sensible things;17 had bowed to such pressures; and had found that neither the policies adopted, nor the pledged aid, produced real foreign-exchange inflows to compensate for the wounding realities of bilateral leverage. Although the US use of food aid in 1966 ‘worked’, in the sense that it extracted (probably desirable) Indian policy changes, the process killed bilateral leverage by revealing, not only the uncertain quid (or dollar) pro quo, but also its potential for politicization and abuse. Indeed, in 1971 (during

Policy Dialogue 75 the events leading to the independence of Bangladesh), the USA once more suspended economic aid—this time in a purely political context, for seven years, in the wake of an avowed ‘tilt to Pakistan’; but Indian policy was not changed as a result. Aid 1965–85: towards multilateral sector dialogue The temporary withdrawal of the US from bilateral aid to India in 1965–6, the lasting wounds left by ‘leverage’ from that period, and the long US absence in 1971–8, rendered bilateral, aid-based approaches to macro-dialogues with India almost infeasible. No one donor looms sufficiently large. At least one of the largest, the UK, in principle prefers to leave such policy dialogue to the Consortium, and is increasingly prevented, by its emphasis on tying and aid-trade linkage, from entering usefully into such dialogue without meeting commercial objections from exporters. Anyway, a growing proportion of aid to India is multilateral; total gross bilateral aid to India, since 1970 typically below 0.7 per cent of Indian national income, would, even if co-ordinated, be too small to exert major influence on macroeconomic policy. Also, there are more positive reasons for the move towards sectoral dialogues, with donors’ wishes mediated through the World Bank. The Bank has in most years been much the largest donor (via IDA) of concessional funds to India—its pledges in 1981–2, for example, totalled 1307 (Rs crores), or 60 per cent of total OECD aid to India (GoI, 1982–3, 271–2).18 The Bank’s overriding role has increased as other donors, especially the US, became less important. Bank aid has also been more reliable than that of some donors and, through IDA, more concessional than that of most. Through its Delhi office and Indiaspecific staff in Washington, the Bank has many more specialists with long-term expertise and links to Indian experts, in many fields, than any one bilateral donor (probably even than all combined). The presence of many senior, Indian passport-holding Bank staff (in at least one case moving between a senior research position in the Bank and senior policy responsibility in GoI), and of an Indian Executive Director at the Bank (usually a retired senior civil servant), at once strengthens and eases the Indo-Bank relationship, in a way not feasible for a bilateral donor, which can hardly appoint senior Indians to domestic policymaking positions. Hence many bilateral concerns are expressed through the Bank as Consortium chairman. This had happened already with the Bell Mission in 1964–5. Then, however, several factors—US predominance in bilateral flows to India; Bank liberalizing pressures that seemed to some Indians to take the US side against them; Bank wishes to release publicsector resources for agriculture by re-balancing industrial activity more towards the private sector; a perceived conflict between Bank agricultural strategy of ‘progressive farmers’ in irrigated areas and stated Indian perceptions of regional and vertical equity— prevented the Bank from adopting the low-key, give-and-take role that it required (and later sought) for effective sectoral dialogue. Most seriously, the Bank and the Consortium were seen by leading Indians as linked with one overpowering donor, the USA (both in seeking to influence Indian macro-policy excessively and perhaps ideologically, and subsequently in not succeeding in mobilizing the aid promised as quid pro quo), partly because of high sensitivities on both sides, but mainly because of the 1965–6 crisis. The Bell Mission— while advocating many sensible measures, most of them later adopted independently by GoI—was in some respects a ‘false start’ for multilateral sectoral dialogue (Frankel, 1978, 271–2; Rosen, 1985, 107).

76

Does Aid Work in India?

In the post-1971 situation of many smallish bilateral donors, with the ring held by the World Bank as major donor, even bilateral leverage has tended to be sectoral, tactful, and mediated through the World Bank. For example, the US—which since 1979 has resumed project aid to India at some US$90–100m yearly—has, in preparatory documents discussed at Consortium meetings, pressed for import liberalization, export development, and preference for the private sector. In February 1983 the Bank, US AID and ODA (UK) jointly surveyed training needs for social forestry, and recommended several private initiatives, with the Forestry Department as ‘catalyst rather than provider’; these proposals were favourably reviewed in Delhi, itself worried about the expansion of the costs of the Indian Forest Service, The US, mainly via the World Bank, has also encouraged India towards diversified private-sector provision of contraceptives. Such initiatives, because sectoralized and Bank-mediated, have not provoked crises or ideological conflicts. All is not sweetness and light. Despite evaluations suggesting high economic rates of return to aid for agricultural credit in India (World Bank, 1981a), the USA has stopped aiding NABARD,19 mainly because of objections that interest-rate subsidies make for a low (or even negative) financial return. The USA has (until 1989 unsuccessfully) pressed the Bank and other donors to do likewise, apparently because conclusions about the scale and inefficiency of subsidized rural credit, partly valid in Latin America, are being applied to India. The US belief in an almost exclusively private-sector role in oil exploration and development led them in 1983 to block the proposed Bank ‘energy affiliate’—which could have been of great value to India—and, on some accounts, has led the Bank to overstate, to donors, the private-enterprise preferences conveyed to India and others in policy dialogues about energy. However, it is in general correct to say that bilateral aid relations with India have been improved greatly, in respect of their capacity to ‘carry’ policy discussions without tension, by their mediation through the Bank. This is not at all because the Bank has systematically co-ordinated its policy dialogues with other donors. This has taken place only where bilateral expertise and cash, complementary to the Bank’s, have flowed—with the Nordic countries on family planning, with Britain on coal. But India has tried to reduce the number of multi-donor projects, where timing and procedural wrangles impel the sort of ‘consultation’ that can easily turn into ‘ganging up’ against the recipient. In most sectors the Bank, too, has carefully avoided the appearance of such ‘ganging up’. The difficulties of World Bank-mediated policy dialogue are different and more fundamental. The first difficulty is the growing pressure inside, and upon, the Bank to remove increasing parts of aid to India from IDA terms. The USA, and to some extent the Bank itself, feels that low-income African countries—which, unlike India, have no prospects to borrow commercially—are priority candidates for ‘soft’ IDA terms. However, the returns to capital aid seem far higher in India than in most of Africa (Papanek, 1972; Mosley, 1987). Most African countries, it could be argued, require research, training, and technical assistance—some of these, perhaps, from India—before large increases in capital aid can work. Aid cuts for India, moreover, seem a sorry reward for steady growth, good aid administration, democracy, and persistent and extreme mass poverty. If, as is envisaged, Bank funds for India are increasingly to flow on IBRD terms, then an increase in the scale of those funds will not be as much of a compensation to India for the decline in IDA money—nor, therefore, as much of a help in making the transition

Policy Dialogue 77 to more liberal policies20—as both India and donors might wish. This is because—while loans from ID A are so concessional as to be worth over 85 per cent of a grant of the same amount—the IBRD rate is now not far below commercial bank rates available to India. If the Bank, its members, or other ‘donors’ increasingly push India from near-grant to nearcommercial terms of capital transfer, then—apart from possibly incurring quasi-Brazilian risks in the long run (see Chapter 1)—they inevitably, and greatly, reduce their power to persuade, or to help, India to improve its economic policies, and hence the effectiveness of capital transfers. This is especially important in view of the growing recognition, in GoI and at the World Bank, that the structure, extent, and quota-orientated nature of Indian import restrictions still severely harm both efficiency and equity—but that, if the changes (on which GoI has long embarked) are to accelerate, supportive aid expansion is needed, both to afford the transitional import increases and to help ‘compensate the losers’ and thus increase the political feasibility of liberalization (Bhagwati and Srinivasan, 1975, 245; Sukhatme, 1983, 34–6; Krueger and Ruttan, 1983, 4–8 to 4–9). The continuing decline of IDA, and hence of World Bank aid, for India is, in view of this growing consensus, most unfortunate. The second difficulty arises from the changing images of the World Bank. It has always favoured efficient use of market incentives in development. However, from President McNamara’s ‘Nairobi speech’ in 1973 to his retirement in 1981, the Bank was increasingly seen as liberal, expansionist both in its own lending and in its view of government, and poverty-orientated; under Presidents Clausen (1981–6) and Conable (1986-), the image has become blurred, partly by US pressures on the Bank, partly by its own frequent reorganizations, and partly by public and apparently ‘ideological’ conservative postures in Bank research and, in Africa, policy dialogue. Such documents as the 1981 ‘Berg Report’ (though not the follow-ups on African policy), and the 1986 and 1987 World Development Reports (Lipton, 1987; Singer, 1987), while not reflecting Bank-wide consensus, did sound an unfortunate note of dogmatism as regards the role of government and the sufficiency of price policy. The reality of Bank actions—whether in loan operations (country-desk or project-officer level) or at the top (in formal policy statements (e.g. World Bank, 1986a) or Presidential commitments)—has changed far less than the image. For example, Bank Group lending was only slightly more concentrated on poverty-related sectors or countries in 1973–81 than before or since (Beckmann, 1985), and the poverty focus has been frequently re-emphasized by McNamara’s successors, not only (though also) by almost every Annual Speech to the Governors. Most recently the Conable Bank has embraced ‘adjustment with a human face’. Yet there is a persistent, nagging feeling among outsiders and aid recipients, including India, that the Bank, in the course of pressing for lower levels of government involvement in several parts of the economy (and of removing barriers to private-sector development), is downplaying poverty issues. Since the 1973–81 Bank positions—expansionism, poverty-focus, etc.—were in close accord with GoI’s official stance, the apparent blurring of such positions was bound to impair prospects of smooth policy dialogue, mediated by the Bank, at sector level The apparently blanket formulae— devalue, deflate, decontrol—applied by donors in Africa, were in reality due more to the Fund than to the Bank, and were in part inevitable, given past African policy errors, worsening terms of trade, and droughts. However, given the increased co-ordination after 1980 between the Fund and the Bank, the smell of dogmatism affected the Bank as well— and reminded Indians of the macro-leverage of the later 1960s.

78

Does Aid Work in India?

There are thus two threats to sectoral, Bank-mediated dialogues with India: a Bank image on occasion dogmatic or blurred; and reduced Bank emphasis on Indian lending via IDA. Such threats are to be regretted for several reasons. (i) Sector dialogues have produced some desirable outcomes. Urban priorities have moved from so-called ‘low-cost housing’ (for middle-level officials) towards slum upgrading for the poor. In fertilizer production, electricity distribution, and elsewhere, the dialogue has also been fruitful.21 (ii) The Bank’s own formulation of its positions in sector dialogues involves a useful learning process. (iii) Above all—while content in development policy normally matters more than form—the most important lesson from sectoral Indo-Bank dialogues lies in the procedures evolved. These help Indian policy-making and analysis. For instance—even if T and V extension or rotational irrigation have flaws or limitations—Bank finance to test such models, in prolonged consultation between Bank and Indian experts, stimulates countermodels, and debate about just when each approach works or fails (e.g. Moore, 1984; Feder and Slade, 1986). Such procedures need (a) Bank money, and (b) free debate within India. As for (a), a leading Bank irrigation specialist wrote in April 1984: I cannot imagine a situation where the Indian irrigation establishment would have been prepared to listen to expatriate advice if this advice, at least in the initial stages of our work, would not have been a condition of our financial contribution to the programme. Today I would guess that some of our technical advice would be accepted even if it would not be concerned with specific lending operations. [Acceptance of our expertise in] data evaluation, training of engineers, training of diploma holders, groundwater hydrology, and even road planning, design and construction have become independent of our lending programme.

This raises a second-generation worry. If bilateral macro-leverage was too invasive and abrasive, has multilateral sector dialogue been too cosy: too much a mutual acceptance, modification, and convergence, through learning, of the assumptions and codes of, for example, the irrigation establishments in the World Bank and the GoI? Do Bank and Indian engineers tacitly agree: to overlook the problem of corruption; to overstress technical fixes, and under-state the need for appropriate, regionally distinct, institutional change; to train at too high a level, neglecting ‘the 80 per cent of non-professional staff who determine day-to-day management of secondaries (Wade, 1982, 181)? Analogous risks apply in other areas of Bank sectoral specialism. Some of the lessons for policy dialogue are discussed below. But the central point is that major experiments, initially financed by aid and then openly evaluated and debated, are the essence of successful dialogue. Free, lively and informed debate in India, notably in the Economic and Political Weekly, has been as necessary to good sectoral dialogue as is some experimental use of cash by donors and GoI alike. Uma Lele and George Rosen both contrast ‘absence of foreign economists and vigorous discussion of agricultural policy in [Africa] today with the situation in South Asia’ around 1960–70 (Rosen, 1985, 231). We doubt the clear linkage between ‘foreign economists’ and free and informed domestic debates on economic policy, but the latter are probably essential for useful ‘policy dialogue’ with the former. Freedom of debate on economic policy—except to some extent during the 1975–7 Emergency—has prevailed in India; but a cosy, top-down style of sectoral dialogue might muffle such a debate, or listen to it too late.

Policy Dialogue 79 There is also a worry about how far the Bank—as its lending emphasis on India, and perhaps its total real resources, decline—will choose to maintain in India (not just in Delhi) expert staff in ‘critical mass’ and with individuals assigned for periods long enough to become really familiar with the sector and with Indian experts (and structures), and senior and assertive enough to carry weight with the Bank HQ in Washington DC Junior line staff are often prone (i) to over-interpret signals, real or assumed, about changing priorities in management or policy22 and (ii) to give priority—in accordance with bankingstyle career signals, in this case real and not imagined—to ‘getting the money moving’ over getting the project and the policy agreed and right Yet a strong case does exist for more decentralization from ‘aid HQ’ to senior staff in the Third World. This has been to some extent recognized by USAID in increasing the local cash limits up to which local staff may authorize projects, and by the UK through its Development Divisions located in such capitals as Lilongwe and Bangkok, and each responsible for decisions for several countries. Several donor agencies told us that experienced local staff acted as invaluable ‘buffers’ against swings of ideological fashion in donor HQs, keeping projects on course and preserving continuity in sectoral dialogue. While accepting this in theory, donor HQ in practice often enforces its own positions, not those of decentralized country offices, in policy dialogues. Sometimes this means pushing through central ‘standards’, even when inappropriate. Sometimes, donor HQ rightly sees the need for ultimate policy overview when the donor’s local staff, keen to push projects through, finesse central policy guidelines. In one striking example—a very large irrigation project under construction in India—local donor staff failed to clarify the compensation of tribals for loss of unregistered land rights in the huge areas to be inundated by the project. Only prolonged campaigning by the agency’s sociologists at HQ eventually forced a better outcome. However, in general, donor HQ—like a Foreign Office, fears that senior personnel, based long-term in (for example) India—side with Indian officials against donor agency advice, and are seen by such officials as ‘not really meaning it’ when they convey agency doubts or criticisms. So donor HQ does assert itself, sometimes rightly. Also, continuity of senior donor experts is less useful if recipient contacts are often changed, posted, etc.—a more serious problem in Africa, but not unknown in India. And a close, lasting relationship among experts can sometimes degenerate into buddy-buddy establishmentarian setting of a static ‘hidden agenda’. Despite such risks and conflicts, the criticisms of Bank positions in policy dialogue, discussed below and in Chapter 7, illustrate the need for more, more expert, more senior, more continuous, more project-localized Bank expertise in India. Pressure in 1988–9 to cut Bank project management costs may waste, not save, money. The problem of ‘ideology’ in policy dialogue ‘It is important that US goals in India23 not be framed in a narrowly ideological fashion…. The ideological interest of the US is well satisfied by the fact that India is a reasonably effective working democracy… [The] amount of State action or private enterprise is a decision for India to make’ (Rosen, 1967, 268). The Bank-led consortium of OECD donors surely shares the interest in democracy in India. Yet India has received much less aid-perperson than most LDC non-democracies. Also, many pressure groups in donor countries

80

Does Aid Work in India?

felt that democracy was threatened by the State of Emergency in 1975–7; however, although the Consortium countries met shortly after the Emergency was declared, only the UK sought to obtain any response by them to it, and was overruled. So donor ideology did not link elections or press freedom with aid—as possibly it should have done; the results, most of the time, would have been to raise India’s share in world aid. Did, and does, the donor community link ‘ideological’ views on macroeconomic policy with aid? The Bank was, and is, less vulnerable to this criticism than any one donor. It repeatedly stresses its commitment to efficiency in both public and private sectors, not to any particular borderline between them. However, some of its recent publications—notably on Africa (World Bank, 1981), but see also (World Bank, 1986, 1987; Lipton, 1987; Singer 1987),—alongside its role in structural adjustment loans, have led some critics to believe that, for some senior Bank policy-makers and researchers, ‘efficiency’ means not only cost-effectiveness plus competition, but also a smaller public sector (Toye, 1987,47–9). The Bank’s close link to the IMF—with its frequent advice, and even conditions for shortrun support, favouring reduced public expenditure, devaluation, and decontrol—increases these suspicions. These recent fears reawaken echoes of India’s experiences in 1965–70 and 1975. Thus Frankel wrote of the events of April 1975: The central government, meanwhile, [in the wake of the oil price explosion and the foreignexchange shortage], was again confronting pressures from the IMF for ‘stabilisation’ as a condition of a credit line to cover the massive balance-of-payments deficit. IMF recommended stringent fiscal discipline to control the money supply and complementary economic policies to freeze wages…and provide incentives to private investment. These recommendations were also then taken up by the World Bank as part of the negotiations on development aid then in progress between the Consortium countries and India. The Bank pushed even harder [than the IMF for such changes, and secured] a growth strategy with no more than marginal adjustments in the interests of more equitable distribution. (Frankel, 1978, 515–6)

Doubt can properly be cast on such accounts of pressures on India by the Bank, other donors, or the Fund. The point, however, is that (as any half-dozen random issues of the Economic and Political Weekly will confirm) such interpretations are increasingly prevalent in India. In the early 1970s they led to official unease about aid, clearly expressed in planning documents—and to strident, influential, and politically effective demands on the nationalist Right (Swamy, 1971) as well as on the Left, for ‘self reliance’ in the sense of near-autarky. The fall in real aid to India, past and proposed, represents in part a response to such unease and demands. Their plausibility has been enhanced since 1980 by the close co-operation between Bank and Fund. This has arisen because the acute, debt-and-drought-linked, balance of payments crises of much of the developing world have induced donors to shift aid from projects to general import support. This has led the Bank to impose conditions less on projects, but more on overall policies, in the context of ‘structural adjustment loans’ (SALs), and in the process to co-operate more closely with the IMF in medium-term lending operations (e.g. ‘extended financing facilities’ or EFFs) and conditions in numerous, mainly African and Latin-American, countries. Such operations and conditions are declared to be non-ideological, but in many cases have involved no published analysis (and, in some cases known to us, no unpublished analysis either) of the

Policy Dialogue 81 effects on the poor of implementing the conditions. These conditions almost always convey a standard message, the ‘four Ds’: devalue, decontrol, deflate, denationalize. Stabilization support from the IMF—especially short-term loans on the higher tranche, but even EFFs—almost always carry several or all these conditions. They have tended to leak into several sets of conditions for World Bank structural adjustment lending, despite its longer-term explicit goals. ‘Cross-conditionality’, moreover, impedes Bank lending if Fund conditions are even slightly under-fulfilled. Also, several bilateral donors are little disposed to give new aid, unless Bank or Fund conditions have been met. India has so far not needed to look for foreign help on quite such terms, and both the Bank and the Fund have since 1986 shown increasing interest in ‘adjustment with a human face’; however, memories of 1965 and 1975 loom large in India, as do present realities of donors and the ‘four Ds’ in Africa. No plausible alternative stabilization programme, radical or nationalist, permitting stabilization in a balance of payments crisis yet not at all involving the ‘four Ds’, has yet been worked out The Bank and Fund are bound by their Charters to collaborate. India has not yet needed a SAL-EFF package. Yet such counter-arguments do not quite succeed in defusing the charge of ideology. The charge bites deeper because Indians recall their experiences in 1965–6 and 1974–5—disturbing precursors of recent African styles of BankFund ‘policy dialogue’—and fear that the ideological hardening in the West will bring even tougher conditions in any future crisis (cf, Patel, 1986, 191). Some Bank officials adopt styles closer to those of some prominent western political leaders than to economic analysis. Whether their arguments are right or wrong, they cannot demonstrably be linked to efficiency or equity. Their adoption may reflect pressures upon the Bank by its donor members.24 It is perhaps more a question of statements and style than of substance.25 Also, aid donors, even the Bank, have changed the substance less than the IMF which, in the early 1980s, was driven by cash shortages to much more restrictive conditionality than the Bank. For example, in East Africa, the Bank was noticeably more long-term in its conditions, and less sceptical about public-sector activity, than the Fund. Yet the Fund has generally avoided public statements or high-profile documents that smack of paternalism and dogma.26 Meanwhile, some positions taken by some Bank officials and researchers, especially in 1984–6—and a quite new sense in some parts of the Bank that alternative evidence, analyses, and persons, however cautious or scientific, were not entirely welcome, even if efficient—have somewhat imperilled the Bank’s credibility with many Indians in its attempted dialogues.27 To avoid such peril is important. In India the dialogue has been impressively successful, and even during this difficult period was carried out by undogmatic Indian and Bank experts well qualified to keep it that way. Also, the Bank’s role in policy dialogue with India cannot be filled by any bilateral donor. There is little prospect that bilateral or commercial transactions might lead to healthier or less arm-twisting dialogue than multilateral aid. Indian officials’ fear of excessive reliance on commercial bank flows may have been influenced by alleged US pressure on Brazil to accept, in return for financial support (ultimately from the Federal Reserve), and US military landing rights at Recife airport. One could equally mention successful Soviet pressure on India to repair, and provide spares for, Egyptian aeroplanes during the 1973 war—pressure conditioned upon trade credits as well as bilateral aid. As a general approach, bilateral macro-leverage on India is so discredited, not least by the memories of

82

Does Aid Work in India?

1965–6, as to be dead. However, the risk that ideology may be seen by GoI as creeping into Bank relations with it—especially if the poverty-focus seems to recede, while ID A aid to India dwindles—is a serious threat even to Bank-mediated sectoral dialogue. A good sign, in autumn 1987, was the Bank’s major new research and policy attention to the impact of structural adjustment on poverty groups; this could be helped and broadened by new Indian research to evaluate the successes, failures, and prospects of transfer to other countries, of their many different programmes, some Bank-aided, for poverty alleviation. A new difficulty since 1987, however, is that the reorganizations of Bank staff may bring to Indian affairs, both in Delhi and in Washington, persons of considerable distinction but limited Indian experience, and some rather prolonged senior vacancies. Risk of rigidity The Bank now represents a single conduit for most dialogue with donors above the project level. In India, it is working in a big country with many different physical and political environments. The Bank thus faces a special need to avoid imposing one monolithic view, even if that view appears to Bank staff to be not ‘ideological’ but purely technical or efficiency-oriented. The responsibility is so great, not because GoI cannot do projects without the Bank or other donors—about three-quarters even of public-sector project funding remains domestic—but because, as the overwhelmingly major cash-linked source of outside advice, the Bank, if rigid or monolithic, risks using its prestige to overpersuade India of spuriously general solutions, sometimes based on small, costly, and regionally specific pilot projects. Yet the Bank’s own wish to move money swiftly and at low administrative cost must favour exactly these general, technocratic, all-India approaches. Has this been a serious problem in practice? We concentrate here on the risk of rigid procedures; the consequences at project level are considered in Chapter 5, and the implications for TC and institution-building are taken up in Chapter 6. In respect of aid to irrigation: the World Bank’s standard package for improving…canals included (1) lining the main canal, (2) land levelling below the outlet,…(4) rotational irrigation between…sub-outlet blocks, and (5) formation of water users’ associations at the outlet level, to implement the rotational irrigation schedule; [and more recently] (6) building the main system so as to allow for ‘flexibility’ [, i.e. for] varying quantities of water to be delivered to each location over the course of the crop season, in line with changing crop-water requirements. Perhaps the major single weakness [is that] the package is to be applied more or less uniformly everywhere, with little research being done to match the ingredients against the environment. This makes for ease in project formulation and lending. (Wade, 1982, 171–2)

However, it is too rigid an approach to serve as a good basis for sector dialogue. Such rigidity is especially dangerous where other donors (e.g. the USA) explicitly funnel their views on Indian irrigation policy through a single source of policy advice, the Bank. For example, canal lining (as against telegraphs and other informational improvements) is a cost-effective way to improve water use only in sandy soils with well-managed canals. ‘Rotational irrigation below the outlet is a waste of time except in some specific soil conditions and where water is very scarce.’ Water users’ associations probably make sense only where water is both

Policy Dialogue 83 moderately scarce and rather reliable. ‘Flexibility is much too ambitious a criterion [where, as is usual] outside the North-west, canal managers cannot allocate water evenly from head to tail within distributories and minors’ (Wade, 1982, 172–9). Some of these criticisms would be questioned by Bank irrigation engineers; others appear to have been taken account of in recent projects. Even so, it does look as if—in its urge to develop a testable policy model—the Bank for several years over-generalized, from the successful experience in re-organizing the Chambal irrigation system of Rajasthan, to create a blueprint for far less appropriate places and structures, including ones where organizational inputs and Bank help and overview, as well as soil and crop-mix, are far less suitable. This conforms with the description of a leading Bank expert: The first objective of our technical assistance was to induce the Indian irrigation [authorities in 1973–7] to use more modern design standards. During this period, we succeeded to upgrade some of the project designs. We argued with our engineering colleagues in favour of more and better water control structures, and, water measuring devices (Personal communication).

