State Capability in India 0192856618, 9780192856616

Citation preview

State Capability in India

State Capability in India T. V. S OM A NAT HA N A N D G U L Z A R NATA R AJA N

1

3 Great Clarendon Street, Oxford, ox2 6dp, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Oxford University Press 2022 The moral rights of the authors‌have been asserted First Edition published in 2022 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2022934672 ISBN 978–0–19–285661–6 DOI: 10.1093/​oso/​9780192856616.001.0001 Printed in India by Rakmo Press Pvt. Ltd Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

T.V. Somanathan: To my supervisors from whom I have learnt so much: J.C. Gordijn, A. Muttulingam, K.R. Koroma, D. Banerjea, Amitabha Bhattacharya, S. Machendranathan, K. Ashok Vardhan Shetty, M. Sathiyamoorthy, P. Sivakami, N. Narayanan, Capt. A. Nagarajan, R. Balakrishnan, Santha Sheela Nair, S.A. Subramani, J. Shivakumar, Robert M. Voight, Robert Van Pulley, Anil Sood, Achim von Heynitz, Fayezul H. Choudhury, Dr. Chhattrasal Singh, T.P. Nagarajan, T.R. Ramasamy, Syed Munir Hoda, K.S. Sripathi, S. Malathi, R. Bandyopadhyay, D.K. Mittal, Chuck McDonough, Bertrand Badre, Nripendra Misra, Dr. P.K. Mishra, Dr. Girija Vaidyanathan, Rajiv Gauba Gulzar Natarajan: To my parents—​Suseela Natarajan and K.G. Natarajan—​to whom I owe a debt of gratitude that can never be repaid.

Foreword India looms large. It vies with China for a place as the world’s most populous country. India is home to the world’s biggest democracy. Depending on the measure, it ranks between third and sixth in economic size. Its vast territory ranges over a subcontinent, with plains, mountains, and rivers that define much of the immense landmass of Eurasia. Indian entrepreneurs span the globe. India matters. Readers will find many books about Indian politics, economy, society, and grand strategies and policies. Yet few books have examined a fundamental issue: India’s capability as a state—​a state capability that connects policy designs with execution. This gap in policy studies seems especially ironic—​and unfortunate—​because students of India acknowledge pervasive failures of implementation. Good intentions, high hopes, and big plans have often produced only weak outcomes. T.V. Somanathan and Gulzar Natarajan, the authors of ‘State Capability in India,’ are scholar-​practitioners. Dr. Somanathan worked with me at the World Bank, where he adroitly and enthusiastically applied his executive skills to guide the administration of one of the biggest multilateral institutions. Together, the authors combine a deep knowledge of the international scholarly literature with examples from practical experience. Somanathan and Natarajan are wary of managerial panaceas or ‘how to’ instructions; instead, they offer wise guidance based on analysis, careful reasoning, and good judgment. They are honest and astute observers of their homeland. In recent years, Francis Fukuyama at Stanford, Lant Pritchett at Harvard, and others have directed attention to factors that increase governmental effectiveness. For many countries, democratic governance alone has not established sturdy legitimacy. Governments need to demonstrate that they can do things—​produce results that their citizens value. This scholarship has revived interest in how institutions contribute to good economic performance—​and thereby has prompted economists to recall topics that were once studied under the heading of political

viii Foreword economy. The successes of export-​led economies across East Asia also encouraged examinations of which policy and governance ideas worked—​ and why. Somanathan and Natarajan have researched questions of state capability, the element of policymaking that has received the least attention. Although their laboratory is India, the issues they analyse appear across developing—​and even developed—​economies on other continents as well. This book is both Indian and international. The book begins with surveys of increased corruption and decreased competence across the Indian state at various levels. The authors examine possible weaknesses in the organization and structure of government, personnel, and technological competence. They suggest that the primary problem of the Indian Civil Service is not a bloated bureaucracy but rather a stretched, constrained, and continually audited force that has become reluctant to exercise judgment and decide. The Indian judicial sector, reflecting numerous constitutional priorities, has assumed major roles in the execution of policy—​largely without the responsibility of coherence and the imperative for action. The pursuit of perfect fairness, admirable in spirit, extracts a cost to the public good. Somanathan and Natarajan explain how the Indian system has led to decision-​paralysis. The understandable interest in coordination has at times led to a highly centralized system that disempowers line departments and officers facing problems in the field. Procedural fixes can help but do not really solve the question of balance among central and local roles, authority, and accountability. The authors examine the possible benefits, but also the limitations, of private-​sector solutions. While recognizing the usefulness of incentives, competition, and markets, the book explains that private-​sector efforts are often constrained by concerns about political acceptability, administrative feasibility, and the complexity of subjective judgment. Privatized systems that ignore these factors often break down. Somanathan and Natarajan are practical men. Their ability to diagnose weaknesses and organizational maladies has led them towards a path of humility, not acceptance of futility. Development is hard. There are no silver bullets or ideal processes. Unintended consequences abound. Therefore, the authors rightly stress attention to the context of problems. They recall Charles Lindblom’s thesis of ‘muddling through,’ not as