Some Bank experts place more stress than do their critics upon the experimental, learningby-doing nature of their part in policy dialogue (while also frankly recognizing the financial sticks and carrots used). However, the language of the above Bank version—‘modern’, ‘upgrade’, ‘better’—does suggest a global technocratic wisdom: a uniform technical package for, rather than environmentally specific institutional approaches to, better water management. It is uncomfortably echoed, in the rhetoric of some Bank officials and documents, over macroeconomic policy (everywhere): liberalization, reform, redefined to mean (mostly) not better access for poor people to assets or services, but smaller states and fewer controls on prices or private firms (including monopolies?). It is not only Indians who are concerned by, and may be harmed by, such rigid versions of dialogue. Over-generalization, from successful pilots to general solutions, is a widespread problem of rigid policy dialogues, not confined to irrigation: (i) Community development, in the 1950s, was thus ‘generalized’ by US donors and Foundations from Etawah, Uttar Pradesh to places with far fewer skilled personnel and far less physical potential for rural development (see p. 218). (ii) Support for co-operative agricultural product collection and processing has been similarly generalized by EEC. The original Anand project succeeded partly because it replaced the local monopolist (Polsons) in handling milk, for which small producers urgently needed collection to avoid spoilage, in an area that enjoyed unusually favourable conditions and personnel inputs. This pilot has been extended to several other areas and products, based on a national authority working through somewhat top-down ‘co-operatives’. (iii) Extension reform along T and V lines has, essentially, been spread out by the Bank from a single tested example of dramatic success in India—again in the ideal environment of Chambal—to several Indian (and African) States. The World Bank’s references to Chambal, and advocacy of T and V, apparently make no mention of ‘pilot project effect: that agricultural performance was so good because the World Bank was intensively involved’ and pulled Indian staff out of less-favoured areas.28 ‘A recent survey in the Chambal Command Area reveals that the extension system has virtually ‘gone to sleep’ since the area ceased to be a focus of interest’ (Moore, 1984; see, however, Feder and Slade, 1986).

84

Does Aid Work in India?

In any sub-sector (e.g. extension, agro-processing), there is always a danger of inferring a single solution for many environments on the basis of exceptionally carefully managed pilot projects in one or two. The danger is much greater if a single main outside bearer of expertise, cash and ‘policy dialogue’, able to grant or withhold cash and hence patronage from embattled officials, insists upon such a solution as a condition for sub-sectoral aid. The ‘pilot effect’, and the unwisdom of generalizing from it, are recognized in general terms. A senior Bank official with wide Indian experience told us: Bank projects go better than non-Bank projects. There is an outside source of troubleshooting in the event of delay, and of appeal if things go wrong within GoI. It is known that, if there is local deficiency, Bank staff will object at State or Central level. There is more kudos for timely fulfilment—and there are more visitors.

However, this general recognition does not always produce due caution in particular areas of sectoral policy dialogue. This is especially important because donors, in seeking to evaluate aid-effectiveness, stress replicability as the key to successful sector dialogue. One good project, in India where aid has ranged from 0.5 to 2.0 per cent of GNP, may simply replace government money that would otherwise have paid for the same project; what is felt to matter most, therefore, is innovative projects that are cost-effectively replicable, and sector dialogue that explores their worth first, and advocates replicability, if appropriate, only subsequently. We concur with this donor assessment. However, despite the Bank’s massive expertise, we do see dangers in its sometimes rather rigid transfers, as major protagonist in policy dialogue, from pilot project to national blueprint. Career incentives in any large bank, and in some aid agencies, tend to reward ‘moving money’ in big amounts; if the project later fails, the rewarded money-mover has usually moved as well, and is seldom penalized. Changing fashions It may seem inconsistent to claim both that policy dialogue can impose rigidity—as donors and recipients together come to accept a specific blueprint for general use—and that it is subject to swings of fashion. Unfortunately both claims are valid. Blueprints once set are applied far too generally, but they are re-set every few years, following changes of mood, as much as of scientific evidence, in economics or agronomy or engineering. Development studies are notoriously prone to swings of fashion. Often, these swings are embodied in ‘impact statements’, on every major project, required by national legislatures from bilateral aid project managers, or sometimes even from multilateral agencies. To insist that all project officers design or evaluate bridges or irrigation schemes with major reference to the role of any one sex or age-group, or to non-formal education, or to the preservation (in countries where poor people are endangered medically) of endangered non-human species, is probably not very sensible. Yet such fashions, once formalized into requirements, exert influence long after the next fashion has replaced them. There is thus a real danger that sectoral dialogue—and project preparation—may become cluttered with large amounts of almost empty ‘busyness’. More seriously, the fashions can distort projects—as with UN pressures towards the accelerated spread of intra-uterine devices in the 1960s (Cassen, 1978, ch.3)—producing ‘successful’ dialogue but bad policy outcomes.

Policy Dialogue 85 The problem is not simple, because at least two ‘development fashions’ have lasting importance. It matters greatly that the project cycle, at all stages, should investigate and seek to improve the impact of the project—directly and via other activities29—on poor children, women, and men. It is also of durable importance that the bias of large capital funding—domestic or foreign—away from rural people, and especially from food-producing smallholders, be exposed and corrected. The language of donors, and of GoI, increasingly reflects these lasting truths. Their new-found place in the fashions of development has therefore done some good to policy dialogue, because rhetoric is heard and quoted back, and carries some commitments. Changing fashions, moreover, can in a healthy dialogue involve mutual learning. Patel (1986, 215) recalls GoI’s ‘difficulties in the early 1960s in persuading the World Bank that rural electrification was important and deserved World Bank assistance. Later, in the early 1970s, the Bank had begun to criticize us for not devoting enough attention to rural electrification!’ By the early 1980s joint analyses had convinced both India and the Bank that remote electrification had grid costs so high as to be a cost-ineffective way to help the rural poor. Unfortunately, however, if a new policy guideline has been adopted for reasons of fashion, it may be abandoned later even if it was and remains correct.30 In 1967—after two droughts had cruelly exposed the effects of India’s prolonged under-emphasis on agriculture; and after the new cereal varieties had demonstrated their huge potential—an influential, and otherwise perceptive, expert could complain of: pressures [that] have contributed to a set of policies favoring expenditure in the low-return sector of agriculture. It is here that the weight of American policies associated with the American aid program to India can make a difference in the short run. In the longer run, the increase in the strength of industrial and urban political groups and the pressures and ideas generated within these groups would hopefully lead to a different political power structure and a set of policies that would continuously stimulate industrial growth. (Rosen 1967, 262–3)

There is hardly a worse use of policy dialogue than to intensify the strong Indian pressures for steel mills before canal maintenance, especially in 1967, but even today. Yet most commercial pressures in donor countries, and many pressures within aid agencies for technically safe spending in big units at low administrative cost, still favour such misplaced heavy-industrial emphases. Similarly, the ‘focus on poverty’—apparently a central feature both of World Bank policy guidelines and of the Indian Seventh Plan—remains vulnerable, both to the combineharvester salesmen and to the genuine, if misguided, believers in trickledown. Academics may feel they can gain promotion by saying (researching, theorizing) something new: currently that poverty is after all easily cured by growth, or that agriculture responds rapidly to urban development even if governments discourage such responses. While (moderately) new again, such fashions are, however, still untrue. Today’s academic fashion is too often tomorrow’s blueprint for sectoral policy dialogue.

86

Does Aid Work in India? Choosing sectors and topics for dialogue

The fashions of development analysts play a large part in determining what sectors and topics donors emphasize in policy dialogue. The wish of donors to find blueprints, usable in many different projects at low administrative cost, can lead them to turn these fashionable sectors—and topics for the handling of any given sector—into rigid project guidelines, even in places as big, and as politically and environmentally diverse, as India. The focus of policy dialogue upon one powerful donor, the World Bank, leaves a recipient with few alternatives but to accept yesterday’s fashions, turned into today’s and tomorrow’s blueprints by Bank experts. Fortunately, most of these are highly competent and fully aware of the risks in this process. In assessing the contribution of sectoral dialogue to aideffectiveness in India, however, we have to enquire whether the key sectors and topics have been correctly specified, and whether there is enough room for flexibility if new facts (not new fashions) challenge the fashions of yesterday. One major example of how sectoral policy dialogue may be pushed, by development fashions, to seize upon ‘wrong issues’ is in agricultural credit (see Chapter 2, pp. 58–61) and World Bank, 1981a). Credit is an important sector for agricultural development in some parts of India. Also—because of the research efforts of the All-India Rural Credit Surveys of the Reserve Bank of India (together with many micro-studies), together with the long period of experiment with alternative channels of agricultural financing and refinancing— there is major Indian capacity to engage in dialogue in which both Indian and World Bank authorities could learn as well as teach. Further, most World Bank evaluations show very satisfactory rates of return to its lending for agricultural credit, not least in India. However, the Bank may be focusing its policy dialogue upon the wrong issues. First, in the successive negotiations with the Agricultural Refinance Corporation and its successor, NABARD,31 the Bank has concentrated almost entirely on the need to reduce and recover overdues. Yet, despite the importance of ensuring adequate financial returns so as to keep lenders viable, it is surely even more important not to choke off credit when—as is amply proven for India, even allowing for past overstatements32—the economic rate of return is more than acceptable. Further, research in Tamil Nadu (Ch. 2, p. 59) clearly confirms that many so-called ‘overdues’ are formal roll-overs and/or are secured against gold or jewels, so that even the financial return is better than it looks. Bank emphasis should surely shift, therefore, to the need to get larger shares of credit to small and/or labour-intensive farmers. That emphasis is implicit in the Bank’s ‘poverty focus’ (and the classification of many of these credit-supportive loans as poverty-orientated ‘rural development’), and may even improve repayment rates, which are better among smaller farmers (Lele, 1974; Lipton, 1976).33 Second, there is the issue of whether the extra assets, acquired with extra rural credit associated with aid, are employment-generating or otherwise poverty-orientated. A World Bank (1981a) evaluation of ten IDA projects commenced in 1970–3, to support Indian farm credit, takes a very favourable view of loans for tractor hire and purchase, but appears to overlook conclusive counter-evidence (Agarwal, 1980,1984; Binswanger, 1979; and see Chapter 5 below) that, while privately profitable, the shift from animal to tractor draught displaced labour without raising output in most of South Asia—even in the extremely favourable environment of the Punjab. The Bank and USAID rightly stress collecting

Policy Dialogue 87 overdues, avoiding large subsidies (which tend to benefit almost entirely wealthy farmers), and ensuring that institutional credit can be embodied in extra investments with satisfactory returns. But an opportunity is being lost of re-orienting the discussion towards the means by which—without substantial credit subsidies, which are often diverted to richer farmers (or intermediaries) and which threaten the viability of lending agencies—credit can be focused on smaller farms, and on labour-intensive fixed and working capital. To concentrate the debate on how to collect overdues—instead of on how to select efficient activities, embodying equitable types of credit—is implicitly to weaken Indian resolve to use credit to increase the productivity of poorer farmers. In defence of the ten IDA projects, they antedated the poverty-focus of Mr McNamara’s 1973 Nairobi speech, so that ‘the tractor components were not designed to reach small farmers’ (World Bank, 198 la). However, there seems in general to be extremely inadequate effort to focus the policy dialogue on the poverty issues, especially where poverty-reduction and high efficiency are compatible. We have mentioned the apparent failure of Bank-Fund negotiators in 1975 to raise the question of how the—doubtless necessary—stabilization measures might be designed to minimize harm to the very poor. Also, although both Indian work and responses by individual Bank staff (see, for instance, Ahluwalia, 1977, 1986) on the ‘poverty/basic-needs’ debate have been of high quality, such work, even the Bank’s own, penetrates too little into the making of aid policy. The Bank’s own Report on Consortia and Consultative Groups (1978) makes it possible to see what was placed on the agenda for Consortium meetings, The nutrition, life-expectancy, literacy, or landholding of India’s poor—and the likely effects on these variables of alternative aid policies—are not issues with which the Consortium donors appear to have been confronted by the Bank secretariat, despite its excellent, ample staff work on these matters. Apart from the choice of topics for policy dialogue, the choice of sectors greatly influences aid-effectiveness. We saw how, as late as 1967, a perceptive US specialist actually proposed (even) less agricultural emphasis for India and its major donors. Not until the mid-1970s did that balance begin to be put right; as a former Chief Economist at USAID rightly remarked: Foreign aid to India was largely directed to industrial rather than agricultural development… from 1951–2 to 1970–1, only 2 per cent of foreign loans and credits (overwhelmingly aid) were earmarked for agricultural development, compared to 61 per cent for industrial development… GoI, much criticised by foreign observers for underemphasising agriculture, allocated 17 per cent of investment in this period to agriculture and irrigation. (Mellor, 1976, 244–53)

The proportion of Bank Group lending worldwide that comprised ‘agriculture’ plus ‘rural development’34 rose sharply in the 1970s, but has fallen quite sharply, in the budgetary projections (made each year for the next quinquennium), since 1982. While policy dialogue clearly does best by emphasizing the right sectors, it is in India not so obvious what those sectors are. Even though the proportion of public expenditure supporting agriculture, irrigation, etc. remains somewhat too low in India—though the shortfall is nothing like as big or harmful as in most of sub-Saharan Africa—is agriculture, etc., the sector where the industrial donors, and their main conduit for policy dialogue (the Bank), are relatively best placed to engage constructively in policy dialogue? Increasingly,

88

Does Aid Work in India?

GoI takes the view that Bank expertise, and its capacity to raise appropriate funds and supporting technology elsewhere (especially in the Consortium), are greatest in energy, heavy industry, power and transport. Fortunately the Bank has also established locus standi with GoI in several agricultural sectors—irrigation, extension, credit, some aspects of research—but it is more the Bank and other donors than GoI that seek rural emphasis in aid and policy-dialogue. Even the donors’ enthusiasm has been damped in the 1980s by the low current and projected world prices (largely created by western farm policies) for major Indian farm products—though in India world prices are probably an imperfect basis for establishing sectoral rates of return.35 Despite India’s own proven skills in public policy and investment for agriculture and rural development, a major donor input in these sectors, in policy dialogue as well as aid, remains likely to increase aid-effectiveness, for two main reasons. First, while most Indian specialists now recognize that ‘20 per cent of public investment for 70 per cent of workers and 35 per cent of output’ is too low a priority for agriculture, the immediate political pressures for domestic spending—usually less profitably—on industry and infrastructure remain stronger. Foreign cash and policy support for agriculture and rural development, if based on real expertise as is the Bank’s, can greatly help GoI to defend, against strong shortterm pressures, the larger rural allocations that it knows to be right. Second, India is in many technical, research, delivery, and economic fields a world leader in formulating agricultural policy (if not in mustering resources or power to implement it). ‘Policy dialogue’ is needed to help the Bank and other donors to learn lessons, about both what to do and what to avoid, for testing and possible application in other developing countries. The proper selection of topics, and to some extent of sectors, for concentration in sectoral policy dialogue is most effectively advanced by the open discussion, between donor and recipient experts, of carefully conducted project evaluations. However, their contribution to dialogue has been hampered by some donor and recipient procedures: (i) only the Bank, and recently US AID, regularly evaluate returns and other outcomes from major aid projects; (ii) use of the main source of good project evaluation is restricted because objections by recipient governments prevent the Bank from publishing OED’s evaluations;36 (iii) operating divisions within the Bank—and probably other donor agencies—quite often decide to hold up the last few per cent of the cash on large projects; this means that completion reports (and therefore evaluations) cannot be prepared; (iv) in general, evaluations take place much too soon after project completion: comparison of the very thorough analysis (much fuller than an ordinary evaluation) of the Muda project in Malaysia carried out shortly after completion (Bell, Hazell, and Slade, 1982), and the OED’s (also excellent) impact study based on research done a few years later (World Bank, 198 1b), shows that a few years of full operational experience reversed many of the apparently favourable equity effects of the project; (v) many evaluations are hit-and-run jobs in which an expert, genuine but often in a quite inappropriate discipline, is given a quick tour of many sites by persons with clear initial positions or special interests. A major bilateral donor’s 1976 evaluation of tractor aid to India appears to have been carried out in a couple of weeks, by an agricultural engineer untrained in economics, guided by local tractor distributors; (vi) most seriously, there is no systematic attempt, by such agencies (or LDCs) jointly, to compare evaluation documents and draw lessons for future project cycles. Nor do most big projects in LDCs—aided or not—begin, as they should, by reviewing relevant evaluations.

Policy Dialogue 89 Bank, US and (recently) UK project evaluations have reached high standards of professional competence, but too much aid to India is still evaluated in unsatisfactory or unpublished forms. In 1982 a major EEC evaluation of aid to dairy development (see p. 193) allowed itself to be almost entirely guided by Indian project administrators who, while known to be dynamic and competent, are widely regarded in India as being somewhat reluctant to release information, and rather mistrustful of objective research. A British official evaluation of aid for a private-sector fertilizer factory erected for an Indian subsidiary, IEL, with a majority shareholding held by a British company (ICI), fails to disclose the extent of the interest-rate subsidy to the donor country firm involved, presumably on the grounds that it was commercially sensitive information (Haley and Hesling, 1986). Aid effectiveness would be improved by frank, published evaluations of past projects. However, the Bank evaluations are generally full, fair, and accessible (though unpublished). Aid effectiveness would also gain if, whenever major new projects or sectoral policy-dialogues were being considered, systematic review of good evaluations of past projects in that sector were first jointly undertaken by donors and recipient together. Can the dialoguer deliver? Most assessments of policy dialogue (including the World Bank’s OED evaluation of structural adjustment loans, e.g. those to Turkey and Kenya) use, as the major criterion, whether the professionals—donor and recipient—are satisfied that each party has ‘delivered’ according to the conditions agreed. In the India-Bank dialogue: (i) Bank doubts about financing NAB ARD centre upon repeated Indian failure to deliver the promised reductions of overdues; (ii) Bank satisfaction about lending to irrigation centres on successful Indian ‘delivery’ of proposed design changes, and appreciation by senior Indian engineers of the virtues of these changes; (iii) the case for assigning ‘substantial (Bank) lending resources to the industrial sector’ is made in a 1984–5 Country Programme Paper in terms of a long series of liberalizations during 1981–3—of imported inputs; of investment licensing; of coal, steel, and cement pricing—together with the need to compensate GoI for the political and perhaps distributional costs of these consequences of past sectoral dialogue. There is something worryingly self-confirming about such assessments, whether they end up praising India for delivering, or blaming India for not delivering. One should also evaluate whether what is delivered secured the desired results cost-effectively: i.e., whether the conditions were well chosen. (i) In rural credits, should NABARD be helped (or pressed) to collect overdues, or, rather, to redirect resources to intermediaries in ways that favour small farmers—anyway better repayers—as final borrowers? (ii) In irrigation, were the ‘delivered’ design changes really more important than institutional changes? Did they really provide good results cost-effectively outside north-west India? (iii) In industrial policy, while the Bank praises the liberalizations, should it be justifying and structuring further aid to the sector in terms of the need to relax the political constraints upon ‘how much further and faster the GoI can move’, and the need for aid ‘to support and sustain the adjustment process which has now been initiated’? Should not the Bank rather address—by the structure of its industrial lending—the recognized ‘concern from both the Left and the Right that in a world of rising protectionism, it will be increasingly difficult to [offset] import liberalizations with export expansion [and] that the burden of adjustment inevitably falls most heavily on those least able to bear it’ (World Bank, pers. comm.)?

90

Does Aid Work in India?

On the Indian side, confidence in policy dialogue is imperilled when donors do not deliver aid that has been pledged, after GoI has accepted the short-run political cost of (possibly economically desirable) changes. In 1964–5, the Bank’s ‘Bell Mission’ secured commitment to a package of liberalizing reforms, and provided ‘informal assurances of a significant increase’ in aid to US$1.5bn a year by the end of the Fourth Plan. After the 1965 war and drought, the June 1966 devaluation took place ‘encouraged by informal assurances of greater external support’ (T.Singh, 1974, 351), and following up the April meeting at which ‘Asoka Mehta placed virtually all of India’s major Fourth Plan schemes before George Woods for his consideration and approval, (obtaining) assurances of a fiveyear aid commitment (at) US$1.2bn a year’ (Frankel, 297–8). However; By November 1967…[the USA] had to inform the Consortium it could not keep its April pledge to provide US$380 million of the US$900 million already promised for 1966/67. Thereafter, actual new commitments by the US for non-project aid to India showed a steady decline from US$300 million in 1966/67…to US$200 million in 1968/69. Reflecting America’s reduced role, total new commitment for project aid [exclusive of food] fell from US$888 million in 1966/67 to US$534 million in 1967/68 and US$594 million in 1968/69 [including debt relief]. (Frankel, 1978, 322)

Indeed, partly reflecting ‘sharp increases in debt repayment’ (Mellor, 1976, 223), net aid to India fell much more sharply, from US$1.3bn in 1964–5 and 1965–6 to US$0.5bn. in 1970–1 and 1971–2 (Chaudhuri, 1978, 99). It is absurd to expect aid donors to put up and shut up. They owe their taxpayers evidence that the aid voted has been well used. But most aid to India has comprised loans, with a grant-element of total aid that has been really high (over 80 per cent) only for a few years. Especially in that context—at least as much as donors have the right and duty to see that India ‘delivers the goods’—so the recipient has the right to expect that aid pledges, made in return for policy changes, will be fulfilled. The substantial cuts in the 1980s in the real value of IDA aid—plus major redirection (actual and planned) of IDA aid away from India, largely in response to US judgements and pressures—render it rather unlikely that, at least, implicit pledges made in past sectoral dialogues will be fully honoured on time. Even a big rise in IBRD (non-concessional) flows to India can hardly make up for significant falls in IDA flows. If Consortium members wish to provide signals favouring liberalizations, they will have to deliver the promised quid pro quo, whether through IDA or otherwise. A repeat of the 1964–9 experience could damage the (generally ‘aid-effective’) sectoral multilateral dialogue, perhaps as terminally as non-deliveries of pledged aid then wounded the (much less satisfactory) bilateral macro-leverage. It is worrying that, since 1985, successive Bank ‘progress reports’ on sub-Saharan Africa—while praising many of its governments for policy reforms—have repeatedly, but fruitlessly, upbraided donors for not delivering the promised capital flows in return. Both Africa and India since 1985 suggest that donors might usefully re-examine the lessons of India in 1965–71. Conditionality cuts both ways. References Agarwal, B. (1980) ‘Tractorization, productivity and employment: a reassessment’, Journal of Development Studies, 16:3.

Policy Dialogue 91 Agarwal, B. (1984) ‘Tractors, tubewells and cropping intensity in the Punjab’, Journal of Development Studies, 20:4. Ahluwalia, M.S. (1977) ‘Rural poverty and agricultural performance in India’, World Bank Reprint Series No.60. Ahluwalia, M.S. (1978) ‘Rural poverty and agricultural performance in India’, Journal of Development Studies, 14:3. Ahluwalia, M.S. (1986) ‘Rural poverty, agricultural production, and prices: a re-examination’ in J.Mellor and G.Desai (eds), Agricultural Change and Rural Paverty, Delhi, Oxford University Press. Bell, C., Hazell, P. and Slade, R. (1982) Project Evaluation in Regional Perspective, Baltimore, John Hopkins University Press Bhagwati, J.N. and Srinivasan, T.N. (1975) Foreign Trade Regime and Economic Development: India, New York, Columbia University Press (for NBER). Binswanger, H. (1979) The Economics of Tractorization in South Asia, New York, Agricultural Development Council. Burch, B. (1979) Overseas Aid and the Transfer of Technology: a Case Study of Agricultural Mechanization in Sri Lanka, Ph.D. thesis, Brighton, University of Sussex. Cassen, R. (1978) India: Population, Economy, Society, London, Macmillan. Chaudhuri, P. (1978) The Indian Economy: Poverty and Development, London, Crosby Lockwood and Staples. Chopra, P. (1968) Uncertain India, Bombay, Asia Press. Desai, P. (1972) The Bokaro Steel Plant: A Study of Soviet Economic Assistance, New York, American Elsevier. Farrington, J. and Abeyratne, F. (1980) Farm Power and Water Use in the Dry Zone of Sri Lanka, Part I, Research Study No. 43, Colombo, ARTI Farrington, J. and Abeyratne, F. (1982) Farm Power and Water Use in the Dry Zone of Sri Lanka, Part II, Research Study No. 52, Colombo, ARTI. Feder, G. and Slade, R. (1986) ‘A comparative analysis of some aspects of the training and visit system of agricultural extension in India’, Journal of Development Studies, 22:2. Frankel, F. (1978) India’s Political Economy 1947–1977: the gradual revolution, Princeton, Princeton University Press. Government of India (1982–83) Economic Survey, Delhi, India Press. Haley, G. and Hesling, S. (1986) An Evaluation of Indian Explosives Ltd.’s Expansion of Fertilizer Production at Kanpur, EV 308, London, ODA. Herring, R.J. (1983) Land to the Tiller, New Haven, Yale University Press. International Labour Office (1971) Matching Employment Opportunities and Expectations (‘Ceylon Report’), Geneva. Jha, P.S. (1980) India: a Political Economy of Stagnation, Bombay, Oxford University Press. Krishna, R. (1983) ‘Growth, investment and poverty in mid-term appraisal of Sixth Plan’, Economic and Political Weekly, 18:47. Krueger, A.D. (1974) ‘The political economy of the rent-seeking society’, American Economic Review, 64:3, June. Krueger, A.D. and Ruttan, V. (1983), Trade sector policies and the impact of assistance’, in Vol. I of A.Krueger and V.Ruttan (eds) The Development Impact of Economic Assistance to LDCs, Economic Development Center, University of Minnesota/US AID/Dept. of State. Lele, U. (1974) ‘The role of credit and marketing in economic development’, in N.Islam (ed.) Agricultural Policy in Developing Countries, London, MacMillan/IEA. Lele, U. (1979) The Design of Rural Development: Lessons from Africa, (revised edition with postscript), Baltimore, Johns Hopkins University Press.

92

Does Aid Work in India?

Lele, U, (1987) ‘Growth of foreign assistance and its impact on agriculture’, in J.Mellor et al. (eds) Accelerating Food Production in sub-Saharan Africa, Baltimore, Johns Hopkins University Press. Lipton, M. (1976) ‘Agricultural finance and rural credit in poor countries’, World Development, 4:7. Lipton, M. (1985) ‘Indian agricultural development and African food strategies: a role for EC?’, in W.M.Callewaert and R.Kumar EEC—India: Towards a Common Perspective, Netherlands, Peeters-Leuven. Lipton, M. (1986) ‘Aid-effectiveness, prisoners’ dilemmas and country aid allocations’, IDS Bulletin, 17:2, April Lipton, M. (1987) ‘Agricultural price policy: which way at the World Bank?’ Development Policy Review, June. Mellor, J. (1976) The New Economics of Growth: a Strategy for India and the Developing World, Ithaca, Cornell (for Twentieth Century Fund). Moore, M.P. (1984) ‘International development, the World Bank, and India’s new agricultural extension program’. JDS, 20:4. Mosley, P. (1987) Overseas Aid: Its Defence and Reform, Brighton, Wheatsheaf. Papanek, G. (1972) ‘The effect of aid and other resource transfers on savings and growth in less developed countries’, Economic Journal, vol. 82. Patel, I.G. (1986) Essays in Economic Policy and Economic Growth, London, MacMillan. Rao, V. and Narain, D. (1963) Foreign Aid and India’s Economic Development, Delhi, Asia Press. Rose, T. (1985) Crisis and Recovery in Sub-Saharan Africa, Paris, OECD. Rosen, G. (1967) Democracy and Economic Change in India, San Francisco, University of California Press. Rosen, G. (1985) Western Economists and Eastern Societies, Delhi, Oxford University Press. Singer, H.W. (1987) ‘The World Development Report 1987 on the blessings of ‘outward orientation’: a necessary correction’, Journal of Development Studies, vol. 24. Singh, A. (1973) Impact of American Aid on the Indian Economy, New Delhi, Vora. Singh, T. (1974) India’s Development Experience, London, Macmillan. Sukhatme, V. (1983) ‘Assistance to India’, in vol. II of A.Krueger and V. Ruttan (eds) The Development Impact of Economic Assistance to LDCs, Economic Development Center, University of Minnesota/ USAID/Dept. of State. Swamy, S. (1971) Indian Economic Planning, New York, Barnes and Noble. Toye, J. (1987) Dilemmas of Development, Oxford, Basil Blackwell. Toye, J. and Clark, G. (1986) ‘The aid and trade provision: origins, dimensions, and possible reform’, Development Policy Review, 4:4. Wade, R. (1982) ‘The World Bank and India’s irrigation reform’, Journal of Development Studies, 18:2. World Bank (1981) Accelerated Development in sub-Saharan Africa: an Agenda for Change, Washington DC, The World Bank. World Bank (1981a) Agricultural Credit Projects: A Review of Recent Experience in India, Report No. 3415, Washington DC, The World Bank (Operations Evaluation Division). World Bank (1981b) Impact Evaluation Report—Malaysia: Muda and Kemubu, Irrigation Projects, No. 3587, Washington DC, The World Bank (Operations Evaluation Division). World Bank (1982) Focus on Poverty: A report of a task force of the World Bank, Washington DC, The World Bank. World Bank (1986) Poverty and Hunger, Policy Paper, Washington DC, The World Bank. World Bank (1986a) World Development Report, Washington DC, Oxford University Press for the World Bank. World Bank (1987) World Development Report, Washington DC, Oxford University Press for the World Bank.

Chapter 4 The Systemic Effects of Aid and Donor Procedures

Introduction So far aid to India has been examined in a number of partial aspects—its implications for macroeconomic management, for the alleviation of poverty and for the development of Indian policy-making. In this chapter, a more holistic approach to aid is adopted, focusing on the question of whether aid has important ‘systemic effects’ on India’s political economy, and whether particular procedures of aid-giving could or should be modified in order to reduce adverse systemic effects, or increase beneficial systemic effects. The reason for asking questions like this is the fear that foreign aid may, in certain circumstances, have strong distorting effects on the recipient’s economy, and therefore that an evaluation method which concentrated only on the results of individual aided projects would fail to detect something important about the overall aid-giving process. Before proceeding to an analysis of aid project evaluations in the next chapter, it is necessary to tackle these larger and more comprehensive questions. One would expect, as a first guess, that the distorting effects of foreign aid on India’s political economy would not be large, especially in comparison with small, poor subSaharan African countries which receive aid. The first reason for this is the smallness of the aid flows to India relative to the size of her economy, as illustrated in Chapter 1. They are simply not large enough to be the cause of ‘Dutch disease’, the loss of competitiveness in tradeable goods brought on by a major discovery of oil or natural gas, or a major rise in a commodity price or (by analogy) a major inflow of foreign aid. The second reason why one would not expect aid-induced distortions to be large in India is India’s thirty-year experience of managing aid inflows precisely to avoid the worst forms of microeconomic distortions. India’s strong administrative capacity relative to other developing countries (particularly, again, countries of sub-Saharan Africa) has permitted an effective learning process to take place in the field of aid management. The broad conclusion of this chapter is that India has succeeded in minimizing the adverse systemic effects of foreign aid. A relatively sophisticated set of defences has been established over the years to buffer the wide variety of political and economic pressures that can emanate from aid donors. Although a number of problems remain unresolved between the Indian Government and the aid donors, they are not likely to be a major disruptive force in the overall aid relationship. Two key components of the defensive strategy should be noted—the multiplication of the number of donors combined with the centralisation of control over the channelling of aid to the domestic economy. By encouraging the giving of aid by a large number of countries, and using aid from the USSR and its allies to create extra room for manoeuvre in its relations with the Aid-India Consortium (which DOI: 10.4324/9780203840153-5

94

Does Aid Work in India?

represents the bulk of donors), India has side-stepped many of the more extreme problems of aid-dependency. By centralizing its own relations with donors through one government department, it maintains a tight control over the aid process and creates an important buffer between donors and the internal agencies that are associated with aid-funding. In order to simplify the discussion of systemic effects, a basic distinction is made between multilateral and bilateral aid. This is a reflection of the fact that aid-giving procedures vary significantly as between the multilateral and bilateral agencies, and therefore the types of problems and conflicts which can be created are also different in each case. A secondary distinction is made between project aid, where some discrete developmental activity is being wholly or partly financed, and non-project aid (which includes debt relief, food aid, and general balance of payments support). Within these broad categories, it should be noted that, quantitatively speaking, multilateral project aid is the largest category of aid to India. Generally speaking, India’s relations with the multilateral agencies have (after the conflicts over the 1966 devaluation) been fairly smooth ones. Some friction has been experienced over the requirement for international competitive bidding. But this subsided as the agencies conceded a margin of price preference to bidders in the recipient country and certain other developing countries. The multilateral agencies have also increasingly moved towards a more relaxed attitude to the financing of local costs and this has reduced another potential area of conflict With the abandonment of attempts to use aid leverage to influence macroeconomic policy, such leverage as now exists operates at the more acceptable level of sectoral policy. The long gestation period of multilateral projects remains their most substantial systemic drawback. The most serious systemic drawback to bilateral project aid has been source-tying, i.e. the requirement that the imports which the aid finances should come only from supplies within the donor’s own country. This is objectionable to the recipients because it can confine their choice to a range of goods that is unsuitable in design or quality for their project’s purposes, or because by limiting competition it allows the donor-country suppliers to charge higher prices that would be possible against the full force of international competition. Bilateral project aid is the second largest component of India’s aid, and in the period up to the 1970s, source-tying was widespread and represented a serious problem for India. During that time, however, India managed to accumulate a considerable expertise in comparing aid offers for projects with ruling international prices, and did not hesitate to reject offers which involved heavy price-loading by donor national suppliers. Since the mid-1970s, the extent of formal source-tying has changed. The United States, a major bilateral donor, now permits procurement in India and other developing countries, although not from its major industrial competitors. The Netherlands and Scandinavian donors, along with the Germans and Japanese, impose relatively few procurement restrictions. But just as the removal of tariff barriers can be followed smartly by the erection of nontariff barriers, so the reduction of formal source-tying can be, and in some cases has been, accompanied by new donor practices which have the same effect as source-tying. Certain bilateral donors have in the 1980s set aside part of their aid budget for high-visibility export promotion purposes. These involve an increase in the intensity of aid-tying, because procurement is not tied merely to the donor country’s suppliers, but is given to clinch a contract for a particular national firm, as in the recent UK aid for sales of power stations and helicopters to India. Often here not only is international competition eliminated, but

The Systemic Effects of Aid and Donor Procedures

95

also intra-national competition among firms in donor countries. In addition, it can be argued that there is less control over whether the aided project is sound from a developmental point of view. Another donor practice which can be an effective informal method of tying aid that is formally untied is the rapid switching of country aid between different sectors. Failure to publicize calls for tenders and the specification of donor country consultancy firms are other methods which also have the same restrictive impact as source-tying. All of these methods of informal source-tying remain a serious impediment to the recipient of aid achieving value for money in its spending of aid. The absence of much finance for the local costs of projects has been a cause of project delay and inappropriate project choice until recently. Now, alongside the multilateral agencies, both the UK and the USA have found ways of significantly increasing the local cost element of their bilateral aid. But, to the extent that local costs are usually also recurrent costs, this favourable development has raised a new problem. Are recipients of local recurrent cost aid being committed by donors to large expenditures in future years which their fiscal systems may not be ready to cope with? The previous exclusion of local costs was designed to finesse this problem, but now aid can require future commitment of recipients’ revenues. But whether the recipient will wish to withdraw from that commitment ought to depend on the social profitability of the project. If the project has been well chosen in the first place, finding the necessary revenues from internal sources ought to be an attractive proposition. In marked contrast with the experience of many aid-receiving developing countries, especially in Africa, India does not appear to have had serious problems with donor procedures of project appraisal, monitoring and reporting. Although these procedures do differ quite markedly among donors, and although they are compatible neither with each other nor with India’s own internal project procedures, the Indian authorities appear to have coped well. However, if some standardization of procedures between donors could be agreed, this would probably cut down administrative delays and be welcomed by the Indian side. An important choice which faces donors is whether to support its aid programme by a substantial local mission. Practice on this varies between donors, with the World Bank and USAID being well represented in India, but the UK High Commission being very shortstaffed for handling aid issues and relying heavily on intermittent special missions sent out from London. Local missions are a costly use of aid money and their effects are far from predictable. However, if donors are committed to making a reality of ‘policy dialogue’, an effective local mission is probably indispensable. If this is so, it might be worthwhile to consider the idea of a joint mission of aid donors, as the way of minimizing the cost of keeping the policy dialogue working well. A joint mission of aid donors could help by raising the average quality of aid field staff and of standardizing the functions which they would perform in relation to the recipient government (which presently show the same inter-country variety as do other donor procedures). This suggestion, however, does have political implications for the recipient, which may be reluctant to countenance a ‘united front’ by donors on aid policy issues. This reluctance could be reduced if dialogue was confined to project or sector issues, but may still be a stumbling block in India, which has always shown itself anxious to maintain a distance between its policy-making and the views of particular donors, either as individuals or as groups.

96

Does Aid Work in India? Many donors: help or hindrance?

Does the number of donors from which a developing country accepts aid matter? Does it make any difference whether a certain amount of aid is given by three donors or by thirty? The effect of larger numbers of donors works in two opposing directions. On the one hand, the proliferation of donors can strain a recipient’s administrative capacity to co-ordinate all the different forms of aid activity to the benefit of the economy, rather than to its distraction and disruption. Ultimately, a recipient can suffer from ‘aid overload’ when donors are allowed to start up too many different activities and to compete for scarce domestic resources in a vain attempt to complete them all. On the other hand: There may be some benefit in multiplicity (of donors) when it widens the recipient’s choices and permits a range of innovation which might not arise if only a few agencies were in the field. (Cassen, 1986, 223)

Thus it is not the absolute number of donors that matters, but the number relative to the level of administrative capacity. Further, the greater a recipient’s competence in aid management, the more it can gain the advantages of many donors—in terms of choice and innovation—without having to pay the serious costs that can arise. Essentially, India is an example of a country with the administrative capacity to take full advantage of a multiplicity of donors. Before considering that judgement in greater detail, it may be useful to sketch in some information about the main sources of aid and their trends over the last decade. Table 4.1 shows the distribution of utilizations of aid among the four groups of donors which are conventionally identified separately. These are the Aid-India Consortium, consisting of thirteen nations plus certain international organizations, whose activities are co-ordinated by the World Bank; the USSR and East European countries (of which the USSR is by far the biggest contributor); the OPEC countries and related organizations; and the ‘others’, notably the European Community (EC) and Switzerland. Table 4.1 also shows the shares of each group in India’s aid flow in percentage terms. It is clear from these figures that by far the greater part of India’s aid derives from the Aid-India Consortium which, even after the dwindling US contributions in the mid-1970s are taken into account, never provided less than 70 per cent of the total, The share of the USSR and its allies is normally very small. Their unusually high 10 per cent share in 1973–6 reflects emergency wheat assistance during the bad harvests of that time and this peak has not been regained since. After the oil price rise of 1973, OPEC and individual oilrich Middle Eastern states made a significant contribution to aid flows, but with the falling oil prices of the early 1980s this is unlikely to be a permanent feature of the composition of India’s aid. The ‘others’, including the European Community, are at best a make-weight in the total. But the significance of the role played by the USSR and allies in giving aid to India is somewhat larger than its small share in total aid flows seems to indicate. This is for three reasons.

The Systemic Effects of Aid and Donor Procedures

97

Table 4.1 India: aid utilization by source Rupees (crores)

1970–3 (average)

1973–6 (average)

1976–9 (average)

1979–82 (average)

1983–4

Aid-India Consortium

728.3

1.006.7

1,121.7

1,674.3

2,309.9

USSR and East Europe

301.1

138.0

52.1

35.2

150.5

0.0

199.7

184.9

40.9

80.0

OPEC and related Others Total

5.5

52.4

26.2

69.4

35.8

763.9

1.396.8

1,384.9

1,833.2a

2,576.2

Percentages Aid-India Consortium

95

72

81

91

90

USSR and East Europe

4

10

4

2

6

OPEC and related



14

13

2

3

Others

1

4

2

4

1

100

100

100

100

100

Total

Source G.O.I., Economic Survey, 1982–83 and Economic Survey, 1983–84, Table 7.3. Note a Total greater than components due to error in source

(i) Soviet government aid has been concentrated on sectors (steel, power) which the Indian planners have regarded as particularly vital in their strategy for economic development; (ii) The bulk of India’s aid derives from a group of western donors whose aid policies are harmonized—not fully harmonized, needless to say—through their participation in the Consortium. Soviet aid contributes more to widening the Indians’ choice of projects than would one additional western donor; and (iii) because of geopolitical competition between the USA and USSR, the latter is particularly keen to substitute for the former when sometimes American promises of aid fail to materialize. The classic illustration of factors (i) and (iii) is provided by the history of the Bokaro steel mill in the 1960s. In a fit of liberalism, and despite its considerable scepticism about the Indian heavy-industry, public-sector strategy of the Second and Third Plans, the US Administration of the day agreed to finance the construction of Bokaro. But it failed to secure the necessary legislative approval for this aid. At this point, the Indian Government was able to turn successfully to the USSR for the necessary funding. The manoeuvre was costly, both in terms of the delays in the implementation of the project and the subsequent indifferent performance of the plant, but the presence of the USSR as an aid-giver did help to maintain the Indian plan strategy basically intact (Kothari, 1970, 413). As has been argued previously, the ability to benefit from the multiplicity of aid donors depends partly on there being many donors willing to give or make concessional loans, but—granted such an international environment—it then depends on establishing the administrative machinery to control the aid giving process in the recipient’s national interest. The main features of the Indian Government’s aid management policies have been: (i) Responsibility for all negotiations with donors, for administering the documentation of aid inflows and for monitoring the implementation of aided projects is centralized in

98

Does Aid Work in India?

the Department of Economic Affairs within the Ministry of Finance. The DEA operates procedures which aim, as far as possible, to stabilize project spending even when the level of aid fluctuates—i.e. to separate as far as possible decisions on the size and composition of the investment programme from the arrival or non-arrival of aid for specific projects. (ii) The concessional element in aid finance is largely withheld from the agency in India whose project is being funded with aid money. The aim here is to dampen the incentive of state and local government, parastatals and other agencies to alter their development programmes just in order to take advantage of a new facility being provided free by a donor. That is regarded as being disruptive of overall planning priorities and an incentive to use sources of supply (i.e. donor country suppliers) which might otherwise not be chosen (Cable, 1980, 75). Of course, it is not possible to maintain this position entirely. Since aided projects do involve an abnormal level of work and documentation, agencies must be given some incentive to participate in them. As a result the central government normally retains 75 per cent of the concessionality of aid finance, while passing on 25 per cent to the agency concerned. (iii) Where aid is capital aid, the Indian Government operates a series of controls over the policy of public sector financial institutions which ensure that the end use of the aid is in line with pre-determined plan priorities. (iv) Where aid is in the form either of sectoral aid or aid for maintenance imports, the Indian Government has almost complete discretion over its use, with the result that donors will normally not be told exactly what it has been used for. As a result of aid management practices such as these, the Indian Government has exercised great care and vigilance over the absorption of aid Her success in this regard has been widely recognized. In a book that summed up India’s aid experience in the 1960s, it was said that ‘the multiplicity of donor agencies, each with their several points of view, gives India ample room for manoeuvre in obtaining a balanced pattern of aid. Cold War differences and the policy of non-alignment enhance her bargaining power still further’ (Eldridge, 1969, 70). At the end of the 1970s, the UK Select Committee on Overseas Development ‘were impressed by the determination of the Indian authorities not to allow aid to distort their priorities. The Indian authorities have maximized their own room for manoeuvre within existing resource constraints at the same time as they have minimized their dependence upon aid donors’ (House of Commons, 1979: xvii). Kalecki had also included India among the ‘intermediate regimes’ of which he wrote: (They) are the proverbial clever calves that suck two cows: each block gives them financial aid competing with the other. Thus has been made possible the ‘miracle’ of getting out of the United States some credits with no strings attached as to the internal economic policy. (Kalecki, 1972, 167)

The second half of that assessment, written in 1967, was just ceasing to be true. In the mid-1960s, the question of economic policy strings was placed firmly on the western donors agenda. Faced with the tight Indian system of aid management, what options were and are available to donors? They can usefully be classified according to the terminology developed by Hirschman (1970), that is, options of ‘exit’, ‘voice’, and ‘loyalty’. Loyalty was the initial option, a willingness to support India’s development effort even (in the case of the United States) contrary to better judgement. When loyalty had eroded, as it had by

The Systemic Effects of Aid and Donor Procedures

99

the middle of the 1960s, the options were ‘exit’ or ‘voice’, or some combination of the two. Some donors have indeed exercised the option to exit in their dealings with India. The exits have been of two distinct kinds. Some major bilateral donors have exited in order to try and exercise leverage over political developments in the South Asian region. Both the US and the UK cut off aid in response to the outbreak of the Indo-Pakistan conflict of autumn 1965, the UK giving in to pressure from the USA., which had treaty obligations to Pakistan under CENTO. When the Indo-Pakistan conflict surfaced again over East Pakistan/Bangladesh in December 1971, the British this time did not follow the USA’s ‘tilt towards Pakistan’, and the USA was alone in cutting off new aid authorizations until 1978. Clearly, at times of international conflict, India, like other aid recipients, has felt the effect of the ‘leverage’ which dominant aid donors, like the USA at that time, can exert, both directly on their own account and (sometimes, but not always) through their closer allies who also give aid. Such leverage caused great nationalistic resentment in India, which was convinced of the correctness of its international actions vis-à-vis Pakistan. Donors who exit inevitably point to the fungibility of aid and the feasibility of using aid finance indirectly to fund increased military expenditure (by directing aid to projects that would have been undertaken in any case and thereby releasing resources for military uses). It is difficult to assess the impact of the US exits from aid to India. In geopolitical terms, the USA probably did more to ensure the limitation of the conflicts by denying Pakistan rapid rearmament after the first bursts of armed conflict. But perhaps, in order to be able to do that, some ‘even-handed’ action against India was also necessary for reasons of successful political presentation. The other type of ‘exit’—or rather leverage achieved by threatening ‘exit’—was a joint multilateral and bilateral manoeuvre aimed at major changes in economic policy in the period 1964–6. For various reasons discussed at length in Chapters 3 and 7, this attempted manipulation came to grief, causing further resentment in India—even in moderates like I.G.Patel (1986, 73)—and has not been repeated. Threats of exit have now been replaced with the use of ‘voice’, Hirschman’s third option. Voice operates more gently, but also more continuously, through the institutions of ‘policy dialogue’, the nature and effects of which are discussed separately in Chapter 3. Policy dialogue takes place much more at sector and project level, than at national level. But to be effective, donors still need to co-ordinate their views at whatever is the relevant level. A multiplicity of donor voices favours a recipient government that wants to go its own way, without regard for donor opinion. Pure multilateral exit is also a logical possibility, although one most unlikely to affect India. In Latin America, some multilateral lending appears to have been responsive to the policy stances of would-be borrowing governments (e.g. Peru in 1969–73, Chile in 1970–3, and Nicaragua in 1979–82), particularly on questions concerning the nationalization of foreign private assets (Van de Laar, 1980, 33; Hayter and Watson, 1985, 214–23). India has dealt very circumspectly with such matters and accordingly has not run into these types of difficulties, and is not likely to encounter them in the immediate future. Systemic aspects of multilateral aid The Aid-India Consortium provides the great bulk of India’s aid and co-ordinates the overall level of future aid commitments by major Western countries. Through the Consortium both multilateral and bilateral aid is given, and each broad category is subject to markedly

100

Does Aid Work in India?

different conditions, procedures, and arrangements. Table 4.2 provides details on the composition of Consortium aid, in this respect, as well as a country breakdown of bilateral aid and the mixture of loans and grants. This table gives a good picture of the diversity of forms of the Consortium’s aid, although unfortunately it does not give detail on the breakdown between project and non-project forms of aid. Loan figures are shown gross rather than net, because at this point we are more interested in the spread of donors and types of aid than in the level of overall transfers between the Consortium and India. The gross figures indicate more accurately the complexity of the aid-giving process. Table 4.2: Gross loan and grant disbursements by Aid-India Consortium (5-year averages) US$m 1976–7 to 1980–1 (average) Bilateral

Gross loans

1981–2 to 1985–6 (average)

Grants

Gross loans

Grants

Austria

1.6



2.8



Belgium

10.2

0.1

4.8



Canada

24.8

22.4

37.0

12.1

3.8

12.8

12.6

14.2

Denmark France

36.2



46.4



156.6

3.9

144.8

0.9

Italy

4.4



3.4

2.1

Japan

95.4

12.6

102.2

11.0

Netherlands

75.2

17.2

43.0

36.3

Norway



18.1



20.8

Sweden

1.6

56.7



47.4

W. Germany

Switzerland

3.2

9.1

0.6

16.0

UK

12.0

229.0



155.3

USA

56.2

132.5

52.2

106.8

Total

481.2

514.4

449.8

422.9

Multilateral IBRD

146.8

359.8

IDA

478.2

927.8

EEC

37.9

76.6

UN

94.4

109.7

IFAD





Other

12.2

Total

637.2

132.3

1305.4

186.3

1,118.4

646.7

1,755.2

609.2

Ald-India Consortium

17.8

Source World Bank (1987) India: an Industrializing Economy in Transition, Report

6633-N, vol. 3, Tables 4.3(b) and 4.3(g) adapted.

The Systemic Effects of Aid and Donor Procedures

101

From Table 4.2, one can conclude that, over the decade 1976–86, Consortium aid was more multilateral than bilateral. Multilateral aid to India at this time was almost exclusively project aid. When the World Bank began after 1980 to offer balance of payment support through its Structural Adjustment Lending (SAL) facility, it was recognized both by the Bank and by the Indian Government that such a facility for non-project aid would not be appropriate for India. This form of lending carried with it conditions, that is, specific economic policies which the borrower promised to adopt in return for the right to borrow. It is difficult to think that the Indian Government would have relished borrowing on such terms, and difficult also to imagine that the Bank would have wanted to resort to sanctions, if India had borrowed but failed to meet the conditions of the loan. In the absence of SAL (or similar policy-based) lending, most multilateral aid has been of the traditional project type. The procedures for multilateral project aid are therefore the most important determinant of the systemic effects of Consortium aid. Major types of problem arising from multilateral project finance have been classified by Harvey (1983, 23–6) as follows: Currency problems The World Bank does not lend dollars just because its loans are denominated in dollars. It lends those currencies which it has borrowed on world markets, and thereby insulates itself from problems of currency fluctuation. The currency of its Bank borrowings does matter to the borrower, because that is the currency which has to be repaid. Thus if it is a hard currency, which subsequently strengthens against other currencies, repayment will become more onerous. Procurement problems Multilateral loans normally include a condition that procurement financed by these loans is subject to a procedure of international tendering. This is designed to give all the contributors to the multilateral finance (and indeed all other countries’ suppliers) an equal opportunity to win the contract, subject only to normal commercial criteria. The recipient country may believe that, for various reasons—including the need to encourage domestic industries—a full international tender procedure is not appropriate. Delays Delays arise from the lengthy project cycle of identification, preparation, appraisal, and negotiation of project lending. This cycle requires agreement between the multilateral institution and the borrowing government and its agencies on a very wide range of matters thought likely to affect the ultimate success of the project (Baum, 1982). Borrowers thus face a choice between accepting substantial ‘interference’ in project design in order to reduce delays, or accepting substantial delays in order to gain more agreement to their own conceptions of correct project design.