Foreword  ix a description of struggling systems but instead as cautionary guidance for implementation amidst complexity. They seek flexibility for feedback loops, iterations, and adjustments. This book does not want officials to succumb to avoid risks; it wants them to recognize that in a world of ‘second best, at best,’ a search for perfection can stymie the good. People tackling hard problems should be on watch, identify mistakes, fix them, learn, and keep trying—​with open eyes, ears, and minds. Somanathan and Natarajan help by sharing their experience—​from study and practice. The authors urge support for successful champions of positive change. Those leaders in the field, or in specific sectors, have usually learned to account for multiple factors. Their successes further cumulative effects. With a number of good examples, reformers can create a group that adds up to more than the sum of its parts. Somanathan and Natarajan point to the value of pilots, competition, and anchor interventions. Simplified policy designs increase the likelihood of success. They favour processes that I once described as an ‘incremental strategy to achieve radical reform’. Institutions and adaptations together can create change simultaneously. Over time, the state can improve integration among successful reforms. Officialdom also needs to be open to the flow of new knowledge from outside the government as well. Somanathan and Natarajan are patriots. They are also internationalist in their search for suggestions and expertise. They know the India of the past and understand, sympathetically, well-​intentioned efforts that went awry. They recognize India’s potential—​the India that can be, for both Indians and the world. They have written a thoughtful and practical book to help India in the 21st Century. Robert B. Zoellick, Former President of the World Bank

Preface Governments, especially in developing countries, are often unable to deliver what their policies promise. In many cases the constraint is capability, not financial resources. The state is complex and massive. It is difficult for outsiders to understand how it functions, while it is difficult for career insiders to appreciate how other organizations function. This mars the quality of many discussions on how to make the state more effective. As officers working at different levels of Government in India, both of us have grappled with trying to deliver what policies say ought to be delivered and to formulate policies that can actually be delivered. Sometimes we succeeded and sometimes we failed. Our failures may reflect our own incapabilities as thinkers, policymakers, and administrators, but they also led us to a deeper study of how to improve the capability of the state. This book is the result. Though several parts of the book are primarily focussed on India, it tries to learn lessons from everywhere and hopefully offers some insights applicable in other countries. Some parts of this book draw upon our own earlier writings over the years, but most of it was written in 2018–​2019. The views expressed in the book are our personal views and do not necessarily represent, and should not be construed as representing, or as being critical of, the views of the Government of India or of any state. Books written by bureaucrats are not usually ‘light’ reading; however, our brilliant and versatile senior colleague Dr. V.S. Gopalakrishnan, of the 1962 batch of the IAS (Maharashtra cadre), has published five books of cartoons under the name ‘Gopal’. So, we decided to leaven this book with timeless cartoons from two of those books, Bureaucrat goes Bonkers and Bureaucrat goes Bananas (published in 1990 and 1993), which look at the lighter side of the issues we have discussed. In 1993–​ 1994, a working group of officers from four states, created to assist the National Institute of Public Finance and Policy in preparing a blueprint for the implementation of Value Added Tax in the states, met many

xii Preface times. Dr. Gopalakrishnan and one of us (Somanathan) were part of that working group and it was then that he had generously given a copy of these books. We are very grateful to him for permission to use them. We wish to thank the Oxford University Press for permission to use material from earlier publications, including from the chapter on the Civil Service in Rethinking Public Institutions in India (OUP—​2017) and The Oxford Handbook of the Indian Constitution (OUP—​2017). We are grateful for several insightful comments on an early draft from Devesh Kapur, which have enriched this book. Special thanks are also due to Mr. G. Raguraman, Principal Private Secretary, and Mr. S. Senthil Kumar, Senior Systems Analyst of the Tamil Nadu Secretariat, for their assistance in their spare time in different ways. One of us (Somanathan) has had the privilege of working with Mr. Raguraman at different times in six different posts from 1994 to 2019; Mr. Raguraman’s superb skills, absolute dedication, spotless character, and selfless public service have played a significant role in some of the experiences which have contributed to this book. Mr. Senthil Kumar helped with a number of technology issues. If this book helps any of its readers contribute to improving the performance of any part of the state, in India or elsewhere, we would feel richly rewarded. T.V. Somanathan Gulzar Natarajan

About the Book The deficiencies in the capability of the state to design and implement effective policies are arguably the biggest development challenge facing developing countries like India. This book seeks to assess state capability in India, identify weaknesses in policy design and program implementation and their causes, and propose some measures to remedy them. Importantly it does so while recognizing political economy constraints and focusing predominantly on the administrative contributors. To this extent, the book’s suggestions are practical enough for adoption by stakeholders at different levels. This book describes the institutional design, constitutional provisions, the organizational structure, and the personnel of the Indian state. It covers a wide spectrum of aspects impacting state capability ranging from ideological narratives and systemic constraints to procedural and personnel management issues to the behaviours and attitudes of individual bureaucrats. It offers a new analytical framework to think about the effectiveness of the state and on the policymaking process. It also offers a nuanced perspective and suggestions on many of the popular themes in public administration—size of the state, generalist and specialist debates, lateral entry, digital monitoring systems in governance, outsourcing and private participation, use of consultants, risk aversion in bureaucracies, performance-based incentives, program evaluations, and so on. Finally, being participants and observers in the bureaucratic system, the authors describe reality without always seeking to locate it in the framework of existing academic literature, thereby offering fresh insights and enriching the discourse on state capability.

Contents PA RT I   I N T R O DU C T IO N 1 Introduction 

1.1 State Capability  1.2 Determinants of Development Outcomes  1.3 Weakness in Policymaking and Program Design  1.4 Weakness in Implementation  1.5 Public Policy Design and Implementation in Perspective  1.6 Scope of the Book  1.7 Structure of the Book 

2 State Capability: A Conceptual Overview 

2.1 What Is State Capability?  2.2 Policy vs. Execution  2.3 The Difference Between Policymaking, Program Formulation, and Program Implementation  2.4 Flexibility Is Essential—​but May Be Impossible  2.5 The Overlap Between Policy Design and Execution  2.6 Addressing State Capability Failures 

3

4 4 8 9 12 14 15

17

17 23 24 25 25 28

PA RT I I   T H E I N ST I T U T IO NA L D E SIG N O F T H E I N D IA N STAT E 3 Constitutional Provisions and Structure of Government 

33



42 43 45 45 46



3.1 Constitutional Provisions on the Executive  3.2 Ministerial Responsibility  3.3 Organizational Structure—​Who Rules India?  3.4 Regulatory Bodies 

3.4.1 Constitutional Position of Regulatory Bodies  3.4.2 The Need for Regulatory Bodies: A Conceptual Overview  3.4.3 Regulation Through the Traditional State  3.4.4 Independent Regulators  3.4.5 Increasing Influence of the Legal Aspect  3.4.6 Trade-​Off Between Consultative Processes and Conventional Political Representation 

33 35 36 41

47

xvi contents

3.4.7 Advantages and Disadvantages of Independent Regulators  3.4.8 Regulatory Institutions: An Economic Analysis 

48 50

4 Personnel of the Indian State 

53

5 Civil Service Effectiveness: An Assessment 

65

4.1 The Higher Civil Services  4.2 The Design of the AIS  4.3 The Intermediate and Lower-​Level Services 

5.1 Criteria of Effectiveness  5.2 Assessing Effectiveness of the Indian Civil Service  5.3 Causes of Poor Performance  5.4 Misaligned Incentive Structure  5.5 Recent Trends Affecting the Civil Service 

53 58 60 65 67 71 78 80

PA RT I I I   STAT E C A PA B I L I T Y C HA L L E N G E S A N D W E A K N E S SE S 6 The Size of the Indian State 

89

6.1 The Bloated State?  6.2 How ‘Bloated’ Is the State: A Reality Check  6.3 Why Does Much of the State Remain Thin? 

89 91 105

6.4 What Are Possible Solutions to the Thin State?  6.5 Recognizing the Reality 

111 115

7 Challenges in Policymaking 

119



127







6.3.1 Litigation on Personnel Management  6.3.2 Long Delays in Recruitment  6.3.3 Unpredictable and Hidden Fiscal Costs of Personnel 