102

Does Aid Work in India? Project bias

When the multilateral lender lends only to cover foreign exchange costs and not the local costs in domestic currency, there may be pressures towards choices of excessively capital intensive and/or excessively import intensive techniques, despite a general awareness that an ‘appropriate’ technology is what is required. The currency aspects of multilateral project finance have not been a major source of difficulty for India. There are various reasons for this. First, the currency exposure problem was essentially confined to the 1970s. Under the regime of fixed exchange rates which lasted until its collapse in 1971–3, the exposure problem was confined to the risk of an occasional revaluation: it was rare and dramatic rather than chronic and continuous. After 1980–1, the Bank changed its currency policy, pooling its available currencies and lending to each borrower roughly the same mix of currencies. This spread each borrower’s exposure into a basket of currencies, and virtually extinguished the exposure problem. Even during the 1970s, however, India was in a relatively fortunate position to deal with currency risks from aid loans. As a regular borrower, any inability on its part to forecast the currency of the loans would have been offset by the acquisition of a basket of different currencies through repeated borrowing. As a borrower with good access to IDA funds in the 1970s, India enjoyed long grace periods before repayment started, which allowed the debt burden to be eroded by inflation. These two favourable factors moderated considerably the currency exposures generated by multilateral finance in India during the 1970s. The procurement conditions of multilateral loans, like currency exposure, posed problems most severely in the 1970s. This was because the requirement for international tendering is most obviously in the interest of the aid recipient when the project is of a con ventional turnkey kind, for which there is no domestic manufacturing capability. Examples are power stations or textile mills. Such projects dominated multilateral lending in the late 1950s and 1960s and international tendering therefore provided a substantial safeguard in securing value for money for India, particularly compared with bilaterally funded projects. However, as the purposes of multilateral finance began to change in the early 1970s, and as India’s own productive capacity diversified under the impact of import substitution, so the advantages of the international tendering rule became less clear-cut. Some friction between the Government of India and the multilateral agencies ensued. As agriculture (and specifically irrigation) grew in importance as a sector for multilateral aid, so some disagreements occurred between the Government of India and the World Bank about whether the rule requiring international competitive bidding should apply to canal construction. The Bank wished the rule to apply in this case, while the Government of India considered that value for money would be sufficiently ensured by competition for these contracts among domestic civil engineering firms. For a period in the 1970s, this failure to agree may have influenced the Bank’s lending policies, leading it to favour a strategy of land levelling and channel construction rather than heavy investment in main canal systems (Wade, 1982,172). More generally, the Bank’s international tendering rule has been a controversial one in India. Critics, both inside and outside India have made a variety of claims about its effects on project choice. These claims are often contradictory, but that does not allow us simply to ignore them. It is sometimes argued that, where India has the capability to produce a

The Systemic Effects of Aid and Donor Procedures

103

required project component, the domestic version should be purchased, in order to ‘reduce imports of technology’ (Economic and Political Weekly (EPW), 1982,1590). The question of relative cost and value for money is simply ignored in this line of criticism. This surely is not in India’s interest. A different argument is that the Bank unduly favours foreign main contractors, even when Indian bidders are equal in competence and competitive in price and that the choice of a foreign main contractor introduces a bias in favour of imported equipment. Thus, of three equivalent fertilizer plant projects (Nangal, Sindoi, and Haldia), the first two, financed by multilateral aid had a 25 per cent higher foreign exchange content than the latter, which was bilaterally funded and undertaken by a domestic main contractor, the Fertilizer Corporation of India (Payer, 1982,143; EPW, 1982b). This alleged bias is interpreted as an attempt to hobble India’s development of an independent technical capability. The Bank’s position is that the imported equipment is more advanced, more reliable, or less expensive, and that fair foreign competition is a stimulus, not a brake to the transfer of technology. This dispute thus boils down to the question of whether the existing procedures ensure a ‘fair’ competition. It does not seem useful to try to come to a general conclusion on this. All one can say is that conclusive evidence of systematic unfairness is lacking. On the whole, the GoI does not incline to endorse these more sweeping criticisms of international competitive bidding. Certainly the Ministry of Finance sees this practice as helpful in limiting the pressures which arise from domestic suppliers which have spare capacity and which believe that they are ipso facto entitled to government orders, regardless of their high production costs and prices. It is a way to avoid the perpetuation of the effects of past unwise investment decisions. The Ministry of Industry naturally has a different perspective, being responsible for the entities created by past investment. In an important round of discussions about a new set of super thermal power stations, it found itself taking up cudgels for Bharat Heavy Electricals against the Ministry of Energy but it proved to be fighting a losing battle (EPW, 1982a). The GoI’s anxieties about international competitive bidding have thus been effectively limited to peripheral areas, such as civil engineering or irrigation. By the early 1980s, these anxieties had been resolved successfully by a change in World Bank policy on procurement conditions. A 15 per cent margin of preference was granted to domestic contractors in both areas; in order, as the Bank’s semi-official pamphlet describes it, ‘to foster the development of local capabilities’ (Baum, 1982, 21). However, international competitive bidding remains the means by which the Bank believes that efficient and economical procurement can be achieved ‘in most cases’ (Baum, 1982). Systemic aspects of bilateral aid Multilateral aid is free from the problem of source-tying precisely because of its requirement in most cases for international competitive bidding. In contrast, bilateral aid’s major disadvantage for the recipient is that it is ‘tied’ in various different ways. Sourcetying is the requirement that the foreign exchange value of the aid should be spent only with the donor country’s national suppliers (or, in some cases, a restricted set of other specified geographical sources of supply). Project-tying is the limitation that aid must be spent only on an agreed, identifiable ‘project’. (Reverse-tying is advantageous for the recipient, because it allows loan repayments to be made in commodity exports rather than

104

Does Aid Work in India?

free foreign exchange.) Bilateral aid is less project-tied than multilateral aid in India, as Table 4.3 shows. But it is seriously source-tied, in a way that multilateral aid is not— see Table 4.4. The bilateral donors’ reason for the double-tying of aid is primarily the need to demonstrate in the domestic political arena the concrete results of aid, which is much easier with the ‘project’ format, and the visible list commercial contracts from aid projects. However, donors are also concerned that their aid should secure their national suppliers additional exports, that is, exports over and above the level which would have been purchased in the absence of aid. At first glance it might appear that extra exports must necessarily be generated by double-tied aid. This is not the case, because aid is to some extent fungible in the hands of the recipient, i.e. the recipient may be able to use aid money for a project which would have been undertaken anyway in the absence of aid, and thereby free resources for another purpose (unspecified). When aid is fungible in this sense, the giving of aid may not produce additional exports for the donor country—it would depend on what the other unspecified purpose or project actually was. Naturally, given that national commercial advantage is a major political motivation for aid-giving, donors are unhappy about fungibility, which even double-tying does not always eliminate. Table 4.3 Gross disbursements of Consortium aid in 1979–80 (US$m) Consortium member

Project

Austria

Debt relief

Food

1.4

Other 0.4

Belgium

7.8 17.1

Total 0.4 9.2

Canada

22.9

40.0

Denmark

12.1

3.8

15.9

France

29.2

11.8

41.0

W. Germany

91.5

56.1

147.6

4.0

23.5

81.0

0.4

99.5

99.9

10.4

20.8

Italy Japan

53.5

Netherlands Norway Sweden UK USA

10.4 7.7

33.5

41.2

102.4

149.2

251.6

56.0

165.3

385.7

IBRD

148.7

148.7

IDA

546.7

546.7

Sub-total

695.4

695.4

Total

1,081.1

5.8

5.8

1824

221.3

Sub-total

1824

396.1

396.1

970.0

1,665.4

Source World Bank (1981), Economic Situation and Prospects of India, Report 3401-IN, p.241, Table 4.2(a).

The Systemic Effects of Aid and Donor Procedures

105

Purchases made with tied aid funds tend to be more expensive than if the same items were acquired by worldwide unrestricted sourcing. Even those commentators who claim that multilateral aid has no advantages over bilateral aid have to acknowledge that tying ‘reduces the real value of aid below the nominal value’ (e.g. Bauer, 1984, 66). Those who have tried to estimate the extent of price inflation as a result of aid-tying in the Indian case come up with estimates of price differences of between 20 and 30 per cent (Riddell, 1987, 209). If these figures are anywhere near correct, and they have not been seriously challenged, then it is clear that source-tying of aid provides a means of support for donor country export industries which are relatively uncompetitive in the world market. This inference is consistent also with the data in Table 4.4 on inter-country differences in the extent to which bilateral aid is source-tied. It is clear that, with the sole exception of Italy, countries which are relatively liberal in their tying arrangements are either those which are very firmly committed to progressive aid policies (the Scandinavian countries, the Netherlands), or those who have reason to regard their export industries as very competitive across the whole range of exports (West Germany, Japan). Table 4.4 Tying status of DAC bilateral aid, 1982–3 (Percentage of bilateral ODA gross disbursements) United

Consortium member

(1)

Of which: Technical co-op (2)

Of which: ‘Cash’ (3)

Partially untied

Tied

(4)

(5)

Of which Technical co-op (6)

Austria

3.3

1.0

0.4



96.7

14.6

Belgium

24.7



4.6



75.3

53.4

Canada

17.0

3.2





83.0

13.0

Denmark

63.9

40.2

19.0



36.1

7.5

France

36.6

11.3

19.3

9.1

54.2

45.3

W. Germany

70.4

19.9

13.1



29.6

14.7

Italy

54.8

31.7

11.7

1.1

44.1

3.3

Japan

55.3



3.5

21.4

23.2

13.4

Netherlands

57.1

33.5

17.2

25.9

17.0



Norway

65.1



56.0



34.9

16.5

Sweden

81.2

24.1

45.7



18.8



Switzerland

66.0

5.6

60.4



34.0

12.9

UK

23.7

0.5

23.1

1.1

75.2

35.9

USA

36.8



36.8

16.5

46.7

11.4

Source OECD (1985) Twenty-five Years of Development Cooperation: i Review, Paris; adapted from Table 1,244 Note DAC members who are not Consortium members (Australia, New Zealand, and Finland) are not included. Tying status relates to all bilateral aid, not just aid to India

106

Does Aid Work in India?

What effects does this variety of tying behaviour have on the Indian system of planning and budgeting? As previously indicated, once the GoI has approved a project for inclusion in its development spending budget for a given year, it attempts to implement it to plan, without close regard to the timing of the arrival of the foreign funds intended to finance it. In the period up to the mid-1960s, the long delay between the authorization and the utilization of aid tended to be blamed on the fact that much bilateral aid was project-tied and source-tied. Certainly, at that time the Indian administration was in the process of learning how to manage aid inflows sensibly, and from 1966 onward, the utilization rate of authorized aid markedly improved (Sukhatme, 1986, 20–1). Despite this improvement, it still remains normally the case that aid receipts arrive after the project outlays have been made. It is possible in principle that delays in the arrival of aid receipts could cause delays in the implementation of the project. But that would normally happen only if, in a particular year, the country was experiencing an overall foreign exchange constraint. In many recent years, that has not been the case. In the absence of an overall constraint, the existing stock of foreign exchange can be managed to cover the interval between the project’s requirement for foreign exchange and the arrival of the earmarked foreign funds. A number of more difficult management problems do, however, arise in connection with offers of bilateral project aid. An important one is how to choose which of the available project aid offers to accept. When donors express interest in financing a particular project, the Ministry of Finance has to decide whether to accept in the light of estimates of the effects which the donor countries’ tying procedures will have on the prices charged by their countries’ supplies. Clearly, a multiplicity of donors, with a wide variety of different tying procedures, makes this a complicated choice. Choosing correctly requires a detailed and up-to-date knowledge of competitive world prices of the relevant products. India is much better placed than many small African countries to marshall this information and to act upon it. She has indeed on occasions refused offers of aid on the grounds that the donor’s suppliers’ price quotes were highly inflated and at least one case was reported to us where such a refusal was quickly followed by a renewed offer at half the original quotation! But there? also cases reported in the late 1970s when India accepted bilateral aid almost reluctantly and without enthusiasm because of the combination of the low priority of the project and the inflated price of the goods: purchase of the UK merchant ships with bilateral aid might be cited in this connection. The Indian Government successfully created the impression that they were doing the UK a favour by accepting the ships, rather than that the UK was doing India a favour by giving the ships. Indian aid officials are said by some to believe that they can out manoeuvre bilateral donors (Hayter and Watson, 1985, 245). On this occasion at least, such a belief was vindicated. Behind the public manoeuvrings, however, lie serious issues about the distortions which tied bilateral aid can cause. One often encountered is a bias towards excessively capitalintensive and foreign exchange-intensive projects. Such a bias was evident in Indian plans of the 1950s and 1960s. But it was almost certainly not caused by bilateral aid, although the latter may have given it some minor sustenance. The chosen development strategy of rapid industrialization without careful regard for employment considerations was the major reason for excessive intensity of capital and imports in early Indian development projects. Bilateral aid merely underwrote this strategy at the margin. Another frequent distortion arising from tied aid is the absence of normal standardization in imported equipment,

The Systemic Effects of Aid and Donor Procedures

107

with consequent problems of raising the local cost requirements for maintenance and staff training. This distortion is most serious in small countries with no indigenous capacity to manufacture capital goods. The Indian strategy of rapid industrializalion was helpful in ameliorating this problem. The proliferation of imported equipment types was often followed by the licensing of foreign manufacture of a reduced number of types, and finally by the emergence of a small number of successful suppliers, either foreign or Indianowned. The sequence did not always work smoothly or in the manner expected—but visible progress was made in standardization, as for example in the production of an indigenous tractor (Bhatt, 1979). Thus on the whole, aid has not been responsible in India for these two well-known non-price distortions, which often accompany aid in African economies. Nevertheless, the onset of recession in the late 1970s undoubtedly intensified the complexities of choosing between alternative tied offers. As spare industrial capacity increased rapidly in donor countries, some tying practices have become stricter, countries like West Germany which were liberal have become resentful of the tying practices of others and entirely new forms of commercially-oriented aid have been brought into existence. Examples are the mixing of aid with export credits and the creation of new ultra-soft loan facilities. These practices are embodied in France’s credit mixte scheme, Japan’s soft loan finance, Canada Aid Trade Fund and the UK’s Aid and Trade Provision (ATP) (Toye and Clark, 1986) India, like other recipients, has been affected by these general trends. Though the share of tied aid has not risen, the cost of not being able to spend a unit of aid freely has gone up, because in a more competitive world market better bargains are available for those who can shop where they wish. Three additional dangers arising out of the recessionary pressures should be noted. Commercially-oriented aid begins with one donor country firm identifying a business opportunity for a capital goods sale in the recipient country. It is ‘firm-led’ for that reason, with the donor government acting merely as a facilitator of export sales. In this situation, there is real difficulty in ensuring genuine competition between all competent firms within the donor country. Firms may even enter into informal understandings about which aidfinanced export sales are to be regarded as ‘theirs’. When this happens, the aid is effectively ‘triple-tied’—by project, by national source and by firm. Donor country governments have no particular incentive to prevent this happening. An export sale is an export sale, regardless of which firm makes it and, if the price is inflated, this is positive from the donor government viewpoint because it actually improves the export statistics. This situation can be a tempting opportunity for corruption, and the manipulation by unscrupulous firms of both donor and recipient aid processes. Commercially-oriented aid can also lead to highly inappropriate purchases. Its standard justification is that ‘even the poorest countries need railways, power stations, transmission equipment, construction materials, or water supply and irrigation systems’ (Hurtado, 1984). No doubt they do. But inappropriateness does not normally manifest itself at that level (although commercially-oriented aid has in fact been used to sell sophisticated aircraft to very small, very poor, African countries). Inappropriateness arises because, in the drive for sales, the recipient country is sold the wrong type of bus, power station etc. for its environment; or it is sold the right type for an environment which itself needs changing. India has certainly become increasingly involved with this type of commercial aid from the UK. The £230m Rihand coal-fired power station (1982) and the recent £150m captive

108

Does Aid Work in India?

power station for the Bharat Aluminium Company (1984) are examples. Here the UK Government was more anxious about gaining orders for GEC and NEI than about the development consequences for India. The third danger is that aid agencies alter their policies in the non-commercial component of their aid, with the aim of gaining commercial advantages. The aid financing of India’s purchase of twenty-one Westland helicopters in 1985–6, at a time when the UK company was on the verge of collapse has been cited as an example of the osmosis effect of ATP aid on non-ATP aid (Toye and Clark, 1986, 312, note 1). A single-firm deal, which puts the presumed interests of the firm ahead of the likely developmental benefits for the recipient could become more acceptable in the normal bilateral aid programmes of donors in the wake of the expansion of explicitly commercial aid schemes. In India, this may already be happening. Apart from this spillover effect, there are other ways by which donors can design aid programmes to capture extra commercial benefits. It is possible for donors to make rapid switches between project areas, moving funds from one to another every two years or so. This leaves the recipient locked into the purchase of further capital equipment and maintenance imports which can only be bought from the donor country and which have to be funded with free foreign exchange. India appears to have experienced this problem also with one or two donors. If source-tying remains the major negative systemic feature of bilateral aid in India, what are the prospects for its significant relaxation? One frequently suggested method is the so-called ‘additionality alternative’. This is a proposal that aid-recipient countries should endeavour to ensure that they purchase additional exports from the donor country of a value related to the amount of aid given. ‘Full additionality’ would imply extra exports equal to the value of aid, but lesser ratios might also be mutually agreeable. As is clear from the discussion at the start of this section, the additionality alternative is—at least in the form of full additionality—an attempt to get to grips with the fact that aid is to some extent fungible, even when double-tied. It is far from clear that this is possible, even in principle, for it seems to require counter-factual knowledge, i.e. knowledge of the level and composition of imports under ‘normal’ trade relationships, defined by the absence of aid. Apart from this, serious problems would arise in the mechanics of carrying out the additionality undertaking. In India, the private sector has the dominant role in making import decisions, and there are powerful arguments for further relaxing import controls rather than reinforcing them to achieve additionality objectives. Within the public sector’s direct area of control over imports, additionality would have to operate mainly with respect to bulk items rather than capital goods and components; it might interfere with procurement associated with multilateral aid finance, where as already discussed, competitive international tendering is required and it would be a positive disadvantage where the donor country had an over-valued currency—because then all exports bought additionally from that country would be over-priced. All in all, the bureaucratic requirements of ensuring that additionality is effective appear to be insurmountable even in a relatively well-organized country like India. This judgement is confirmed by India’s brief experience. Between 1967 and 1969, the USA attempted to impose the concept of additionality on to its bilateral aid relations with India. It did not succeed: this ‘episode was short-lived but costly in economic terms and in political good will’ (Sukhatme, 1986,19). There does not seem to be any realistic way of getting around the aid fungibility problem.

The Systemic Effects of Aid and Donor Procedures

109

If additionality is not a genuine alternative, some donors may cling to source-tying as a surrogate safeguard for their commercial interests. Other donors, more subtly, may proceed with formal un-tying, in the expectation that it will have little impact on the export benefits they achieve through aid. Just as, when tariffs are dismantled non-tariff barriers can have a similar protectionist effect, so the formal untying of aid can be neutralized by other non-tying means. Two examples of such ‘other means’ are worth remarking. First, even when bilateral aid donors do not source-tie their aid, they usually do not adopt the same requirements for international competitive bidding as the multilateral donors. It is often the case that invitations to bid for bilaterally aided projects are publicized only in the donor and the recipient country. Thus if the project calls for equipment in which both donor and recipient countries’ industries are internationally uncompetitive, the economic benefits of untying the aid will be effectively frustrated. There is a case for changing donor procedures with regard to publicity and tendering to bring them fully into line with those of the multilateral agencies to ensure that untied bilateral aid produces the best value for money for the recipient. Second, the role of project consultants is also likely to be important. Consultancy assistance, under bilateral aid, is almost always provided by donor country nationals, selected by the donor agency. The use of donor national consultancy firms can often result in procurement being tied de facto, although it is not tied de jure. For the import-content of project aid, therefore, some of the untying which the statistics show may be illusory. Local cost finance As we have seen in Table 4.3, about two-thirds of aid to India is normally project aid. Apart from international tendering and problems arising from tying, the chief remaining cause of friction in the project aid relationship concerns local cost finance. Tied aid automatically precludes local cost finance, but even when aid is not linked to a given geographical source of supply, it may still be given for spending on imports rather than domestic expenditures. These expenditures may be either complementary capital expenditure or maintenance and running costs. Historically, aid agencies have fought shy of meeting such costs, for two very powerful reasons. They have argued that long-term political support for aid in donor countries depends on not eroding the commercial benefits which flow back to their national suppliers. Secondly, they see local costs as a continuing commitment which they are unwilling to accept in the light of the legislative requirements for public expenditure control in donor countries. A number of Consortium countries, including Belgium, West Germany, Japan and the USA are known for their reluctance to finance local recurrent costs (Jennings, 1983, 514–5). There are various ill-consequences of this reluctance. Firstly, the search for local counterpart funds can seriously add to the delays in the implementation of a project. Thus: It would be reasonable to assume that a shift in a donor’s policy with respect to the provision of funds for local expenditures would enhance an aid agency’s capacity to minimise the time required for project completion. (OECD, 1981, 12)

110

Does Aid Work in India?

Second, the absence of donor finance for local expenditures can distort investment programming, with too many resources directed towards new capital investment and too few towards the maintenance in good operating order of previous slices of investment. Third, lack of local cost finance biases investment against projects which directly and explicitly benefit the poor. This last effect has been extensively commented on. For example, it has been stated that: It is possible to cite a number of cases, (including) all UK aid to India until the introduction of Retrospective Terms Adjustment, in which a country-specific upper limit on the extent to which aid could be given in local cost form caused potentially poverty-focussed projects… to be delayed, truncated or else simply passed over in favour of other projects with a lower local-cost content. (Mosley, 1981, 221)

When discussing the extent of restrictions on local or recurrent cost finance in aid to India, it is again necessary to distinguish between multilateral and bilateral aid sources. For multilateral sources, particularly the World Bank and IDA, the pressures for greater local cost financing of projects were held at bay in the 1960s by the IDA decision to grant India ‘programme aid’. These were credits to be used for miscellaneous industrial imports. Programme aid released domestic resources, which would otherwise have been required for these imports, to be used to meet the local costs of aided projects. The Bank itself, however, did not view this IDA experiment very favourably, because it did not see its function as one of giving balance of payment support only distantly related to specific projects, particularly as it harboured some doubts about the overall success of Indian planning. It therefore gradually relaxed its criteria for providing local cost finance for projects. Gradual widening of the criteria has had the effect that local currency financing became the Bank’s preferred method, compared to programme loans, of contributing to the foreign exchange component of development programmes sums over and beyond those required to buy the imports on Bank-financed projects. (Van de Laar, 1980, 48–9)

How much local cost finance is permitted in any Bank loan is a matter of judgement by the Bank in relation to individual loans and the overall portfolio of loans to a country at the relevant time. The need for the borrowing country to have an important stake of its own in any project is one key influence in the judgement, as is the Bank’s wish to be able to shape the organization and execution of the project and the policies in the sector to which the project belongs. In certain areas of recent lending to India, the Bank has been criticized for not getting this balance right In its US$200m package of loans for agricultural extension between 1977 and 1982, the Bank’s loans amounted to half of the total costs of the projects. It is suggested that this aid ‘has committed the Indian taxpayer to considerably enhanced recurrent expenditure on agricultural extension once the World Bank disappears from the scene’ (Moore, 1984, 2,12–13). This is not disputed by Bank officials (e.g. Feder and Slade, 1984, 35). They argue, however, that the new style of extension is more effective than the old, although the increment in effectiveness and the cost/benefit ratios have not yet been comprehensively assessed. But it is on such assessments that the affordability of these aided projects will depend.