7.1 Why Does India Underperform in Policymaking?  7.2 Good Policymaking Process: A Conceptual Overview  7.3 Translating Theory into Practice—​Managing the Trade-​Offs  7.4 Weaknesses in India’s Public Policymaking  7.4.1 Excessive Fragmentation in Thinking and Action  7.4.2 Excessive Overlap Between Policymaking and Implementation  7.4.3 Inadequate Non-​Governmental Inputs and Informed Debate  7.4.4 Lack of Systematic Analysis and Integration Prior to Policymaking 

7.5 Specialist vs. Generalist—​or Specialization of the Generalist 

105 109 109

119 121 124 126 127 129 130

132

contents  xvii

8 Decision Paralysis 



8.1 The Problem  8.2 Judiciary  8.3 RTI and Its Effect on Civil Service Behaviour  8.4 Vigilance and Investigative Agencies  8.5 Comptroller and Auditor General  8.6 Finance Function 

9 Challenges in Execution 

9.1 Execution in Perspective  9.2 The Systems Approach to Public Service Delivery 



9.3 The Paradox of Elections and Census  9.4 Coordination vs. Empowerment 



9.2.1 Problems at the Frontline 

137

137 142 156 162 168 172

177 177 180 183

186 190

PA RT I V   P O S SI B L E S O LU T IO N S 10 Addressing Policymaking Challenges—​I 

199



199 201 202



10.1 Structural Reforms 



10.2 Improving Integration and the Flow of Knowledge From Outside Government  10.3 New Public Management Approaches  10.4 Performance-​Based Incentives in Government 





10.1.1 Reduction in Fragmentation  10.1.2 Separating Policymaking From Implementation  10.1.3 Decentralization/​Enhanced Delegation of Authority 

10.4.1 Quantification and Measurement Problems  10.4.2 The Administrative Uncertainty Principle  10.4.3 The Moral Issues With Public Service Delivery  10.4.4 Institutional Culture, Incentive Distortions, and Unintended Effects  10.4.5 Maintaining Credibility  10.4.6 Scale Dynamics  10.4.7 Conclusion 

10.5 Other Reform Ideas 

10.5.1 Restrict Civil Service Entry to Those Who Have Completed a Course in Public Administration?  10.5.2 Creation of Civil Services Board/​Enacting a Civil Services Act?  10.5.3 Curb Post-​Retirement Opportunities by Aligning Retirement Ages  10.5.4 Training in Specific Skills Using Modern Online Learning Technology 

199

207 207 210 212 213 214

215 218 219 219

219

220 221 221 221

xviii contents



10.6 Personnel Management 

10.6.1 Career Progression—​the ‘Informed Generalist’  10.6.2 Reforming Appointments/​Postings  10.6.3 Promotions into All India Services  10.6.4 Reforming Empanelment  10.6.5 Lateral Entry 

223 223 226 229 229 234

11 Addressing Policymaking Challenges—​II: Overcoming Decision Paralysis 

245



256 259



11.1 Judicial Action  11.2 The Oversight Agencies  11.3 Regulatory Institutions: Future Evolution  11.4 The Minimum Viable Product Approach 



11.5 Designing for Scaling of Complex Public Policies  11.6 Simplifying Policy Design 

263 268

12 Becoming Better Policymakers—​Behavioural Change 

273



11.4.1 MVP in Government  11.4.2 Why the MVP Approach? 

12.1 Bridge the ‘Trust Deficit’  12.2 Enabling Legitimate Private Benefit Is Good Public Policy  12.3 Logic Is Sometimes an Enemy of Prudence  12.4 The World Is Second Best, at Best  12.5 Comfort With Holding Multiple Hypotheses About an Issue  12.6 Comfort With Ambiguity  12.7 Detachment From ‘Achievements’  12.8 Not Generalizing From a Few Experiences or Data Points  12.9 Top–​Down vs. Bottom–​Up Perspectives on State Capability  12.10 Difference Between Partial and General Equilibrium  12.11 Beware of the Law of Unintended Consequences  12.12 Policymaking Is Often About Iterating With a Draft Solution  12.13 Negotiating Policy Choices  12.14 Policies Generally Generate Probabilistic and Not Binary Outcomes  12.15 Willingness to Make Bayesian Updations  12.16 Follow Rational Scepticism  12.17 Tolerate Failures and Allow ‘a Million Marginal Revolutions’  12.18 Avoid the Predecessor–​Successor Syndrome  12.19 Personal Effort Is Not Systemic Improvement  12.20 ‘Do No Harm’ Principle  12.21 Just Improve Basic Administration  12.22 Development Is Very Hard, a Little Humility Would Do No Harm!  12.23 Making the Behavioural Changes Happen 