The Systemic Effects of Aid and Donor Procedures

111

For Bank Group aid, the issues of local cost finance and international competitive bidding were linked to each other. Even when a specified portion of local currency financing had been agreed for a project loan, it could be reduced as a result of the ICB requirement, if foreign contractors could compete effectively in activities which had been set aside for local expenditures—e.g. construction. This was one of the factors that fuelled the dispute between the GoI and the Bank over the ICB rule in the early 1970s. This dispute was effectively settled in 1973 with the grant of a 7.5 per cent margin of price preference to Indian domestic civil works contractors. This has since been raised to 15 per cent. Aid from bilateral sources, because of its initial high degree of tying, was much slower in making provision for a local cost element than was multilateral aid. For example, as late as 1977, only 2 per cent of UK aid to India was local cost aid (House of Commons, 1979, xxvii). A major change of UK policy occurred in 1979, when significant debt relief was granted to India (the so-called ‘Retrospective Terms Adjustment’ or RTA). Debt service payments were made available to India in rupees to meet local costs associated with UK aid projects. As a result, local cost aid rose to over 30 per cent of UK aid to India in 1982–3 and is likely to remain at approximately that share for the next twenty years. This is much more closely in line with the local cost proportion characteristic of Bank lending. When US aid to India was resumed in 1978, it was even more dramatically ‘untied’. India was allowed to spend aid funds either in India, or in the US or in a developing country. The only sourcing that was definitely ruled out was from suppliers in developed countries other than the US itself. Even this might have proved restrictive if the projects supported necessarily had a high foreign cost element. But that has typically not been the case. Project choice has centred on irrigation, social forestry, and primary health care, including family planning and child nutrition. As a result, by far the largest part of current US aid now consists of local cost finance. These local cost expenditures include some recurrent costs in certain health and irrigation projects. It has been thought necessary to offer this in order to get the projects concerned accepted by the GoI. Fears about excessive commitment have been met by stipulating a maximum percentage for recurrent costs and entering into an understanding with the GoI that the recurrent element will gradually taper off over time. Clearly, here is a bilateral aid example of committing the GoI to enhanced recurrent expenditure in certain fields once the pump-priming of aid has been completed. Again, whether this is or is not a problem depends on the social profitability of the aided projects in the longer term. As the percentage of local and recurrent costs has grown in aid finance in the 1980s, increasing attention has been paid by donors to the cost recovery aspects of projects. Taxation reform is again on the Bank’s priority agenda (see, for example, Newbery and Stern, 1987). But improved tax raising is seen as only part of the solution to the sustainable financing of recurrent expenditures. In several important aided sectors in India, cost recovery has not been good. Repayment of rural credit and of housing loans in site and services urban projects have been notorious for low recovery rates. But by moving away from foreign costs-only rules, the donors have almost necessarily involved themselves in seeking to raise cost recovery rates across a wide spectrum of public activity, including irrigation and transport, health and education. This involvement brings with it some difficult decisions which centre on the need to reconcile cost recovery with the povertyorientation of projects. This is an area which will demand much detailed empirical analysis

112

Does Aid Work in India?

of beneficiaries’ genuine ability to pay, or re-pay, as well as sensitive judgement on where to strike the balance between financial sustainability and poverty alleviation. Donor procedures Since the mid-1970s, donors have become increasingly aware that their own procedures of aid-giving may impose significant avoidable cost on the recipient of aid. By ‘procedures’ is meant the whole cycle of political and bureuacratic activity that accompanies the actual transfer of resources. It is a wide-ranging term covering the donor country’s method of authorizing its aid budget, making aid policy, programming aid by country and sector, project preparation and approval, the actual disbursement of funds, and the ex-post monitoring and evaluation of aid-funded activities. Complying with these procedures requires effort and resources from recipients, in terms of negotiation and documentation. Failing to comply also imposes costs, either in terms of delays to existing projects and programmes, or in terms of lost opportunities of gaining additional foreign resources. Some of these costs would appear to be imposed unnecessarily. Although these procedures all have the basic aim of the efficient mangement of aid funds, they assume a great variety of different forms as between different donor countries. The supposition must be that some standardization between donors is possible and would be beneficial in reducing recipients negotiation and compliance costs. The standard argument against this has been that standardization of procedures cannot be carried out without very difficult changes in the constitutional frameworks of the donor countries. However, a recent study of the problem concluded that this explanation was not entirely valid, on the grounds that: During the course of this study, a small number of donors have made changes in their procedures which appeared previously to be impossible from a legalistic point of view (e.g. multi-year commitments, field missions) without major revisions in the bureaucratic or legislative framework of their aid programmes. (OECD, 1981, 73)

Thus there does appear to be some scope for beneficial co-operation between donors, to agree standard formats for project appraisals, disbursement certifications, evaluation reports, etc. Project appraisal methodology has often been identified as a source of difficulty for aid receiving developing countries. This is partly because these methods are often too sophisticated for domestic planning capacity and partly because donors (e.g. the Bank, the EC, Sweden, and Japan) insist on the use of distinctly different appraisal methods (Murelius, 1981, 91–3). It is clear that these difficulties are much less severe for India than for a range of other countries, including Brazil, Peru, Nigeria, Tanzania, and Thailand. Nevertheless, the Indian Governments capability for project appraisal does not appear to be particularly strong. In the mid-1980s, the Project Appraisal Division of the Planning Commission consisted of only fifteen mainly junior professionals, handling a total of only 150 project appraisals per year (Little, 1987, 44). Despite the limitations in the scope and quality of the work of the division, its existence is sufficient to ensure that the Indian Government no longer sees the variety of donors’ project appraisal requirements as a major issue of aid policy, compared with the tying of aid. Needless, if donors can standardize their requirements for appraisal in a way that reduces India’s compliance costs, this will be welcomed.

The Systemic Effects of Aid and Donor Procedures

113

Especially difficult procedural problems can arise from having several donors involved in a single project—the ‘multiple-donor project’. The co-ordination requirements here are much more stringent than for the single-donor project and failure to meet them can lead to additional sources of delay in project implementation. The GoI has, as a matter of policy, tried to minimize its involvement with multiple-donor projects. On the few occasions when avoidance has not been possible, steps have been taken to reduce the dimensions of the co-ordination problem. When IDA has collaborated with bilateral donors, an attempt has been made to break down the whole project into components that are separable for managerial purposes, i.e. the completion of one part is not critical in the sequence of other parts and joint activity on a single site is not necessary. Nevertheless, some delays have occurred, because separability is rarely absolute, but these have been only minor irritants within the whole scheme of aid-financed projects in India. One important area of procedure where donor practice varies noticeably in India is with respect to the use of locally-based aid agency staff. The World Bank mantains a local mission in New Delhi as does USAID. The UK in marked contrast, has very few permanent staff handling aid in its High Commission, and relies on ad hoc visits of aid staff based in London or its Asian Development Division in Bangkok to handle the peaks of aid work. These differences derive from donors’ differing evaluation of the costs of local missions in relation to their benefits. The costs are considerable and have to be deducted from funds that could otherwise be used for resource transfers. The benefits are unpredictable but can be considerable. Clearly, if policy dialogue is to become a major feature of the aid relationship, there are great potential benefits in conducting it locally, continuously and through personnel who have been doing the job long enough to possess deep local knowledge and to be trusted by the recipient side. Having a local mission does not, unfortunately, always ensure that this happens. Three major difficulties are: (i) merely having a local mission does not ensure that it has the authority necessary to be effective. If headquarters has strong priorities and is not prepared to permit a realistic decentralization, the local mission will be frustrated in its attempts to mould country policy. It has been alleged, for example, that the Bank’s large resident mission in India is allowed only to supervise the implementation of Bank aid projects, and that the big decisions on initiation and appraisal are deliberately reserved for headquarters in Washington (Hayter and Watson, 1985,72). (ii) For a task where much depends on personal relationship, the right quality of top staff is essential to success. Lack of leadership will leave the mission ineffective. Here again, it has been alleged that all is not well within the Bank’s India mission, because of conflicts and jealousies between expatriate and local Indian staff (Hayter and Watson, 1985, 74). (iii) Familiarity may breed contempt of a local mission. Suggestions or criticisms may be discounted by the recipient government as being ritualistic only in nature. The Indian Government’s position (reflected by Patel, 1986, 215) is that the dialogue is mutually educative, and that India has learned, for example about integrated strategies for agricultural development, from the aid agencies. Off the record, Indian officials can give the impression that they believe that they are cleverer than aid officials, and can always out-manoeuvre them (Hayter and Watson, 1985, 245). None of these difficulties needs to be viewed as insurmountable. If the best way of making aid bear good developmental fruit is the policy dialogue, it does seem desirable that donors with an aid programme beyond some minimum size should take the trouble

114

Does Aid Work in India?

and expense to have an effective local mission. From a simple cost-reducing perspective, it would be economical and effective for aid donors to fund and staff a joint mission, sharing the costs and co-operating in an international search for high quality staff. Such a proposal, however, despite its technocratic attractiveness, would face an obvious political disadvantage. It could easily be seen by the Indian Government as the formation of a ‘common front’ of donors to put pressure on it for policy changes on the issues the size of the public sector, trade liberalization and the treatment of foreign private investment which have cause disquiet with western aid donors intermittently over the years. By relatively prudent macroeconomic management, India has avoided the situation of many developing countries, which are borrowing from the international financial institutions on terms which contain detailed and explicit conditions for policy changes. Even when it borrowed 5 billion SDR from the IMF Extended Fund Facility in 1981, the Bank played a very minor role and only normal forms of IMF conditionality applied: there was no question of Bank/Fund crossconditionality to bring about changes in India’s economic policies. Having successfully avoided such cross-conditionality in the aid process, the Indian Government would need to assure itself, before accepting a joint local mission of aid donors, that it was not also accepting aid cross-conditionality by the back door. References Bauer, P.T. (1984) Reality and Rhetoric: Studies in the Economics of Development, London, Weidenfeld and Nicolson. Baum, W.C. (1982) The Project Cycle, Washington DC, IBRD. Bhatt, V.V. (1979) ‘Indigenous technology and investment licensing: the case of the Swaraj Tractor’, Journal of Development Studies, 15:4, July. Cable, V. (1980) British Interests and Third World Development, London, Overseas Development Institute. Cassen, R. and Associates (1986) Does Aid Work? Report to an Intergovernmental Task Force, Oxford, Clarendon Press. Economic and Political Weekly (1982) 17:40, 2 October. ——(1982a) 17:44, 30 October. ——(1982b) 17:25 December. Eldridge, P.J. (1969) The Politics of Foreign Aid in India, London, Weidenfeld and Nicolson. Feder, G. and Slade, R. (1984) ‘A comparative analysis of some aspects of the training and visit system of agricultural extension in India’, Washington DC, mimeo. Harvey, C. (1983) Analysis of Project Finance in Developing Countries, London, Heinemann. Hayter, T. and Watson, C. (1985) Aid: Rhetoric and Reality, London and Sydney, Pluto Press. Hirschman, A.O. (1970) Exit, Voice and Loyalty, Cambridge, Mass, Harvard University Press. House of Commons (1979) The Pattern of United Kingdom Aid to India, First Report from the Select Committee on Overseas Development, Session 1978–79, London, HMSO. Hurtado, M.E. (1984) ‘The hard-nosed touch’, The Guardian, 11 May. Jennings, A. (1983) ‘The recurrent cost problem in the least developed Countries’, Journal of Development Studies, 19:4, July. Kalecki, M. (1972) Selected Essays on the Economic Growth of the Socialist and the Mixed Economy, Cambridge, Cambridge University Press. Kothari, R. (1970) Politics in India, New Delhi, Orient, Longmans. Little, I.M.D. (1987) ‘A comment on Professor Toye’s paper’, in Emmerij, L. (ed.), Development Policies and the Crisis of the 1980s, Paris, OECD.

The Systemic Effects of Aid and Donor Procedures

115

Moore, M. (1984) ‘Institutional development: the World Bank and India’s new agricultural extension programme’, Journal of Development Studies, 20:4, July. Mosley, P. (1981) ‘Aid for the poorest: some early lessons of UK experience’, Journal of Development Studies, 17:2, January. Murelius, O. (1981) An Institutional Approach to Project Analysis in Developing Countries, Paris, OECD Development Centre. Newbery, D. and Stern, N. (1987) The Theory of Taxation for Developing Countries, New York and Oxford, Oxford University Press for the World Bank. OECD (1981) Compendium of Aid Procedures, Paris. OECD (1983) Development Co-operation: Efforts and Policies of the Members of the DAC, Paris. Patel, I.G. (1986) Essays in Economic Policy and Economic Growth, Basingstoke, Macmillan. Payer, C. (1982) The World Bank: a Critical Analysis, New York, Monthly Review Press. Riddell, R.C. (1987) Foreign Aid Reconsidered, Baltimore and London, John Hopkins University Press and James Currey, in association with ODI. Sukhatme, V. (1986) ‘Assistance to India’, mimeo, September. Toye J. and Clark, G. (1986) The aid and trade provision: origins, dimensions and possible reforms’, Development Policy Review, 4:4, December. Van de Laar, A. (1980) The World Bank and the Poor, Boston, The Hague and London, Martinus Nijhoff. Wade, R. (1982) ‘The World Bank and India’s irrigation reform’, Journal of Development Studies, 18:2, January. World Bank (1981) Economic Situation and Prospects of India, Report 3401-IN for official use only.

Chapter 5 Project Aid to India

Role and evaluation of project aid Net aid to India, by 1985–6, was only 0.8 per cent of GDP. However, project aid matters much more than this suggests. As the Appendix to this chapter shows (pp. 204–6), in India (unlike most other LICs): (i) the role of foreign capital is best assessed by the ratio of gross aid to public development investment (21 per cent in 1985–6) or to imports (11 per cent); (ii) a large, growing part of aid is project-linked; (iii) its relatively hard terms (tied loans rather than grants) mean that project returns must at least raise GNP enough to cover loan repayments; and (iv) heavy concentration of aid on key sectors, past (steel) and present (irrigation), further increases the importance of good project performance. It might be objected that projects, nevertheless, do not matter—that only policy dialogue matters—because aid is money, and as such (even if nominally linked to a given activity) is fungible. If so, the recipient might merely use the donor’s money to divert its own elsewhere—i.e. the recipient might ‘put up for aid’ projects that it would undertake anyway, even with no aid. Thus aid allegedly finances, not ‘really ‘these projects, but marginal and only-just-worthwhile public expenditure, for which it releases funds. This objection is plausible for some small projects. However, for aided projects as a whole—or for any big project—the fungibility argument makes sense only to the extent that the recipient has ready access to other, non-aid resources: i.e. to the extent that aid forms a small part of a recipient’s total relevant resources. It would be absurd to suggest that all aided projects would be done even without aid in a country where aid provides most1 project finance; but aid at 5 per cent of gross investment, the current Indian norm, might seem too small to avoid fungibility. But 12 per cent of public investment is less clear; not all this 12 per cent would be done without aid. Moreover, most aid finances public development investment, of which it comprises about 20 per cent; this is, as a whole, surely not fungible. Such significant project-linked resources, accompanied by technical assistance with useful messages that earn respect, have plainly altered the project-mix. Far from being a poor second-best for effective policy dialogue, they have comprised a necessary condition for it. In principle, agreed and cash-linked conditionality could emerge from sectoral, not project, aid; in practice, both India and donors want to see that a proposed approach works on the ground, in the project. If projects fail, how can people be sceptical about them yet hopeful about ‘policy dialogue’? A project cycle comprises identification, preparation and design, appraisal, approval, implementation, and evaluation-cum-monitoring (Baum, 1978). If we are to go beyond exciting but atypical tales of triumph or disaster, we require a systematic ‘evaluation of evaluations’ of projects’ performance. DOI: 10.4324/9780203840153-6

Project Aid to India

117

The large number of Indian aid projects and donors compelled us to concentrate on to a few major donors. The World Bank, the USA, the UK Overseas Development Administration, West Germany’s Bundesministerium für Wirtschaftliche Zusammenarbeit, and the Directorate-General 8 of the European Commission kindly made relevant documents available to us, and we also reviewed two overall Swedish analyses of their aid projects (Swedish National Audit Bureau, 1976; Macedo and Eduards, 1987). However, the World Bank evaluations available to us were much the most numerous (38 evaluations of 43 projects), covered a longer period (completions in 1969–85) and cost more money (US$2.0 bn. of Bank Group aid, plus over US$5 bn. of GoI capital outlay). Also, these evaluations are closely comparable in method, and in our view are genuinely independent judgements. They include highly critical comments, and replies from GoI officials. Partly for this reason, they are not in the public domain. Therefore, although we concentrate on the World Bank evaluations, we have tried to avoid identifying specific projects or persons, and have not referred by title to project documents. Weaknesses of all the agencies’ evaluation documents should make us cautious about drawing conclusions from them. They were usually prepared after visits of a few weeks, made within months of completion. Later and fuller ‘impact studies’ often show different results. Evaluations rarely include points of comparison—whether ‘baseline studies’ of preproject conditions, or ongoing ‘control studies’ of otherwise similar, but non-project, areas. Thus changes in levels or distribution of output or income among persons reached by the project—even if correctly estimated at evaluation—may not reflect the results of the project.2 Surprisingly few of the evaluations, except those of the Bank group (The World Bank), assess the expected economic rate of return.3 Therefore, many evaluations read like laundry lists—seven objectives achieved, six not achieved, all to varying degrees, with no way of judging the net effectiveness of the resources used up by a project. Few evaluation documents look seriously at the project’s impact on poverty, income distribution, or even employment—despite explicit policy commitments, in most aid agencies, to emphasize these issues.4 The Bank group in 1986 decided to monitor the poverty impact of major projects even outside those conventionally seen as falling into poverty-orientated sub-sectors. Poverty monitoring is not easy; nobody wants bogus number-spinning. However, if the problems of isolating a project’s growth effects can be overcome (and economic-rate-of-return estimates made) in a rough-and-ready way, so can the problems of assessing its poverty impact Some evaluations tend to assess projects through the eyes, and with the statistics, of persons whose incomes directly depend on the projects’ public support or repute. Sometimes, aid agencies tend to evaluate only their smaller projects.5 The remedies are obvious; some are under way.6 Nevertheless aid project evaluation is clearly unsatisfactory—albeit less so in India than elsewhere. Some donors, notably the EEC, do not seem prepared to devote significant expert in-house resources to the task, nor to allow genuine independence to the assessors.7 Although cross-section studies of countries (Papanek, 1972; Mosley, 1987,133) allow us to be fairly confident that aid as a whole, at least to Asian countries, has provided growth, evaluation of aid projects has generally been insufficient to draw firm conclusions about what sectors, donors, regions, scales, techniques, or institutions are best, or worst, either

118

Does Aid Work in India?

at achieving growth or at reaching the poor. However, some significant lessons about what makes for good or bad aid project management—by donors and recipients—do begin to emerge from the Indian experience. A major priority, for donors concerned with aid-effectiveness, should be to make better use of evaluations. This need goes beyond ensuring that at least a sample of projects from all donors is evaluated comparably and competently, with proper baselines and controls, for their effects on growth and poverty (and for their institutional sustainability). Each aided project also needs to be compared in such respects with non-aided public-sector projects and with private-sector projects that include foreign lending, in roughly similar sectors and areas. Good evaluation is costly (one to three per cent of project costs; the UK spends below 0.1 per cent) and has a long lead time, especially as baseline data are essential to good evaluations. However, the above process could be made into the beginning of a proper feedback system. In this, a donor (or a LDC government), seeking to appraise, design, or implement a specific project, could readily collate information on project successes and failures in similar environments. At present, even the World Bank, with its rich store of evaluations, does not do this systematically; no other donor, as far as we know, does it at all; and there is no body of accessible project knowledge upon which an Indian State, or a smaller LDC (or donor), could usefully draw. India—with excellent research institutions, many varied projects, different donors, comparable aided and non-aided public and private activities, and some receptive governmental agencies—is the obvious place to start. Co-ordination by computer renders ready access to a body of knowledge feasible; widespread problems in SSA make it necessary.8 The World Bank Group’s project sample Projects and evaluations The Operations Evaluation Division (OED) of the World Bank kindly made available to us 38 evaluation documents, covering 42 Bank/IDA-aided projects completed in India. One evaluation (23 of Table 5.1) was published in November 1973; the other 37 between 1977 and 1985. Apart from one long-drawn-out project, the remaining 37 evaluations refer to 41 projects on which disbursements were commenced between 1969 and 1978, and completed between 1974 and 1985. Details of the projects appear in Table 5.1. There are substantial formal safeguards of OED independence in its evaluations and OED is often very critical of recent project practice. All the evaluations used here include project completion reports (PCRs), either by non-OED Bank staff (sometimes also involved at earlier stages of the project cycle), or by Indian executive agencies. Almost all the PCRs used here are supplemented by project performance audits (PPAs) by OED itself, often supported by outside consultants. Some PPAs involve only assessment of in-house documents, interviews, and writeups. Most projects, however, and all problem projects, were also reviewed in the field by OED missions before these PPAs were finalized. Various involved GoI and state agencies are shown the draft PCR and PPA. Their comments are added to the reports. The openness of the whole process is impressive; however, this might have to be reduced if there were much exposure of exactly who said what about which specific project; hence some necessary vagueness in what follows. In particular, it should not be assumed that any comments in this text have any official support, either from GoI or from the Bank.

Table 5.1 World Bank projects in India

126

Does Aid Work in India?

Table 5.2 summarizes the financial role of the World Bank in the large sample of its projects where we were able to see evaluations. The US$2bn or so of Bank Group credits, reviewed here, appear to have covered less than one-third of project costs (Cols 8 and 10). Only a small part of this ‘shortfall’ was due to cost overruns (Col. 9) and a tiny proportion to co-financing by other donors (fns. b, d, h, j). Most of it is due to the large role played by GoI in project finance. However, the figures are greatly influenced by four big infrastructure projects (Table 5.1, 22a, 22b, 24, and 25) in which the World Bank played a small role. If these projects are excluded, the total Bank input into Indian projects with full cost information, recorded in Table 5.2, was US$1.5bn, comprising 45 per cent of appraised total cost (US$3.3bn) and 40 per cent of actual total costs (US$3.8bn). Although not necessarily a random sample of the PCR-PPAs—either in timing or project-mix—the projects that we have reviewed roughly represent the Bank’s project performances and priorities. Moreover, these are not very different from those of nonBank aid flows (insofar as the latter can be assessed from the very limited information available). Good returns, long delays The first lesson from these projects is that overall performance has been satisfactory. Real returns expected at post-evaluation—normally economic rates (ERORs)—were assessed in 34 cases, and were in double figures in all but four; three of these four had major redeeming features. Where no rates of return were evaluated, 5 out of 8 reports portray a generally adequate performance. The four clear failures or near-failures (Table 5.1, 2, 32, 26, 21) absorbed 10 per cent of the US$1.97m of World Bank group credits evaluated in these documents, and 6 per cent of the approximate US$6.9bn total actual costs (see Table 5.2, notes g and k) of the sampled projects. A lower proportion of failures would have implied an unduly conservative attitude to project finance. Of course, the estimated real ERORs, etc., have major weaknesses. The rates evaluated for farm credit projects are overstatements, since—as one of the PPAs makes clear—some private investments, financed by these projects, would have been financed otherwise without them (but ERORs on most such projects would remain high even if this could be allowed for). Tractor components in projects show seriously overstated ERORs because comparison of net farm income between tractorized and non-tractorized farms neglects the higher degree of irrigation on the former (Agarwal, 1980, 1984). One evaluation complains that the EROR estimates often show a distressingly large ‘judgemental’ range. Four of these evaluations warn that their EROR estimates are especially risky because they had to be made very soon after project completion. Another four warn about their ERORs because these reflect sub-systems—parts of substantial infrastructural or socialservice programmes—whose performance cannot really be separated from that of other sub-systems. All the same, the numbers and comments suggest strongly that the true rates of return on projects embodying World Bank aid to India have, on the whole, been satisfactory. Time overruns, however, pervade these projects (Table 5.3), averaging 51 per cent of scheduled time for the 34 projects with available data. These time overruns delayed the operation of, and hence returns from, the projects. Normally this tends to cut back

42

1,974

137 j

465

74

35

347

559

338

Total

1,800

93 k

335g

74

35

347

559

338

Fully costed projects

4 5 Bank Group credits (US$m)

c

5,657

339

2396h

338e

57

739

113 1

617

Appraised

6,099

363

2334h

349e

64

974

121 1 c

754

Actual

6 7 Total project costs (US$m)

32

27

14

22

61

47

49

55

93

93

103

97

89

76

93

83

8 9 Col. 5 Col. 6 + Col. + Col. 6 7 % %

Notes a For projects in which both appraised and actual costs are given in the evaluation documents, b Two national (ARDC/NABARD) and nine state-level projects, c Includes US$111m from other donors, d Storage is included in ‘other agriculture’. e Includes US$5m from another donor. f Rail, ports, telecom, power. g This (and all subsequent figures in this row) are for four projects only, h Includes US$59m from 2 other donors, i Education, populations, sewerage, urban development j another donor agency provided a further US$11m, for a project not included in cols 5–10 (no appraisal or actual costs available), k This (and all subsequent figures in this row) are for two projects only, l The early, atypical Project 21 (Table 5.1) is here included only in this last row of numbers.

Total1

38

4

4

i

Social sectors

6

5

Infrastructuref

3

55

3

Other ag./forestryd

Ag. marketingd

11

8

6

6

3 Projects reviewed

6

6

b

Nos.

2 Evaluation documents

Irrigation

Ag. credit

Fertilizer

1 Sector

Table 5.2 Bank role in funding some projects

30

26

14

21

55

36

46

45

%

10 Col. 5 + Col. 7

128

Does Aid Work in India?

ERORs, particularly if inflation affect costs more than returns. However, time overruns sometimes reflect correct changes in project design—or even GoI’s decision to increase its input into an apparently successful project, so that completion exceeds 100 per cent, as with three agricultural credit projects in Table 5.1. Conversely, if a project is underfulfilled by x per cent (completion is 100 minus x per cent), then a time or cost overrun is worse than it looks; thus the last irrigation project listed in Table 5.1 overran by one-third of its planned completion time and cost 12 per cent less than expected, but was 35 per cent under-fulfilled. Table 5.3 attempts to standardize for these facts, giving time/completion and cost/completion ratios—and also, in the data for ‘all projects’ in each sector for which calculations are possible, to allow for the fact that delays or cost escalations do more harm for big than small projects. Nevertheless, the picture of delay and under-completion is worrying. A unit of ‘real’ project implementation appears to be taking 1.5 to 2.5 times as long as planned, except in rural credit (1.25 times as long). Comparison of the sixth and the last columns of figures in Table 5.3 shows that the corresponding rise In cost, per ‘real’unit completed, is considerably less serious, except for irrigation projects; in other sectors, delayed completion probably affects mainly ‘finishing touches’ which do not add so much to costs. The relatively good timekeeping and cost-containment of credit aid are, of course, the counterparts of weaknesses: bad recovery of overdues, and little certainty that the extra credit (let alone the aid for it) is indeed necessary and sufficient to add the farm-level components assumed at appraisal and evaluation. Even if persistent delays and cost escalations do not greatly harm a project’s EROR or poverty-focus—and on the latter effect we have little evidence—it is desirable, if feasible at reasonable cost, to reduce them. They must damage financial and personnel planning, not only in donor agencies, but also in GoI, which pays much of the cost of these projects. If trains are persistently late, perhaps the timetable should be changed. The trouble is that one can seldom predict how late a train or a project will be, but Table 1 does suggest that timetables have been more realistic (i.e. delays are less) on recent projects, In any case, we need to know when, and why, delays did not harm ERORs. Often, it was because the projects’ output prices so outpaced input prices that the gap between the two increases (raising EROR) outweighed the effect of delayed sales of output (reducing EROR). This effect is reported for three fertilizer projects and four others. However, first, rises in a sector’s output prices—creating a good EROR estimate at evaluation, e.g. after an oil shock has increased output values from a fertilizer project—can be reversed later; impact studies of the price history of aided projects, comparing input and output prices with those prevailing at appraisal and evaluation, could improve future project selection. Second, price forecasting at appraisal requires more attention; it has not always been consistent with price projections from the Bank’s own commodity specialists. Third, the sensitivity of expected returns on an aided project to errors in price projections should be explored at appraisal and evaluation. This is especially important if a country looms large in a particular market, and/or if there is much investment, by the country or overall, in producing for that market.