247 248 252 253

275 277 280 281 283 284 287 288 289 290 292 294 296 297 298 300 301 303 304 305 305 307 308

contents  xix

13 Addressing Execution Challenges—​I 

311



13.1 Strengthening Supervision 



13.2 Electronic Monitoring Systems 

313



324



13.3 Delivery Units  13.4 Greater Emphasis on the Qualitative Dimension of Accountability  13.5 Enhanced Delegation and Discretion  13.6 Use of Social Media 



13.7 Use of Consultants and Advisors 

336



13.8 The Enforcement–​Harassment Trade-​Off 

342

14 Addressing Execution Challenges—​II: Personnel, Procurement, and Permits 

349











13.1.1 Improving Traditional Methods of Supervision  13.2.1 Conceptual Challenges  13.2.2 Design Challenges  13.2.3 Incentive Distortions  13.2.4 Possible Solutions 

13.6.1 Suggestions on Managing Social Media  13.7.1 A Balanced Approach 

13.8.1 Data Analytics: Weak vs. Strong Signals  13.8.2 Lessons for Enforcement Systems Design 

14.1 Personnel Management: Transfers and Postings  14.2 Procurement 

14.2.1 The Rule Framework  14.2.2 Can Corruption Be Reduced by Legal/​Procedural Reform?  14.2.3 Efficiency Issues  14.2.4 Possible Improvements in the Procurement Procedure  14.2.5 To Tighten or to Loosen? 

14.3 Permits and Licences  14.4 The Real Solution 

15 Addressing Execution Challenges—​III 



15.1 Outsourcing and PPPs  15.2 Audits, Ratings, and Certification  15.3 Digitization/​E-​Governance  15.4 Evaluating Programs 

15.4.1 The Case for Evaluations  15.4.2 Limitations of Evaluation  15.4.3 Institutional Architecture for Evaluation  15.4.4 Ensuring Rigour in Evaluations  15.4.5 RCTs and Their Limitations 

311

311

314 318 319 321

329 330 333

335 341 346 348

349 351

352

353 359 360 366

366 368

371

371 378 380 385

385 387 389 391 393

xx contents

16 A Practical Approach to Better Capability 

399



404 407 409





16.1 Encouraging Positive Deviances  16.2 Promoting Anchor Interventions  16.3 Making Disruptive Changes  16.4 Identifying and Using the Cumulative Effect of Multiple Factors  16.5 Keeping Expectations Realistic  16.6 Epilogue 

About the Authors  Index 

399 400 401

411 415

PART I

IN T RODU C T ION

1 Introduction India has made giant strides in the 75 years since Independence in 1947. Administrators in the 1940s worried about famine; today they worry about what to do with massive surplus grain stocks. Till the 1990s, India constantly worried about foreign exchange shortages; today the Reserve Bank frequently intervenes to ‘sterilize’ inflows and worries about how to invest the burgeoning reserves. Every economic and social indicator has shown a large improvement. And, confounding sceptics, this nation of unequalled religious, linguistic, and social diversity has remained a democracy throughout. And yet, most people would agree that India has the potential to do better. What can developing countries like India do to realize their potential for economic and social development? For many commentators, the problems are macroeconomic weaknesses, insufficient liberalization of markets, excessive import tariffs, inefficient public enterprises, weak institutions, inadequate expenditure on social services, etc. However, such explanations do not address the less visible but arguably more important requirements of development. Development that is equitable and sustainable does not happen by way of one-​off decisions (the so-​called big-​bang reforms), useful and even critical though they may be. Poor learning outcomes, shortages of water, inadequate sanitation, dysfunctional health care, traffic congestion, shortages of affordable housing, stagnant farm productivity, pervasive informality, deficient tax revenues—​all of these are examples of typical problems of developing countries without obvious solutions. In some cases, the objective is clear but there is disagreement on the ‘right’ policy. In other cases, the ‘right’ policy is not, or at least less, in dispute, but implementation is difficult. This book argues that state capability is an important but often neglected factor in development. It dwells on various theoretical and State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0001

4 Introduction practical issues relating to state capability, with a specific focus on India, identifying key weaknesses and suggesting possible solutions so that India and other developing countries can attain their full potential.