Number Actual of proj- cost of ects projects (US$m)

Absolute data (averages) Time overrun (%)

For all projects (actu- Absolute al-cost weighted) data Cost overrun Time % overrun (%)

(averages)

Cost overrun (%)

For all projects (actual-cost weighted)

Fertilizers 5 723 57.2 157.2 53.4 153.4 27.8 127.8 21.9 122.5 Agricultural credit 11 1,211 60.5 127.4 30.0 117.4 9.5 99.5 17.9 111.2 Irrigation 3 453 41.3 229.2 42.6 231.6 51.7 248.0 70.0 276.5 Agricultural 3 64 55.0 278.7 49.4 259.5 18.3 173.3 18.1 167.1 marketing Other agri3 126 76.0 180.0 47.1 151.2 2.0 103.3 −10.2 94.4 culture Infrastructure 2 1,745 41.0 203.5 26.2 164.7 6.5 140.5 8.2 129.4 Method Time and cost overruns (%) are the averages for the relevant projects from Table 5.1 Time/completion (cost/completion) ratios, x 100, show the ratio of ‘time (cost) to complete project’ to ‘time (cost) planned at pre-appraisal’, divided by ‘proportion of project completed’. Thus if a project is 10% overfulfilled but costs 12% more than planned, its cost/completion ratio is (112/110) x 100 or 101.8. If it takes 8% longer than planned, its time/completion ratio is (108/110) x 100 or 98.2. Source Table 5.1. Only projects with data on completion, and on time and cost overruns are used here. ‘Absolute data’ are simple averages of project data from Table 5.1. Data ‘for all projects’ are obtained by weighting each project’s data by its share in actual cost for the projects in the sector in Table 5.1.

Sector

Table 5.3 Time and cost escalations in World Bank-aided projects in India

130

Does Aid Work in India? General features of Bank Group projects

Regionally, although of the 38 evaluations only four or five reported results clearly less than satisfactory, three of these were in one especially ‘difficult’ state. There, only two of the five projects with Bank Group aid were satisfactory. This reflects the difficulties experienced by GoI and other donors with that state, despite considerable natural resources. Perhaps this difficulty is related to donors’ special problems in tackling some thorny issues, notably that of corruption. The financial structure reflects very little donor co-financing; a repeat of a tested success (Project 9); a new donor invited to break a GoI-Bank bottleneck (15). In the only other case (24) the evaluation stated that major problems of donor disagreement and co-ordination, of the type familiar in Tanzania and Bangladesh (van Arkadie, 1986), caused serious delays. The rarity of multi-donor aid projects, then, reflects the expressed, and justified, reluctance of GoI to incur the problems of timing, conflicts of laws, and consistency that habitually bedevil such projects. Multi-donor, non-Bank projects are even rarer. Due to the delays and consequent cost overruns, GoI had to bear a larger proportion of capital costs than was estimated at appraisal. Total Bank group aid on all the projects in Table 5.1 was US$2.0bn; total scheduled project cost was US$5.0bn; total actual project cost was US$7. 1bn. (US$0.2bn. from non-Bank donors). Hence heavy extra costs, far in excess of inflation and often in foreign exchange, were imposed on GoI by the delays. Information on employment and income-distribution effects is very scanty. Also, as we shall see, the analysis on these matters is sometimes questionable. One evaluation stresses that these projects were generally designed before ‘GoI and IDA policies changed in favour of generating more employment and reaching smaller farmers’. However, these evaluations show rather little trend, over time, for project documents to become increasingly concerned with labour use or with poverty-focus. There was a trend in the 1970s in the Bank, as in other donor agencies, to shift the structure of lending towards countries, sectors, and even beneficiaries with acute poverty problems (World Bank, 1983, 1–3, 24–5). But this trend was seldom expressed in clear-cut choices, either within projects or in the policy dialogue. For example, the use of incomedistribution weights in cost-benefit analyses was outlined in a clear way by Central Projects Staff, following Squire and van der Tak (1975)—but only as an advisory guideline. It has never (so far as we know) been used in practice in any major project cycle. The recent Bank decision to monitor poverty impact more widely (see p. 148) is of limited effect because (after rising steadily from 1968 to 1981) the share of Bank Group activity in ‘poverty sectors’ (Beckmann, 1986) has since been in decline. To judge by these evaluations, so has the degree of poverty-focus in monitoring Bank Group projects even within these sectors. From 1988, however, the poverty emphasis has revived. Many lessons of the projects, outlined in the PPA/PCRs, cannot be simply read off from Table 1. We consider three sets of issues. The first set concerns project cycle procedures, donor and Indian, especially in appraisal, supervision, evaluation, and mission composition. The second set deals with analytical choices, in matters such as technology; the use of nonIndian standards, equipment, and consultants; the tradeoff between moving money swiftly and achieving long-term goals flexibly; and the level and nature of intermediation between a big aid donor and many financially miniscule ‘target’ beneficiaries. The third set of issues

Project Aid to India

131

concerns policy goals: ultimately, benefits to the poor, fuller use of labour, efficiency, and growth; more immediately, learning effects for donors via pilot projects, and for recipients via institution-building. Project cycle procedures Post-evaluations detected (inevitably) some slips in pre-appraisals. (A bilateral donor has conceded similar arithmetical errors that would, if detected sooner, have avoided a clearly unsuccessful manufacturing project). Encouragingly, such simple appraisal errors are not recorded, in evaluation documents seen by us, for projects appraised after 1970. The lack of systematic post-evaluation of project-cycle procedures by several donor agencies, however, still leaves cause for concern. Lessons can also be learned from more subtle mistakes. First, integration among aided projects needs more systematic attention, especially at pre-appraisal. The post-evaluation of one surface-irrigation project points out that the ineligibility for credit of many farmers greatly delayed the whole process of on-farm land development; yet a simultaneous farm credit project covering the same area, also Bank-financed, was confined to loans for tractors and groundwater development, and was thus apparently precluded from meeting such farmers’ credit needs. More generally, the calculation of ERORs on all Bank projects is entirely from the standpoint of the particular developing country receiving advice, yet many Bank projects to raise output of the same product in different developing countries can reduce each other’s EROR by glutting each other’s markets. Apart from international integration of project effects, it is not clear that in practice the Bank’s estimates of projectspecific ERORs deduct for the effects of draining key personnel away from other parts of the national system—e.g. towards the new Training and Visit system and away from conventional extension—nor ‘add on’ for the effects of training personnel who can then usefully serve that system outside the project area. These objections apply at least as forcibly to donors other than the Bank—and above all to the donors and projects where evaluations are not done, not available, or not numerate. Second, many appraisals, through their procedures, projections or assumptions, made future project work less fruitful. (i) In several cases, inadequacy of baseline data prevented proper subsequent calculation of ERORs; (ii) sometimes farm budgets at appraisal assumed that all farmers used the same mixes and amounts of all inputs, so that such problems as the unequal, and hence patchy and inefficient, distribution of irrigation-water offtake were not foreseen; (iii) sometimes, assurances by persons eager to proceed—perhaps by staff eager to move the money—were allowed to override careful assessment of likely implementation difficulties; this was reported in four evaluations, two each in respect of conflicts over land acquisition and of inadequate incentives to use project capital. All these three problems are reported in the more recent batch of evaluations seen by us (post-1984, of completed projects that had been pre-appraised in 1973–9), as well as in the earlier batch (1969–74). Third—although the Bank group has shown exemplary flexibility in permitting major redesigns during project implementation (these affect over 75 per cent of its agricultural projects and outlay)—the delays, conflicts, patching-up, and hence reduced ERORs, involved in such flexibility, could have been mitigated by remedying the lack of adequate consideration, at the design and appraisal stages, of alternatives.

132

Does Aid Work in India?

In general, project designs have usually narrowed down the appraisal process to a yes/no–cynics would say a yes/yes–decision, with scant scope for shifting the hardware within a project, let alone for moving to a less hardware-oriented approach altogether. There are welcome signs in some recent Bank projects in India, notably in agricultural marketing, that alternative design ERORs (and sub-project ERORs) are being estimated at appraisal, but this still seems to be the exception; such breakdowns are seldom presented when the project appraisal report reaches the Board. Still less are poverty impacts set against ERORs for alternative design/component sets, leaving the Board to decide. The Bank is probably the best of all donors here, because the least subject to the wrong sorts of commercial pressures, and the most systematic and independent in its post-evaluations. A final problem about pre-appraisals, revealed by these post-evaluation documents, applies also to other stages of the project cycle. It is that the composition of the team, its duration, and its field procedures often militate against obtaining a ‘feel’ for crucial, sometimes non-quantifiable issues. These could sometimes be handled by better team composition. A lawyer might best sniff out problems of land acquisition. An anthropologist could have helped many appraisal missions: to reject such absurd assumptions as that, in two large states reviewed for long-term rural credit projects, there was no problem of shortterm credit (and therefore no likelihood that long loans would be diverted to finance current inputs, or even consumption); to assess the impact of forestry development on tribals; and to explore the unregistered land rights, and alternative opportunities, of persons displaced by big irrigation projects. An appropriately trained economist could ensure that proper baselines and control groups are prepared, especially in pilot projects. One evaluation reveals cost increases due to rock faults—a risk that a hydro-geologist would have foreseen at appraisal. A recent evaluation comments on the lack, at pre-appraisal, of a specialist on the cropping patterns feasible under the project’s irrigation system. A recurring problem with infrastructure projects is the failure to include, in project missions, staff with appropriate macro-planning expertise to ensure that project and system are correctly matched. The cost of such errors and omissions is huge. It would justify expanding the disciplinary base of a hundred six-week appraisal missions with (on average) one extra expert each, even if only one major project improvement, worth as little as US$3m, resulted; we would anticipate a total improvement worth at least five times this sum. Once again, the readiness of most donor agencies to expand appraisal missions, imaginatively but selectively, beyond the most conventional disciplines and ranges of experience is certainly no greater than the Bank Group’s—and their readiness to reveal their appraisal procedures and expertise is usually much less. A wider range of mission staff can complement improved mission attitudes and organizational forms in improving the project cycle, but it cannot substitute for such changes. First, more time by individual agency staff in the project area, less turnover, more ‘field feel’, are essential, if aid agencies wish to learn how efficient but poor beneficiaries are to be identified and reached. One large agricultural project, lasting over five years and costing over US$100m, exemplifies the problem. Evaluators pointed out that, although missions every five to six months would have been desirable, considerably longer intervals elapsed

Project Aid to India

133

between some visits. Compounding this problem was the high turnover among Bank staff, most of whom visited the project only once before responsibility was delegated to someone else. Similar criticisms are quite common. Such brief periods in the field lead to systematic under-estimation of project-threatening patterns of tenure and credit, since these patterns are concealed. Second, even local, competent, motivated, full-time implementing agencies can fail to meet targets, if obsessed by hardware-oriented approaches and/or by top-down management structures that seldom make contact with the needy. Third, project teams cannot achieve the best results, if they persistently evade sensitive, but central, issues affecting the behaviour of both beneficiaries and bureaucrats: civil-service power struggles between new and old agencies; lack of clearly enforced guidelines about what each level of decision-making has to provide for, as well as to exact from lower levels (as when the unreliability of the main system of gravity-flow irrigation discourages water users from taking time on co-operation below the outlet); and sometimes the existence of serious corruption. These three sets of issues are responsible for apparently technical problems in many aided projects.9 Faced with such problems, donors, eager to avoid accusations of meddling in domestic politics, either seek solely techno-economic solutions at project level (changes in physical design of systems, or in pricing of their outputs to users), or try to impose policy conditions at sectoral, state, or national level. However, if the power structure stays much the same, such tactics are likely to be partly frustrated, as abuses or distortions helpful to groups with power will often emerge in other forms, expressing more or less the same power balance. Donors of project aid tread a fine line: how are they to strengthen the forces of efficiency, rationality, and responsiveness in project cycle procedures, without improperly intervening in local or state politics? At least India, unlike many African countries, has enough local expertise and open debate to ensure that extreme political or administrative distortions of techno-economic guidelines are publicly questioned. However, the (self-?) delusion that the issues really are just techno-economic is unlikely to help donors. Although the choice of technique on a project can indeed change the amount of damage done by unclear lines of authority, or by over-centralization, or by corruption, this is possible only if project designers explicitly allow for such non-technical issues.10 In India, the above problems are eased by a competent, long-serving group of Bank experts in Delhi, who move around major projects often, and who usually have relations of mutual respect with similarly competent and long-serving Indian experts. However, much more responsiveness to, and day-to-day local contact with, the intended beneficiaries of big aid projects—often, even if very poor, interacting in complex ways with such power groups—is required than this technocratic model suggests. In this context, two recent evaluation documents may be in error in stressing the advantages of a resident Bank staff in working with Centre officials to promote Bank Group projects in India across departmental boundaries and with less effect from personal and political pressures. Such advantages are likely to be achieved at the cost of slower progress on non-aid GoI investment activities (and on development activities on current account) which lack these special favours. And Bank Group projects, if thus insulated and protected from ‘normal’ administration and politics, are themselves vulnerable after the evaluation is over and the donor has left. The way to better project-cycle procedures lies through much greater presence in the field.

134

Does Aid Work in India?

The problems in project-cycle procedure should not obscure the general success of Bank Group aid in India. There has been much learning by doing. However, many criticisms recur in recent evaluations; Bank project procedures in other developing countries, notably those with less Bank presence and shorter learning, are usually less satisfactory than in India; and other donors’ project procedures are usually inferior to the Bank’s—not least at evaluation. However, the problems raised on pp. 147–8 appear in Bank evaluations too— especially the problems that project performance audits (PPAs) represent swift overviews, often too soon after the event; that project completion reports (PCRs) partly reflect the views of staff involved in the project cycle; and that baselines and control groups are absent or inadequate. All this is especially damaging to pilot projects. Analytical choices in projects Technology The first such choice concerns technology. We have stressed that technical choices appear to be foreclosed too early in project design, generally leaving the appraisal to assess just one way of doing the project That ‘one way’—given the backgrounds of donor project preappraisers, the pressures to obtain project approvals that seem to move money quickly, and the wish to reduce apparent risk—is too often the way of high technology. In a fertilizer project, an incompletely developed process, based on an untried combination of subprocesses, caused losses from severe delays and foregone production. Another project suffered similar effects from complex slip-form silo construction. Such approaches greatly increase the need for special technical and managerial skills. In some earlier aid projects, this directly delayed critical works. India’s large and growing pool of skills has reduced such problems, but several evaluations of recent projects still show high-tech choices causing indirect delays. The cause of indirect delay is often a prolonged dispute between the GoI and the potential donor about whether the high technology of the project necessarily requires foreign consultants and foreign procurement of supplies. Donors’ and consultants’ penchant for high technology not only causes delays and losses. It also, at first, diverts attention from beneficiaries’ preferences and, later, helps to channel such preferences more towards superficially ‘modern’ approaches, rather than towards cost-effective ones. This can imperil the very objectives of aid. The distortions of high-tech affect not only hardware, but even rural credit projects. In one case, farmers’ own tested and cost-effective preferences (for bamboo over conventional steel tubewells) were not analyzed at appraisal, and were only to a small extent financed by the project. In another project for credit to minor irrigation, farmers’ wish to avert risk by using low-input, low-output dug-wells (rather than more elaborate tubewells) was not foreseen at appraisal. Sometimes, credit projects have so defined beneficiaries as to exclude the really poor from credit or benefit, at the same time as claiming that all remaining beneficiaries, even ‘poorer’ ones, are sufficiently advanced and capital-intensive to prefer the most advanced technique. Thus the post-evaluation of a project offering credit for wells reported that so-called small farmers, as readily as large farmers, adopted mechanical pumping rather than the improvement of traditional methods; yet farmers operating less than three acres—70 per cent of all farmers in the project area, and those likeliest to prefer the more labour-intensive, lower-risk method—had not been permitted to borrow under the project,

Project Aid to India

135

Generally the emphasis in credit support upon capital-intensive modern technologies not only ‘unemploys’ workers, but also discourages adoption among the smallest, labourintensive, unskilled, and risk-averse entrepreneurs—especially if, apart from scale and indivisibility, the equipment is experimental, or new to its environment or culture. Both employment and income distribution tend to suffer from high-tech, unless very carefully selected to suit the requirements of local poverty-groups or symbiotic with non-aid, labourintensive investments. Sometimes it is—even though many in the government believe that western donors’ comparative advantage in India lies precisely in high-tech. First, the benefit/cost (EROR) pre-appraisal of all Bank Group projects allows for the fact that under-employed labour, which is likelier to be used up by a less high-tech approach, is less costly to India than its market wage suggests. However, even this does not allow for the benefits of labour-intensive methods in respect of poverty alleviation. (It would do so, if the Bank would only make mandatory its guidelines on incorporating income-distribution weights into ERORs.) Nor, perhaps, has the Bank Group done enough to mitigate undue predetermination (largely by engineers) of project technology in the design stage—which almost certainly discourages low-tech. Second, large programmes—such as five years’ worth of total Indian investment, or even of IBRD/IDA loans to India—should take some risks, aiming at a high average EROR on projects rather than a safe but lowish EROR on each one. Hence, such programmes should aim to install some technologies that look towards India’s more skill-intensive and capital-intensive future. Finally, some Bank projects have shown institutional flexibility and ingenuity in ‘learningby-doing’ how to provide access for very small and poor operators to efficient large-scale and/or ‘high’ technologies. For example, the importance (not least for employment) of timely supply of improved seeds—delivered via a complex and sophisticated ‘hightech’ system—initially dictated a large-farm approach to producting the seeds for use by smallholders. Later, after piloting, it was spread successfully to seed-growers with small farms, through a ‘compact area’ approach. Nevertheless, in most cases high-tech, large scale, capital intensity, and few jobs (per US$1m of project investment) do go hand in hand. They do, also, match the preferences, manuals, and working experience of most engineers trained in western-style institutions— not least those in developing countries. The Bank Group’s ingenuity in sometimes seeking to offset the worst effects of these biases has not wholly removed the damage via delay, lost output, high costs, and foregone jobs.11 Incorporation of project economists throughout the design stage, presentation of alternative designs at appraisal, and use of income-distribution weights in both design and appraisal, appear to provide possible remedies. Foreign expertise A pervasive cause of delay, and hence of lowered ERORs, has been Bank Group insistence that, for some purposes, recipients use foreign expertise or equipment. Such insistence usually raises project costs and delays completion, but can in principle raise ERORs if it improves project performance by enough in compensation. Often, however, this seems not to have been the case, though the Bank’s evaluations point to a learning process.

136

Does Aid Work in India?

This problem—while much less frequently reported in evaluations of Bank projects completed since 1983 or so—is probably persistent among other, less frank donors. It ties in with the wish of bilateral donors to respond to their own exporting firms. For most donors, even for the Bank Group, the problem is typically compounded by donor pressure for ‘modern’ technologies. For example, in a big fertilizer project, the Bank Group appears to have objected to an economically adequate, if perhaps not optimal, process proposed by the Indian executing agency, and to have withdrawn its objection only after long and costly delays. In another example, the Bank’s initial insistence on an (eventually uneconomic) high-tech approach (slip-form silos) was also the ultimate cause of ‘an inordinate amount’ of delay to a storage project, due proximately to GoI-Bank disagreements about whether local consultants could be used. We reject a simplistic framework of ‘good’ local expertise versus ‘bad’ high-tech consultants and hard donor agencies. Sometimes, a donor can take too soft a view on local expertise; a Bank-aided fertilizer project was seriously penalized by the GoI’s decision to force the Indian executing agency to purchase some equipment from a manufacturer in India without adequate experience, against the recommendations of IDA. Sometimes, it is the recipient government that impedes local input; in an irrigation project, both the Bank and the Indian executing agency were frustrated in their wish to rely upon local inputs by GoI delays in counterpart funding, though GoI comments on the evaluation suggest that this cannot happen nowadays. India now has a wider range of competing inputs—and of consultancy skills, often under-employed—and the Bank Group has by now learned to use them better. In a project completed in the early 1980s, for example, foreign and local consultants worked well together, with an exchange of roles: in the later years of implementation, the local firm took over prime consultant responsibility, and the foreign firm supplied the project with specialized assistance on contract. Yet, on balance, donors have not gone far enough in using Indian expertise and equipment.12 Move money or improve project? A third analytical choice in aid project management is between getting the money moving quickly, and achieving long-term goals flexibly even at the cost of some delays. Donors, especially the Bank, are normally accused of overstressing the former. Usually the career incentives in any donor agency are to move money swiftly and at low administrative cost: banks want to lend, and aid agencies want to use the money voted to them by governments— otherwise these funds may be lost to the agency at the end of a fiscal period, and the vote next time may be reduced. If the project turns out badly, the quick-spending officers partly responsible are unlikely to be ‘punished’ in career terms. Often, indeed, they will not be in the same part of the same agency when their chickens come home to roost—though in India, aid agency staff are more likely to stay in the same group of offices. Bank Group project delays do not suggest a mad rush to move money fast. However, using speed as a criterion for efficiency can be misleading. Undue speed at design or appraisal might, in principle, slow down implementation later on. In several projects, a wish to move money early has indeed meant long delays later. However, it remains sensible to have some high-risk projects in a big portfolio, to keep average ERORs

Project Aid to India

137

high. Delay in beginning implementation of a good project is a real cost; reduction of such delays justifies some hurry at the appraisal stage. The questions are: how much? And should more attention be given to identification and design? The balance between two targets that both raise ERORs, but are in direct conflict–moving money fast and getting projects improved–is a fine one. Despite some exceptions, we do not believe that the Bank has pushed too fast in its Indian loans. One clear example of care is shown in a farm credit project in the 1970s. The evaluators point out that the local executive agency by which the project was ‘carried out [was] known to have major weaknesses. [The] comprehensive package of manpower and financial reforms insisted on by the appraisal team, even at the cost of a significant delay in project processing, [is] to their credit’. Intermediation; overdues, externalities, federalism It is the GoI, not the project authority, that guarantees repayment of aid loans—essential if donors are to be able to pick projects with high ERORs even if the recoverable commercial returns are low or negative; and desirable to avoid excessive caution in project selection, which would produce a safe but low EROR on the donor’s portfolio. Formally, therefore, aid loan agreements are between GoI, or its financially guaranteed agents, and a donor, such as IDA. There are three crucial issues, however. First—despite the formal aid agreement—in substance the money passes through a series of intermediaries, often ending in the private sector (Chapter 7). For lending in support of rural credit or minor irrigation, the series might be: IDA; GoI; Reserve Bank of India (RBI); NABARD (formerly ARDC); a state-level land development bank (LDB), or else a commercial bank; a primary land development bank (PLDB), village co-op, or a rural branch bank; and finally private borrowers, mostly farmers. (Analogous series apply where money passes via an industrial development finance corporation such as ICICI). Second, each ‘link’ in the chain has access to other sources of cash. For example, given the level of aid, the GoI can enable the Reserve Bank of India to increase its on-lending to NABARD, so the latter can re-lend to commercial banks in support of their rural lending; farmers can borrow from commercial banks or from village moneylenders, both financed privately. Neither the Bank Group nor other donors can influence this directly. Third, though the loans are normally destined for medium-term to long-term producer credit, extra cash is in reality to a large extent fungible between this use (for which aid is nominally intended), short-term credit, and consumption; and it is fungible at each stage of intermediation. Many of the problems besetting aid loans—notably the problem of overdues—are related to these three interwoven issues: serial intermediation, parallel sources of cash, and fungibility. First, if intermediation is serial, it is hard to apply incentives and disincentives at the right level. Typically a commercial bank, faced with excessive overdues from a rural branch bank serving several hundred borrowers, can threaten to close the branch, or at least to reduce or delay new lending to it (and therefore by it). However, this penalizes not just wilful defaulters, but also (i) the unlucky; and, above all (ii) borrowers who have succeeded in keeping up repayments. Such good payers, indeed, are encouraged to default after all, by the threat that even if they repay they will be unable to borrow from their

138

Does Aid Work in India?

bank branch next year, because it has effectively been bankrupted, by the defaults of other borrowers. Moving one stage up, the ability of a particular rural bank to on-lend is reduced, if NABARD decides to cut its support because it has done badly in retrieving overdues from village branches (at least three substantial cases are reported in the World Bank’s evaluations of rural credit projects). But this incentive to good credit management for the board of directors (of a commercial bank or of a land development bank) is a direct disincentive to a local branch (a village bank or a primary land development bank). If the managers of a branch bank feel that head office—because it has been penalized by NABARD for the past poor performance of its branches as a whole—will be unable to lend it much money for the next crop anyway, why should the branch struggle to repay this season? At the level of the aid donors, by threatening NABARD with the disincentive of reduced funding, they indeed encourage it to penalize each commercial bank head office more heavily for its overdues; but such penalties not only create incentives to default among ‘innocent’ local branches of each bank, and among non-defaulting farmers, but also allow bad financial returns on a project, due partly to intermediation problems, to choke off funds when economic returns may be excellent. This problem is not confined to farm credit (as one evaluation of a large infrastructure project makes clear). Nor has it a neat answer: simply to continue unlimited funding not only leaves the aid donor in the position of appearing to encourage financial indiscipline, but also puts the GoI in the same position, and drains its resources, because the aid commitment to the project is fixed in cash terms, as is GoI’s commitment to finance the gap between donors’ contributions and total project costs, unless all parties agree to reduce the size of the project Apart from this serial intermediation problem, appropriate incentives are hard for a donor to apply when recipients at each level can respond by turning to parallel, competing sources of funds. The Bank has taken the view that institution-building, in farm credit, involves it in gradually raising the ‘level’ of its lending, from state-level institutions to NABARD (ex-ARDC). However, each time the level of donor support is raised, its influence on base-level actions by borrowers is further diluted, because there is one more intermediary lender who can turn to parallel sources of funds. Is all this ‘healthy competition’ in credit markets? Only if price incentives, information, externalities, and distribution of income and power are such as to translate maximization by each agent (farmer, village formal financial institutions, commercial banks) into socially desirable equity and efficiency outcomes. Several evaluations suggest otherwise. In one state credit project, for example, the structure of borrowing for minor irrigation led to a much less labour-intensive pattern of works than IDA had envisaged. Since each borrower had many alternatives, IDA could exert very little influence (even if it has wanted to do so). In many cases, borrowers’ access to these lending alternatives prevented IDA, NABARD/ ARDC and commercial-bank/LDB from controlling negative externalities. Most notably, Bank-ARDC-LDB rules on spacing of wells designed to conserve groundwater (and thus to ensure both the farmer’s return to his well and the lender’s repayment), are not enforcible if alternative credit sources are readily available. Evaluations of IDA aid to rural credit in two states, and a general OED review of such aid, specify this problem. Apart from credit projects, much Bank support for specific rural investments contains credit components, and faces the same difficulties. Thus a 1986 review by the Bank Group

Project Aid to India

139

of groundwater development in an Indian State pointed out that ‘unregulated private development may have a serious limiting effect on usable groundwater’; that ‘poor recovery [of credit for] private investments’ in such [over]use is typical; and that ‘a substantial part of investments through World Bank supported institutional credit are used for private groundwater investments’. Non-enforcibility of economic incentives, due to parallel and serial intermediation, and to fungibility, is a challenge to aid-effectiveness extending far beyond the Bank Group, credit, the rural sector, or India. It is the basic challenge posed by economic externalities— benefits reaped by one party, while costs are borne by another—to market-based approaches to development (of which use of aid is a small part). The choice of level of intermediation, and of appropriate incentives, is complicated in India (and some other countries) by federalism. The World Bank’s early agricultural credit projects sought to help GoI to strengthen state-level local development banks (LDBs) and commercial banking institutions, alongside ARDC’s capacity to supervise and evaluate them. Only later did the Bank move towards direct support of ARDC/NABARD, citing the advantages of greater flexibility, and of capacity for reachout to remoter states and smaller projects. The long, careful build up meant that gains probably exceeded losses from this upward shift in the level of intermediation. Where a donor straight away operates at the centre, at the highest level of intermediation—and especially where its share in the project is small—relations of donor and recipient tend to be less successful. In the case of an infrastructural project in several states, for example, the evaluators found that, ‘IDA’s lack of a (direct) contractual relationship with the state (authorities helped to ensure) that IDA’s influence at state level was minimal or non-existent’. Goals of Bank Group project aid Institutional learning: recipients Most PPAs—and indeed the earlier appraisal documents—express great concern with helping Indian institutions to improve their performance. The increasing use, in the Bank’s own analytical work, of Indian consultants (e.g. of ARDC/NABARD itself, to prepare several PCRs following credit projects) exemplifies this concern. In most fields where there have been several Bank Group projects—credit, fertilizer production, extension— major, and in many respects successful, institution-building efforts have been made. There was plenty to build on; in many areas, e.g. fertilizer technology, the Bank Group may well have learned more from the institutions than it has taught them. A major success of the Bank has been as ‘marriage broker’—or catalyst—showing how different institutions can work together on projects. Bank Group credit support in the mid1970s was critical, for example, in the development of operational links between ARDC and commercial banks, and hence helped to make feasible the GoI’s planned shift from a rural credit system based on co-ops and LDBs, to one based largely on commercial banks. The Bank’s population project has also twinned each of two Population Centres with a different, prominent Indian management institute. A third example has been the use of an IDA loan to ‘catalyze’ the combination of research expertise in an agricultural university with the commercial tasks of seed-farming and processing.

140

Does Aid Work in India?

In any country, foreign visitors are usually better than insiders at uncovering—and financing—unsuspected prospects for co-ordination. Hence, institutional marriage is a donor strength. Lack of attention to locally reared institutional ‘children’, however, is a donor weakness. Water users’ co-operation is discussed in several evaluations, but with reference to exogenous groups or formal co-operatives. The evaluators seldom say much about informal, local, traditional institutions for such co-operation, let alone of the circumstances that sustain such institutions: e.g. some scarcity, but not too much, of the co-operatively allocated resource (Wade, 1979). Nor, at project level, is much attention paid to informal credit sources. Not only moneylenders (where competitive), but also rotating credit and savings associations could well sometimes function as intermediaries with local knowledge (and low costs of securing loan recovery), and could thus perhaps alleviate some of the credit constraints on rural projects, reach more of the rural poor, and avoid some of the problems experienced with conventional, formal intermediaries. Another area of institution-building, normally initiated by Indian official agencies but at least supported by donor projects, needs to be questioned. ‘Island projects’ with exceptionally high levels of services, run by special project-based authorities and with independent resources (often dependent on donors), are against GoI policy as aid media. But this policy appears not to rule out island programmes. One such programme, for a group of areas with similar ecology, is stated by OED to have had ‘good institutional support’ from conventional GoI line agencies. Yet it suffered ‘frequent changes of administrative staff at higher levels. Also, ‘lack of technical expertise’ in the central unit was a problem, because such professionals were ‘almost entirely centred in the [line] Ministry’—with which ‘cooperation did not materialize to the extent desirable’, impeding the programme’s ‘development of technical packages’. Moreover, ‘line department officers who were seconded to district’ authorities of the programme had to implement project ‘activities [not] supported currently by their departments’. Conflicts of authority and interest led to serious consequences, with little effective remedy. These are standard problems faced by island programmes and projects isolated from national systems of line administration, secondment, and finance. Such programmes, though not projects, have a long history in India—especially in rural India (Lipton, 1987). They represent recurring, often courageous, usually well-intentioned efforts to break bottlenecks in, or among, traditionally proud, separate sub-bureaucracies—different states and the centre; different line departments; IAS and IFS—but face major problems. Perhaps donors should be more careful when supporting such programmes, especially since their post-donor life is usually poor and short. When aiding such organizations as Command Area Development Programmes, donors may show insufficient awareness of the possible conflicts between donor-dependent programmes and stronger traditional line ministries. At least one evaluation reveals the seriousness of such problems. This is not to say that island programmes and projects never work, or that new, top-down organizations, supported by ‘Christmas tree’ donor projects, are always wrong. A major evaluation of a Bank Group input into an urban development programme—an exciting, analytical, well-written document that deserves a wide audience (not the ‘restricted distribution’ to which all OED project-specific work is sentenced)—documents a big and dramatic exception to many current platitudes about development projects. This programme

Project Aid to India

141

mixed most of the metaphors for the undesirable: island, Christmas tree, top-down—yet it has clearly worked in itself, though probably less so as a pilot (see pp. 176–7). Institutional learning: donors Much of the history of aid projects in India comprises pilot projects: examples from the Bank Group include the projects above and at least one project each centred upon urban sanitation and upon irrigation. Other donors have also supported pilots, e.g. the UK-aided project on dryland farming systems. In pilot projects, the donor may often be tempted to try out numerous project components, even if they have ‘no essential linkages or interrelations’. With such projects the only connecting item, on which sub-projects are hung, is the stem of the Christmas tree: donor cash and influence. These often sustain a local organization that has little support in—indeed is seen to rival—line ministries, and is therefore unable to survive, in any strong or integrative form, the transition from donor to domestic funding. The transition from support of statewise agricultural credit to direct support of ARDC/ NABARD was a relatively successful instance of piloting. Farm credit can in principle support many things; the Bank Group’s nine state credits, of which a Bank Group impact study was made in 1981, were limited mainly to two sorts of purchase by final borrowers; minor on-farm groundwater equipment, and tractors. The evidence shows that the former was narrow but productively cost-effective (although a Bank evaluation in late 1986 shows that such privately- owned tubewells can be inequitable and may overuse both groundwater and electric power), but tractor credit was often inappropriate. In any event, the relatively narrow range of approved objects for lending (even allowing for fungibility), and the concentration on permanent (indeed almost, though not in the standard ministerial sense, ‘line’) agencies, did permit successful donor evaluation and learning. The nine-project impact study shows that Bank credit projects tackled several, in some ways increasingly difficult, states—and, under GoI guidance, assessed and improved the capacity of the apex organization (ARDC/NABARD)—before the long and costly pilot stage was transformed into direct support for ARDC. A massive piloting exercise in India—a project seeking to upgrade a whole range of services and housing in a big city—was a success as a project but probably a failure as a pilot. Two reasons for these linked facts—the uniqueness of the city served, and the ingeniously opportunist ‘situationism’ of both Indian project management staff and Bank authorities—are well discussed in the PPA. The executing agency was initially an engineering-based, strong, non-participatory body—in essence a reaction or successor to the functional near-breakdown in that city of the traditional modes of local government It was seized on for flexible Bank Group support, according to the evaluation, in ‘patch-up and fix-up style’, as ‘a bold experiment formulated quickly in response to manifest and immediate need’. Thus the executing agency was allowed to get on with the jobs in hand without the usual Bank insistence on appraising each sub-project. This led to charges of ‘Christmas Tree’ status, charges justified in form but not in substance. There was no sign of the increasingly fashionable participatory, bottom-up consultations. Institution-building took priority over project specification. The outcome was almost certainly excellent by measurable criteria, and the pilot led to two further large projects in the same city.

142

Does Aid Work in India?

This apart, however, the initial project piloted very little. The authors of the PPA recognize that the centralized, engineering, non-participatory management style usually ‘works’ only during reaction to extreme conditions associated with perceived failure of older institutions. The PPA stresses that institutions of city government in India must normally be much more democratic—as in an ongoing Bank-aided urban development project in another Indian city, with ‘a flat-structured [Indian implementing] organization that relies on negotiation rather than a vertical-structured [one, that relies on] authority’. Such democratization is seen as the pilot city’s future also; an authoritarian island in a democratic sea is hardly a useful pilot. Even within India, the recognized uniqueness (and extremity) of the ‘pilot city’ rendered it a somewhat doubtful pilot. But the major problem with this pilot is the absence of timely provision for control groups, baselines, and evaluations. Crude ERORs were attempted at appraisal, but were estimated only for some sub-projects. They are going to be very hard ever to verify, because only well after the pilot (i.e. during the successor project) did the executing agency ‘initiate an evaluation of the impact of past investments’—inevitably, ‘based on available, and in some cases limited, data’. There are enough indicators (e.g. death-rates), measurable over time in comparison with other areas, to be fairly confident that overall success was achieved. But it is almost impossible to determine what parts of this large project worked best, and why. This is not said to attack a project that, rightly, gave first priority to a cost-effective attack on extreme need in one Indian city. However, with a pilot project, proper baselines and controls, to permit full evaluation of performance, should have been estimated. Probably, an independent local research institution should have been selected and helped to work with the executing agency in monitoring; that might have added 0.5 to 1 per cent to the cost of this vast project, but would have added much more to its usefulness as a pilot. Lacking such evaluation, one must conclude that this pilot is exceptional; and that, as a rule, many purposes, islands of organization, non-participation, and Christmas-tree styles do not conduce to project success, nor lack of baselines and controls to piloting and donor learning. Ultimate goals: fuller labour use It is chiefly by increasing the prospects for and returns to work—and thus the amount of work that people are willing to do and to hire, and also the amount they can produce and consume per hour worked—that developing countries and aid donors can best advance, simultaneously, the causes of efficiency and equity. It is, however, not surprising that aid projects have often been bad at increasing the incentives and prospects for work. Donors’ own internal incentives and technologies (and consultants) often militate against labourintensive design choices. Another serious impediment to clear thinking about the employment impact of projects has been the rapidly changing theory and evidence about the work-productivity-poverty nexus in the developing world. In the 1960s many believed that vast idle labour reserves existed, waiting to be tapped for development In the 1970s the pendulum swung, and for many economists involuntary unemployment was almost defined out of existence. Only fairly recently has clear evidence come in that involuntary unemployment—as opposed

Project Aid to India

143

to withdrawal from labour compelled by illness, hunger, domestic ‘duties’, or the costs of job search—is significant, fluctuating, increasing, but heavily concentrated on the poor, especially casual employees and women (Lipton, 1983). It is not surprising that aid project work often: (a) has incorrectly analyzed the impact on work duration and hourly work income—the example of tractorization is discussed on pp. 179–81; (b) has incorrectly compared pre-project and post-project circumstances of labouring groups, using ‘partial equilibrium’ methods where these are likely to involve major errors (see pp. 181–2). More surprising, however, is the insensitivity often shown to issues of employment. In particular, most donors, including until very recently the Bank Group, have tended to ‘hive off analytically the big, capital-intensive projects—major transport links, factories, etc.—and to confine analysis of employment (and income distribution) to more diffuse projects such as rural credit or slum upgrading. On the former group of projects, several evaluations strikingly omit any discussion of ways to expand, consistent with economy, the use of unskilled labour. In other cases, the Bank Group evaluation even takes credit for reducing the numbers of unskilled workers. There is a real tension here, between project efficiency (and financial viability) and social efficiency, although the latter is never advanced by adopting a technique that uses more unskilled labour, at a given level of use of all other inputs, than is needed to produce a given output However, social efficiency can be advanced by adopting labour/capital ratios considerably higher than would be implied by maximizing financial or even economic returns without income-distribution weights. On top of income-distribution weights—i.e. even for projects or techniques justified with such weights—appraisals should allow for the indirect costs when people become unemployed in societies with little ‘social security’ other than crime or beggary. Far from this, it does not look as if, once a project is in operation, labour is even priced at social cost by Bank (or other donor) advisers on staffing levels. Nor, when advising against ‘overstaffing’ in some operations, are consultants automatically asked by donors to look for socially cost-effective, and financially tolerable, ways to expand unskilled employment elsewhere in the project. Under-used capacity and excessively high-tech are common and costly faults in large Indian projects, aided or not. This suggests that incentives to labourintensity in construction and use—perhaps associated with attempts to negotiate ‘highemployment, wage-restraint’ deals with major labour unions—could well be explored early in many project cycles. Should donors, linked with that part of GoI which sees their comparative advantage in high-tech, press India to be even more dualistic, half high-tech (with aided projects that aim to slash ‘overstaffing’) and half poverty-focused (with aid projects aiming at labour-intensity)? This would seem unwise; Japanese development since 1945 (perhaps since 1880) suggests that such patterns are feasible, but require a degree of corporate political control over business far beyond what is acceptable in India. The World Bank’s move in 1986, to extend the monitoring of poverty impact to major project cycles outside the traditionally poverty-focused sectors will if implemented be an important move away from such dualism. So could the incorporation of income-distribution weights into project design choices. Even in rural and small-scale projects, donors’ analytical styles may miss out major employment-linked problems. The impact study of Bank Group credit projects—much

144

Does Aid Work in India?

of which financed tractor purchase by non-poor farmers—points out: ‘In southern states, tractor owners interviewed saw as main benefit the ability to manage large (16–20 hectare) holdings without depending on burdensome permanent labour. [There], tractorization has at best no or even a negative effect on employment’. While the problem is recognized, the tone is rather too hopeful, and the following statement that elsewhere in India tractorization ‘created more employment than [it] displaced’is almost certainly incorrect (see Binswanger, 1978; Farrington and Abeyratne. 1982). Tractor aid is important because it is popular with donor exporters and because it typifies the problems of evaluating aid that finances potentially labour-displacing equipment. One major IDA agricultural credit project implemented in the 1970s was intended to finance mechanization only—8,000 tractors (plus implements), 40 self-propelled harvesters, and 200 tractor-drawn harvesters and spares. The analysis of mechanization is confined to the tractors, however; the much more questionable and employment-reducing (and itself much reduced) harvester component is not evaluated. The evaluators reported that as a result of investment in tractors, cropping intensity rose from 140 per cent to 160 per cent and the cropping pattern changed to high value crops. Also cropped area rose by 20,200 ha. and the gross value of production increased by US$30 million per year in the two states combined. However, there are four basic errors in the evaluation. First, the only study specific to this agricultural credit project covered just twenty-five tractor-owners and twenty-five non-owners, spread across six villages in one district of one of the two affected states. Second, it is not correct to assess the output gains from tractors by comparing output levels as between owners and non-owners of tractors instead of between users and non-users. This is to assume that all and only owners are users. In reality, plying for hire is a major use of tractors. Third, the method of analysis assumes that differences in farm output and returns, between tractor-owners and others, are entirely attributable to tractors. In reality, tractorowners are much likelier than non-owners to have access to irrigation and other inputs, as affluence is a cause of both tractor ownership and other advantages. Finally, the survey contained no pre-project baseline data, but only data after project completion. So the evaluators cannot safely conclude that the cropping intensity increased from 140 per cent to 160 per cent. All we can say is that tractor owners had a higher intensity, after the project, than non-owners; they may well have had an equally elevated intensity before the tractors arrived. Identical remarks apply to claimed ‘rises’ in yields and area. These are not rises but differences between rich and poor farmers at a moment of time. Such differences prove nothing about changes in the relative status of users and non-users of tractors over a period. Yet it is on this series of mistakes—echoed in most donors’ and recipients’ evaluations of such equipment—that the high claims for IDA tractor credits rest.13 The failure of tractors to raise output, and the generally negative impact on employment, are fully borne out by other studies of the area, not associated with the PPA itself but (very properly) reviewed there—and even more strongly by more recent work (Agarwal, 1984). The general lack of independent contribution by tractors to output—except where they enable new land to be cultivated—is, indeed, crucial to employment effects. For a major component claimed for the ‘economic benefits’, in this PPA, is ‘labour savings through lowering labour costs per unit of (crop) output’. If the claims for higher total crop output

Project Aid to India

145

due to tractors as such are unfounded, as seems to be the case, then this ‘benefit’ operates wholly through raising unemployment. In such a case, true project benefit depends wholly on the ability of the economy to absorb the released labour productively. Even in those rare circumstances where tractors, although reducing labour use per unit of crop output by a proportion, raise crop output by a higher proportion—almost invariably an area effect, not a yield or an intensity effect—we would ask whether similar expenditures on other methods of relieving draught-power constraints, or indeed on other inputs (such as water, seeds, or fertilizer) raising yields despite such constraints, might have raised output by as much, or more, with a greater share in total benefits (and in input use) for the poor (and under-employed) labourers. In this joint evaluation of two nearby tractor-credit projects, the PPA is generally much more cautious than the PCR about employment effects. Nevertheless, it also accepts the conventional wisdom that ‘labour has become increasingly scarce’. (Sceptics like ourselves read this as a claim that employers do not wish to pay market wages, nor governments to use resources to support seasonal migration from depressed areas.) ‘Labour effects were not incorporated in [the] survey but it was assumed that there was incremental labour demand. During the mission’s field visits displaced tenants or farmers were said to have found alternative employment. All persons the mission asked were of the opinion that on balance tractors had created employment’ (our italics). It is unlikely that the Mission met migrant workers who had returned to very poor nearby areas. Such workers are the likeliest to be displaced by tractors (and by the, unevaluated, threshing machines also provided). A related problem to donors’over-estimation of employment effects is that evaluations— and other project documents—often just compare ‘beneficiary’ and ‘control’ farmers after the project. But one also has to know the two groups’ pre-project (baseline) levels of family and hired employment. Otherwise, one cannot exclude the possibility that the groups’ different post-project levels of labour use are due to differences in endowments between the groups not controlled for. Suppose the latter is not the case. It is still not possible to infer that—because 10.6m. extra family workdays and 14.5m. extra hired workdays are created in one Bank project on the farms receiving credit for extra irrigation—this is the employment effect; it may be more, or less. The family farm-workers who benefit from the project may respond to extra on-farm work chances by supplying less labour for hire off the farm. This will reduce their hired workdays; part of that gap will be filled by extra labour from hired workers without farms, or with farms not benefiting from the project. Numerous other effects, some very large, are feasible, depending on income and price elasticities of labour supply by family and hired workers; and on cross-elasticities of labour supply for one use to changes in attainable wage-rates in other uses, both among and within households; and also on demand elasticities. The only way to compare employment effects is to compare pre-project and postproject income, and levels of hired and family labour, as between two roughly comparable communities (e.g. groups of villages), with separate but initially similar labour markets. Baseline and control sample areas, not just households, are therefore needed. Only thus can the total effect of a project on its area—not just the effect on project households—in respect of, say, family work and hired work, be compared. This is not a nit-picking point. Unfortunately it means that evaluation estimates of the employment effect of projects—

146

Does Aid Work in India?

whether or not aid-financed—may overstate or understate by a factor of two to one. The absence of information in most evaluations about whether extra work required is in the slack season or in the busy season—i.e. about whether such work is likely to add to existing employment, or to substitute for it—makes the problem even more serious. Not only in Asia, but increasingly in Africa too, the poverty impact of rural projects depends less on the effect on ‘small farmers’ than on the contribution to work chances for landless and near-landless labourers. Proper assessment of that contribution is thus a top priority for ‘evaluation reform’. Ultimate goals: less poverty Most of the projects, recently evaluated, were designed and appraised a decade or more ago. Then, many experts believed that growth would usually ‘trickle down’ to the poor, so that projects that efficiently produced growth were the best way to alleviate poverty. Since the mid-1970s, much counter-evidence has appeared. Chapter 2 rehearses the evidence that, at least until the late 1970s, steady growth in India brought no significant decline in poverty incidence or severity—especially when skill-intensive and capital-intensive (as often with foreign-funded projects), growth can fail to benefit poor people for long periods, and can even harm them. An environment of rapid increase in the workforce, and of price policies that discourage the use of labour-intensive commodities or inputs—while not the fault of aid donors—can interact harmfully with aid projects that ‘go for growth’ on the assumption that this alone will ease the task of poverty alleviation. Rather, these two goals are usually complementary; for example less-poor workers are healthier and better-educated, and thus more growth-inducing. But are ‘projects’ appropriate to employ them? Issues of aid and poverty are treated in Chapter 2. Here, we look only at three projectspecific issues, exemplified in Bank Group project evaluations. How do projects define potential participants, and with what effect on poverty groups? Do regional patterns of aid affect its poverty-focus? Does the choice between projects with mainly technology-based, institution-based, or market-based approaches affect the impact on poverty? First, ‘small farmer’, ‘poor beneficiary’, etc. are often defined in ways that exclude many of the poor, and all of the poorest. In one state credit project in the mid-1970—for example, the evaluation points out that ‘marginal farmers’—apparently defined as those operating below 3 acres, not 2.5 acres as in standard GoI practice—‘were not permitted to borrow under the project’. Yet, even in 1970–1, of all operational holdings in the state that were wholly or partly in agricultural use, over two-thirds were smaller than 2.5 acres (Naidu, 1975). Thus, by project completion in the late 1970s, the smallest 75 per cent at least of operational holdings must have been excluded. In another state, a 1970s Bank Group project for credit to on-farm minor irrigation specified a ‘minimum benefiting area of 1.5 ha’ for dugwells; but the 1971 land distribution in that state would have restricted borrowing to the biggest 60 per cent of farmers (Naidu, 1975), and by the completion in 1977 surely to the bigger half. Yet we are told that 55 per cent of wells were completed by ‘small farmers’. In a third Bank-aided credit project, completed in the 1980s, again according to Naidu, the GoI’s small/marginal farmer borderline (2.5 acres) placed 60 per cent of all farmers

Project Aid to India

147

below it; yet, despite the use of this lower cut-off for lending the evaluators still claimed that ‘80 per cent of the lending went to small farmers’. It is not clear that the smallest holdings cannot be made viable even with World Bank credit. The ratio of default to lending steadily increases with farm size (Lipton, 1976; Copestake, 1987). Nor is a tiny farm uneconomic or unviable merely because it suffices to meet only part of a family’s needs. Donors and recipients often reduce the impact on poor farmers by defining the ‘small farm’ category to include the not-so-small. ARDC successfully modified the standard definitions of GoI (on which ‘small’ but non-marginal farmers operate 2.5 to 5 acres, modified to allow for irrigation, land quality, etc.) as ‘too restrictive’ and has edged the upper borderline of ‘small farmers’ higher. In one IDA project, eligible ‘small farmers’ could cultivate up to 14 acres, so that, by project completion in 1977, only the top 10 to 15 per cent can have been excluded from the ‘small’ group (while the smallest half of farmers could not qualify for credit). In another state, the upper limit of ‘small farm’ for project purposes is set so high that 91 per cent of cultivators in the state are ‘small’—while the smallest 65 to 75 per cent of cultivators were too small to be eligible for credits.14 The donors are, of course, fully aware of such problems. They affect not only rural credit projects, but most rural and some urban projects, since these contain large credit components (to help beneficiaries acquire inputs before outputs are bringing in income). The de facto exclusion of the poorest springs largely from the concern of Indian intermediaries to keep administrative costs down, and to secure good collateral.15 Overall, despite the clear intentions of the Bank group, the facts cast some doubt on its claim (in one impact study) that ‘the main effect of IDA participation was on the qualitative side, i.e. increased targeting of credit to small farmers’. The lending conditions excluded the smallest Meanwhile Indian intermediaries have edged up the limit that defines ‘smallness’. Indeed, in the late 1970s an evaluation complained that ‘there seems to be little incentive at present for lending agencies to assist small farmers’. This is reinforced by the high-tech, labour-displacing emphasis of much investment linked to these credits. It might appear that the large anti-poverty outlays under the Sixth and Seventh Plans, especially via IRDP (Chapter 2), would redirect NABARD, as the apex of credit intermediation, towards the poorest farmers; but the great bulk of such programmes, and associated credit, support the acquisition of non-farm assets. In India’s large semi-arid areas, where (as in most of Africa) the very poor are still mainly smallholders, a credit gap can emerge, especially after a new form of technology has been introduced (e.g. some HYVs permit such people to get less poor if, say, fertilizer credit is available). It is especially unfortunate—since smaller borrowers are normally the most intensive users of land, loans and water (Berry and Cline, 1979; Booth and Sundrum, 1984)—that definitional laxity at the upper end, and credit exclusion at the lower, direct so little Bankassisted ‘small farm credit’ to the poorest 50 to 60 per cent of Indian farmers. Of course, though not the poorest, a farmer with 3 to 5 acres is in most of India far from adequately housed, clothed, or educated. But farmers with 0.5 to 3 acres need working capital even more. They are probably better at using it and at repaying the loan. They need all the help they can get—from GoI, NGOs, and donors—in combating the strong local forces that tend to exclude them from formal credit. The regional direction of aided projects also greatly affects their poverty-focus. This applies internationally: probably the best single thing that donors could do for poverty-focus

148

Does Aid Work in India?

is to shift bilateral aid away from some of the heavily over-aided middle-income countries towards the really poor countries in South Asia and in Africa. Bank Group aid, especially through IDA, expressly aims to offset some of this huge imbalance of bilateral aid against large, poor LDCs. To help to offset similar imbalances within India, IDA support of ARDC was used in the late 1970s to open a dialogue to determine where aid for rural credit could contribute most to the less developed states. A major Bank-aided project was intended to direct one-quarter of the funds to nine such states, but in fact forty-two per cent was spent there. If—as the evaluation of this project suggests—ERORs on such credit projects are about as high as on others, such regional focus improves poverty impact, not only because average income of beneficiaries is less in poor regions, but because distribution within backward rural areas is less unequal (Dasgupta, 1977). For both reasons, the bias of many aided projects against the poorest farmers (p. 184) may have less impact in these regions. Also, the proportion of landless labourers—missed out, of course, from farm credits—is smaller there than elsewhere in India. Cash aid to poor states, however, is not a simple way for donors’ projects to benefit poor people. Some poor regions are associated with systematically bad project performance, (see p. 161). The evaluation mentioned on pp. 184–5 points out that remote areas may feature ‘physical inaccessibility of a large number of target group farmers’ or other poor people. Conversely, some sorts of project location, while efficient, provide poor people with special difficulties of access. Two evaluations of marketing projects pointed out that markets selected for aid appear to be the central or intermediary rather than the primary or farmer market type, so that only farmers living close to the market or larger farmers with adequate transport and other incentives would be expected to bring their produce to them. As in most of Africa, so in these more backward, dispersed parts of India: high ratios of very bad land to people require a new approach to infrastructure if the remote poor are to be reached at acceptable unit cost. Development of transport types and organizations less subject to economies of scale and of agglomeration could be relevant here. Analogously, Bank and US AID experience with rural electrification shows that forced-draft dispersion of grid-styte projects—power lines, major transport, and market links—to remote areas can be an inefficient, and ineffective, way to relieve poverty there. More imaginative research and development of local non-grid systems may be more promising, especially in view of the very strong (albeit not causally simple) association between poverty incidence and access to key infrastructural items (Wanamali, 1986). These PPAs exemplify three types of approach to ‘including the poor’ at appraisal and evaluation: technical, institutional, and market. The accepted long-term remedy to the issue of technical choice, more research into poor people’ s technologies, is exemplified by the Indo-UK Dryland Farming Project, but has often been hampered—e.g. in the cases of improved animal implements, and of some bio-gas technologies—by lack of early integration between technical and economic analysis, and of timely field-testing for robustness and profitability at small-unit level. Market approaches to ‘reaching the poor’ are exemplified by water sale from big to small farmers. This has partly corrected for the initial bias, from location of wells on betterendowed, less risk-averse larger farms, and from restrictions placed upon subsequent small-farm well development by well-spacing rules aimed at groundwater conservation. In some cases, water sales were not considered at appraisal; two OED evaluations state

Project Aid to India

149

that, if they had been, their redistributive effects could probably have been increased by appropriate design decisions. Similarly, the World Bank and Canadian site-and-service and slum-upgrading schemes initially sought to forbid renting-out, so as to avoid the emotive accusation of diverting poverty-focused aid to slum landlords; but the prohibition was dropped after evidence accumulated that rentals of rooms permitted both fuller use of capital and, via increased supply and hence lower prices of rented property, some sharing in benefits by the very poorest urban dwellers. Market approaches in the other direction—enabling the poorest to own groundwater and pumping rights co-operatively, for sale to better-off farmers (as with Proshika in Bangladesh; see Wood, 1984), or to develop artisan assets at household level through IRDP in India’s Sixth and Seventh Plans—can also offer promising levers upon poverty alleviation in the context of aid projects. However, a crucial weakness of this approach—in the hands of most official donors, including the Bank group—is their scant interest in specifically ‘pro-poor’ market actions, as opposed to less directional types of marketing projects and market improvement. Yet the widespread fear that undirected markets harm the poor is not as absurd as is sometimes claimed. Participation in most markets requires a threshold endowment—an ox cart, numeracy, knowledge—which the poor are especially likely to lack. Especially in remote areas; or where there is great initial inequality; or where the providers of some commodities (e.g. transport to market) can apply pressure by using their control of others (e.g. credit); or where big operators can affect licences (e.g. of urban retailing) or price policies—in such cases market development, if it is to benefit the poor, must be supplemented by direct Government provision or competition, or indirect government actions to secure these. Unregulated market development may well pull key inputs away from pro-poor providers. A Bank study in one Indian state of private and public tubewells installed up to mid-1986 (many of which were backed by Bank-funded credit or equipment) concluded that ‘without public-sector development, the weaker elements of the farming population may be permanently excluded from irrigated agriculture’. If prices in other markets are set to meet overall policy requirements—which in every country include the placating of the powerful—‘improvement’ of a specific market may be inefficient by the same process that renders it inequitable; in the above case, private tubewells used twice as much (heavily subsidized) electric power per acre served (four times as much at peak periods), thereby bleeding the rural grid in general, and the more poverty-oriented public-sector tubewells in particular.16 Institutional approaches towards ‘reaching the poor’ are also exemplified in Bank Group aid projects. However, several of these, even in India, have shown insufficient regard for the role of existing institutions in concentrating benefits upon the already wealthy. At least one evaluation points out that, in pre-appraisal, the likely distribution of gains from new irrigation devices was ignored—and farm models were developed simply assuming scaleneutrality—despite clear evidence of great initial inequality. We have already highlighted lending rules, and even definitions of smallness, excluding small farmers and perhaps tenants. Smallness of plots on small farms is a further barrier to irrigation. One irrigation project performance audit recognizes the obstacles posed by fragmentation—both because small farms mean small plots, and because of wastage of land in bunds—but appears to dismiss the remedy of consolidation because ‘Bank land consolidation programmes in other

150

Does Aid Work in India?

projects have been unsuccessful’ and because farmers expressed ‘widespread opposition’ on grounds of soil heterogeneity. Time is certainly needed to win farmers’ confidence, and to test out farming systems that will replace the real risk of fragmentation to small farmers. However, the Bank’s absolute objections appear to ignore the long history—and considerable achievements—of official consolidation programmes in some Indian states, and its prospects to help very small farmers in the context of water resource development (Minhas, 1970). Efficient growth Our review of Bank aid projects in India should close by reiterating their generally high real EROR. We have pointed to some possibly overestimated ERORs, notably on projects linked to provision of tractors, but this is not a general feature. Nor is it plausible to dismiss most Bank-aided projects as largely fungible. GoI would not, in the absence of Bank Group aid, have proceeded nearly as far, as fast, or in the chosen direction, with rural credit, urban drinking-water and sewage improvement, the reform of agricultural water management, reorganized agricultural extension, or the shift of urban development away from subsidized formal-sector housing for not-so-poor civil servants, towards slum upgrading and site-andservice. These were largely efficient projects and they produced substantial extra real GNP. However, almost all evaluations assume that a project’s contribution to real GNP per person depends entirely on its economic rate of return. In fact, the rate of saving out of extra income is also important.17 Even with projects loosely in the banking sector, such as aid to LDBs and ARDC/NABARD, the question of savings mobilization is seldom adequately addressed. It is not just a question of gross savings flows out of extra project income, or through new project institutions. The extent to which such flows are additional—as opposed to diversions from other savings, or from earlier savings channels—also matters. So does the efficiency with which extra savings are channelled into high-yielding and/ or poverty-reducing investments. What is more, the important question about savings is, ‘What difference does the project make to the total uses of income?’ And ‘total uses of income’ include possible contributions to public savings via increased tax payments—just as hidden project returns may include the reduction of real claims on health outlays, as beneficiaries become less poor and ill-nourished. Such savings effects can be important, especially with big area-based projects. A final consideration, co-determining the impact of a project upon efficient growth of income-per-person, is its population effect. Widely-spread growth is known to lead to reduced fertility (World Bank, 1984). With a large localized project in irrigation or urban development, this can be an important hidden benefit—and its scale can be affected by choices at the design and implementation stage. Savings and population considerations in large projects should be allowed for throughout the project cycle, in assessing the choices likeliest to generate efficient growth. At present only the rate of return on capital is considered.

Project Aid to India

151

Some projects from other donors No other donor has either such a large number of projects as the Bank Group, nor such a systematic evaluation procedure. We were able to obtain evaluation documents only from the Federal Republic of Germany (BWZ), the European Community (EC), the UK (ODM/ ODA), and the USA (USAID). Few of these documents contained formal EROR estimates, and many were subjective, relying on project officers, or on beneficiary or implementing organizations. Baseline studies and control groups were even rarer than in Bank Group projects. Nevertheless, the evaluations have three advantages. They confirm that both the generally good performance of aid projects, and their major weaknesses (e.g. in respect of intermediation and of employment analysis), are not confined to the Bank Group. Second, because these non-Bank evaluations are attributable to outside consultants, we are permitted to identify specific projects and places. Finally, these evaluations cast light on types of aid not given by the Bank group, notably on food aid (from USAID and EEC) and on technical assistance (from West Germany and the UK). Although these two types no longer loom large in the total Indian aid programme, they may have lessons for Africa, where their role is much bigger. Aid from USAID The USA suspended project aid to India in 1972, before the development of its current evaluation procedures, and resumed it only in 1978. Therefore, few recent US aid projects in India have been completed and can be evaluated. Employment and income-distribution effects of projects could not be discussed, nor ERORs re-estimated in a major review of two ongoing and five planned USAID irrigation projects (Keller et al., 1981).18 The substantial sums spent in evaluating the oilseed growers’ cooperative project (NDDB et al., 1983), which was funded from counterpart funds to GoI sales of US soybean oil aid under PL 480 Title II, appear to have been wasted. This highly favourable review was executed by the two agencies which received jobs and cash from implementing the project, plus the two financing agencies. The implementing agencies did their proper job, showing their results in a good light to their visitors. The most useful USAID project review concerns food aid: in particular the US food aid being used in school feeding (SF) programmes. In FY 1984, this cost US$27.5m and reached 6.7 million beneficiaries, as against US$54.7m (six million beneficiaries) for Maternal and Child Health, and US$34.8m for food-for-work (US AID, 1984, 4). Conventional wisdom is that SF programmes have little nutritional impact on schoolchildren (and/or do not reach the hungriest, who do not attend school) and should give place to other uses for food aid. This study (pp. ii–iii, 25–31), however, shows that school enrolment, especially of girls and in backward districts, is in India increased substantially by SF projects.19 The balance between the three main types of food aid projects—food-for-work, school feeding and maternal and child health—may therefore need reassessment in India. Certainly it cannot now be assumed that food aid, whatever the projects supported, must self-destruct by depressing producer prices, with consequent disincentives to domestic food production. Moreover, with India no longer a net importer of grain, food aid should be seen mainly as a form of balance of payments support (Maxwell and Singer, 1979)—normally from donors whose food surpluses are by-products of their farm support policies, would glut

152

Does Aid Work in India?

grain markets if not given as aid, and are not, if cut, likely to be replaced by other and theoretically ‘better’ forms of aid. Unlike most forms of balance of payments support, food aid can to some extent be targeted upon needy groups, or can indeed be self-targeting. Further—though a ton of food aid increases food availability by only 0.8 tons (Blanford and Plocki, 1977)—this can be concentrated on increasing food entitlements for those at greatest risk (Sen, 1981). EEC, WFP and ‘Operation Flood’ In view of the difficulty of directly targeting food on the needy (Beaton and Ghasseimi, 1982), food aid should ideally consist mainly of ‘self-targeting’ foods—and/or should be sold by recipient governments to finance self-targeting projects. Cereals, especially cheap foods, would seem likeliest—both as aid to be eaten by the poor, and as extra project output—to enhance income of small farmers, and employment of rural labourers. However, for several years Operation Flood (OF) has provided most of food aid to an India now a net exporter of cereals via OF, the EEC/WFP provides India with milk powder and butteroil—foods that loom much larger in rich people’s budgets than in poor people’s. GoI sales of these products are then used by NDDB for projects to raise dairy output and processing. These are seen as good ways to help poor producers, because in some parts of India considerable numbers of the landless poor either own one or two animals each or are being, or could be, helped to own them via IRDP loans. Is OF a sensible use of aid? Does poor people’s actual or potential producer benefit outweigh the facts that cattle use land in ways requiring more capital and less labour (per acre and per unit of output) than crop production, and produce foods consumed mainly by the better-off? One would expect objective evaluation of OF, given the importance of the flows. In 1970–81, Rs. 1.5 billion of aid went to the first phase, OF I, Rs.1.2 bn as counterpart funds from GoI sales of WFP/EEC dairy products—comprising US$150m from the EEC alone. A further Rs.2.3bn ($275m) from EEC counterpart funds, and Rs. 1.7bn from IDA, were projected for OF II, which began in 1978. In 1981 alone, and from EEC alone, Rs.448m of OF aid (48m European Units of Account) were transferred (Zurek, 1982,4,16, and 25). A UN Inter- Agency Mission visited India for a month to evalute OF I and produced a lengthy terminal evaluation (Jasiorowski et al., 1981), and—in effect—a one-man mission later evaluated EEC food aid to OFI and OF II on the basis of three weeks in Delhi and Gujarat (Zurek, 1982). There has also been a great mass of interim evaluations, and of papers by representatives, supporters, and critics of NDDB—an excellent bibliography appears at the close of a highly critical and rather polemical review (by S.George, 1984). However, the central questions remain: have OF operations achieved an adequate EROR? Have benefits from OF gone mainly to poor people? Neither of the evaluation missions was asked in its terms of reference to produce an EROR or any other overall benefit/cost analysis (Jasiorowski et al., 1981, Annex III). Indeed, neither the WFP evaluation of OF I, nor the EEC evaluation of its food aid to India (entirely to OF in respect of counterpart funds), assessed even cost-effectiveness in regard to OFI or its components. Yet a standard EROR on OFI is feasible and most desirable in view not only of the large sums spent on OFI, but also of their even larger expansion, which is replicating elsewhere the central features of OFI: a rather high-tech approach, based on

Project Aid to India

153

exotic crossbred cattle, greater dependence on purchased feed, and state-of-the-art milk processing technology; plus apparently cooperative, but in fact centralized, management institutions. Plainly, the Bank Group estimated an EROR on OF II at appraisal, before committing US $150m of IDA aid to it (Zurek, 1982,15); but a basis in objective economic evaluation of OF I is not reported by the WFP or EEC teams, and no OED documents on OF are yet available. The WFP evaluation confirms that ‘by far the greatest part of milk is used by the medium and high income groups’ (Jasiorowski et al., 1981, 77), and that ‘milk would be too expensive a food, even at the sale price from village co-operatives…for promoting increased consumption [by the poor] in preference to other nutritious, cheaper foods’ (p.69). The claims that OF aid benefits the poor, therefore, rest upon production of OF outputs. There are two problems of consistency. First, suppose that poor rural producers and employees, engaged in dairy production in OF areas, gain substantially from OF. This must be set against the loss by poor people, on a national scale, from the diversion of resources towards increasing the output (and cutting the price) of costly dairy calories for the better-off, and away from doing so in respect of cheap cereal calories for the poor. This diversion is especially serious because marginal land which NDDB seeks to divert towards green fodder on private and common lands (George, 1984, ch. 7) is likely to be diverted from poor people’s activities. Second, even assuming that NDDB’s activities unambiguously, even cost-effectively, help the poor dairy producer, such benefit is nevertheless undermined by the financing of NDDB through ‘free’ aid in the form of milk products, which when marketed must reduce producer prices. With butteroil, dairy aid is about 2 to 3 per cent of domestic dairy production (Zurek, 1982, 25), and much milk production is consumed by the producing family. So this 2 to 3 per cent ratio of dairy aid to dairy output could well mean a 4 to 7 per cent ratio of dairy aid, by value, to marketed dairy products in India. On reasonable assumptions about price-elasticity of demand, that could reduce producer prices by anything from 8 per cent to 21 per cent. OF ‘seeks to help ten million rural families’. Yet OF I reaches only 1.33 million, and scheduled-caste and landless households are badly under-represented (ibid., 4, 42; Jasiorowski et al., 1981, 63–5). Even if OF/NDDB benefits only 10 to 15 per cent of the landless poor, this is better than most government production-oriented activities, aided or not Even if the poorest 76 per cent of farmers provide 23 per cent of NDDB milk, that is a creditable outreach. It is wrong to attack NDDB/OF, on these data, for not reaching the poor. The true criticism is that neither NDDB/OF nor their evaluators have estimated the net benefits to poor people (or others), or the costs of achieving such benefits, in an objective way. Also, we do not know if some districts, sub-projects, times of year, or activities—in this large project, now being replicated with massive EEC/WFP/IDA involvement in OF II—are better or worse than others at benefiting the poor. If poor milk sellers do benefit, by how much? ‘A family’s herd average is 1.3 cows and/or she-buffaloes, each producing 1 to 2 litres daily’.20 Obviously, so little milk cannot support a very significant proportion of income for typical rural families of five to seven persons; a person-day of cattle work normally produces about 10 litres of milk (Jasiorowski et al., 1981, 65).

154

Does Aid Work in India?

Although ‘the poorest 20 per cent [of Indian villagers] own less than 5 per cent of the milch stock’ (George, 1984, ch 7), these poorest 20 per cent own far below 5 per cent, closer to 1 per cent, of farmland, and enjoy only some 5 per cent of total village income. Some evidence has been produced to indicate that the increases in income were generally accompanied by widening income disparities (George, 1984, ch 7) but this in no way rules out major, NDDB-related income gain for poor milk- producers. Perhaps most worrying is the fact that—despite evidence (Moore, 1978) that the mini-herds of small and landless milk sellers make more cost-effective use of scarce land and capital than bigger herds—NDDB/ OF technology may greatly and artificially raise unit costs and risks in cattle production, so that it may well redirect advantages (and perhaps also hidden subsidies) towards larger farmers (George, 1984, ch 7; Jasiorowski et al., 1981, 63). Such technology may displace unskilled labour. The estimate of one extra personday of work per ten daily litres of extra milk must be set against the fact that ‘the new [NDDB] system has replaced to some extent the old methods, [so that] certain employment opportunities have been reduced’ (Jasiorowski et al., 1981, 65). Also, to the extent that extra milk is procured from small family herds, owners will tend to supply their own extra labour—especially if landless, and above all in the slack season. As in other cases (see p.181), so with milk production: proper analysis of a project’s effectiveness in raising employment cannot be based on partial-equilibrium comparisons of labour requirements from ‘before’ and ‘after’ new milk technology alone. Objective evaluation of EROR, poverty, non-owner employment effects, and replicability of this large, expanding and heavily aided scheme is now essential. We hope that ongoing research at the World Bank may provide the necessary information. At present OF is a costly, promising initiative—probably with a satisfactory EROR, moderate poverty impact, and limited replicability—damaged by the confusion of public relations with evaluation, so that learning effects are lost. The story of OF suggests that both the EEC and (with the striking exceptions of IBRD/IDA and IFAD) the UN system may well, in their world-wide as well as Indian operations, damage the prospects of their aid projects by the imperfections of their evaluation processes. UK aid projects in India Around 1983, the UK was spending about £0.5m on evaluation (commissioned and in-house)—well below 0.05 per cent of UK aid—and produced only twenty or so evaluations yearly worldwide, of which about a tenth remain confidential (Cracknell, 1984, v, 3,29). India now receives 35 to 45 per cent of UK gross bilateral aid. Already in 1976–80, UK gross bilateral project aid to India averaged over £40m a year (ODA, 1981,46; 1982,46; 1983, 34–5). For 1981–5, the average rose to over £103m yearly, plus £8m under the aid and trade provision (ODA, 1986, 40).21 We saw thirteen evaluation documents published since 1974 (covering completions from 1973 to 1985) but totalling only £21m. Of these documents, two (Wilmshurst and Stevens, 1975; Henderson, 1974) covered ongoing programmes with a British aid input, rather than aid projects. The other eleven evaluation documents (published from February 1974 to September 1986) are summarized in Table 5.4. One was really a pre-appraisal. Five were of small research-cum-training TC operations, totalling only £831,000 of aid. Two covered small ‘appropriate technology’ projects, absorbing £118,000 of aid.

Project Aid to India

155

Therefore, we could review only three evaluations of substantial UK-aided capital projects in India. ERORs had been calculated by ODA’s evaluators only for these, a decision discussed below. ERORs at evaluation were estimated at 12 per cent on the totally aid-financed £4.5m tractor project (EV 36, and see below), probably an overstatement; at 5 per cent on the £52m pair of fertilizer projects (£7m being UK aid; see discussion of EV4 below); and at 18 per cent on the fertilizer manufacturing project, EV 308, also costing £52m (of which £8.5m was UK aid). Except for the Dryland Farming project (EV 282) and the appropriate technology projects (EV 325), poverty impact was not assessed at evaluation, nor does it appear to have been formally considered elsewhere in the project cycle. The IFFCO fertilizer project consisted of plants at Kalol and at Kandla (Snowdon and Stafford, 1980). It was a two-donor project, with US$20m (£8.4m) from the USA and £7.0m from the UK. Indian sources met the rest of the £49.5m planned (£52.7m actual) capital cost Kalol produced an EROR just over 12 per cent but Kandla produced a negative EROR. This ‘unpacking’ of project components at evaluation, so that separate ERORs can be estimated and explained, is valuable and should be commoner. Kalol was on time, due to competitive tenders confined to ‘pre-qualified’ bidders. Kandla—which would have been stopped by correct appraisal arithmetic—involved risky, interdependent high-tech, in process and scale. Further delay in operation, until a period of low demand, was due to a switch from US to UK aid (ibid., sections 3.2–3.7). Responding to the House of Commons Select Committee (HOC, 1979) which criticized non-evaluation of large capital aid in India, ODA in late 1986 published an evaluation of a major fertilizer expansion project (no. 9 in Table 5.4). It clearly showed good economic and financial returns, and was sustainable and environmentally careful. However, there are several disturbing features, of which the evaluators (Haley and Hesling, 1986, 9, 12) point out only the first two: the captive power plant combined high-tech with GoI insistence on domestic origin, leading to major delays, cost escalation, and hence bad returns; and a wide range of estimates for actual aid disbursed. Capital aid is £6.3m by ODA expenditure records, but only £4.3m on project files (Haley and Hesling, 1986, para. 5.2). Most seriously, though just over half the equity in Indian Explosives Limited (IEL) was (and is) held by ICI (UK), the terms on which the aid money was loaned forward by GoI to IEL—and thus, indirectly, by the UK aid programme to British shareholders—are not specified. Of the project’s costs not covered by aid, part was met by a debenture issue (Haley and Hesling, 1986, para 1.3); in the absence of clear information about on-lending terms for the UK aid, it must be inferred that it was used, at the margin, as a cheaper source of money, enabling ICI and/or IEL to reduce their debenture-borrowings. The two fertilizer projects and the tractor project comprised £20m of the £21.2m UK aid credits in the evaluations of Table 4. The fertilizer projects were adequately evaluated. However, seriously inadequate resources for evaluation of employment, and indeed of efficiency, mark the 1975 evaluation of aid-financed export of 3,850 British-made tractors to India in 1971–4. The method (Dalton, 1976, vii, 6–7) was a small survey in two months in 1975 in which 177 of these tractors were located. The tractor distributors provided transport, interpreters, and introductions to local dealers ‘[who often joined] our visit to the farmers… At the interview [,b]rief notes were made to be written up afterwards’. The possibility of statistical evaluation was thus precluded.

The training of young Indian scientific workers: 5 PhD’s plus 6 under way

1 Birkbeck/Bangalore chemistry link (EV 231) 1982

Opening date

1971

1975–6

Indian University April 1971 Grants Commission

Executing agency

2 Indian manpower Training proGoI project (EV 9) 1977 grammes for about 60 middle and senior level officials at GoI in the UK, designed to improve the management and administrative skills of the participants. 3 British aid tractors 3,850 tractors GoI in India (EV 36) exported from the 1976 UK to India, under a UK/India loan agreement Repayments made by the GoI over a period of 25 years with no repayments during the first 7 years.

Main objectives

Project and date of evaluation

Table 5.4 Some UK-aided projects in India

3