1.1  State Capability ‘State capability’ can be simply defined as the ‘ability of a government to get things done’. It is the state’s ability to effectively design, implement, and deliver what is promised to its citizens. It is no exaggeration to claim that the biggest long-​run development challenge faced by most developing countries, including India, is the weakness of their state capability. A term often used in broadly the same sense (of the state’s ability to get things done) is ‘state capacity’. The term ‘state capability’ is preferred in this book because capacity can also be used in a purely quantitative sense or to reflect resource availability (as in the term ‘capacity utilization’). Other terms sometimes used include ‘state effectiveness’ and ‘bureaucratic capability’. (Note: In this book, ‘bureaucracy’ is used in the non-​pejorative and purely descriptive sense.)

1.2  Determinants of Development Outcomes It would be wrong to suggest that state capability is the most important determinant of development outcomes. It is also necessary to be clear about what state capability is not. Institutions top the list of determinants in development economics scholarship. They are the rules of the game, formal and informal, that shape interactions between individuals and organizations.1 Strong institutions are considered the primary reason for the economic growth of developed countries, and deficient or weak institutions as holding back developing countries.

1 Douglass C North, ‘Institutions’, Journal of Economic Perspectives, 5, no. 1 (Winter, 1991): 97–​ 112 (https://​www.jstor.org/​sta​ble/​1942​704?seq=​1#pag​e_​sc​an_​t​ab_​c​onte​nts).

Introduction  5 A history of colonization matters too, though this is not often discussed in the modern discourse on development: colonial powers benefitted immensely from the selective imposition of tariffs and from unfair terms of trade with their colonies and this had nothing to do with state capability (the ostensibly capable colonial British-​Indian state did not prevent famine). Resource endowments play a big role in development. Botswana or Kuwait, with mineral riches and small populations, have inherent advantages over India, even if their domestic bureaucratic capabilities may not be better than India’s. Knowledge and expertise, or loosely ‘ideas’, are vital to choosing good policies. Often these come from experts and researchers in academia or others outside government. A good policy for increasing farm incomes needs expertise in agricultural science and in agricultural economics. The power of good ideas is enormous. If institutions, resources, and the stock of expert knowledge are taken as given, politics plays an important part in the choice of policy. Good policy ideas from the best academic experts may have no effect if they are not adopted by governments. The choice or adoption of policies by a government is influenced by politicians in government (Ministers) and by politicians and non-​politicians outside government, in the legislature and outside. Their influence on policy is, for the purposes of this book, simply ‘politics’. If bad policies are chosen through the political process, good state capability will not produce good results. The erstwhile Soviet Union had admirable state capability, including the ability to conquer many nations, put man in space and so on, but it pursued unsustainable economic policies. Indeed, a capable state administering ill-​advised policies may do more harm than an incapable one. Therefore, for the purpose of this book, it may be useful to modify the definition to say that ‘State capability is the ability of a government to get the right things done.’ While policy choice falls partly in the realm of politics, the choice of policy in government is also influenced by the staff in government and their ability to assess and articulate the available choices. This may depend on the competence of the staff, their work culture and so on. Thus there is an element of the policymaking function which is non-​political and this falls within the realm of state capability. While in practice there is

6 Introduction Good Development Outcomes Good Policy

(Macro economic, Micro economic, Social, Technical etc. expertise and knowledge )

Political will to resist/ overcome opposition to good policy

Political Skill to garner support for good policy

Ability to tailor & sequence good policy to win support

Effective Implementation

Capability to choose good policy from available options

State capability

Politics More Political

Capability to execute policies well

POLITICS/ADMINISTRATION CONTINUUM

More Administrative

Figure 1.1  Good development outcomes

some overlap, and the distinction between the political and non-​political aspects of policymaking may not always be apparent, the distinction is conceptually important. A capable state is one that, within political constraints, is good at choosing the best available policies and programs to accomplish the desired purpose. However, the adoption of good policy is not sufficient for the achievement of good development outcomes. There is the crucial matter of implementation—​an excellent policy poorly implemented will produce bad results. The preceding discussion is summarized in Figure 1.1. It shows that for a given level of resources, development outcomes are influenced by knowledge or expertise (the italicized part), by politics or political will (the part labelled ‘politics’), and by state capability (the part labelled ‘state capability’). Realization of outcomes, O, can be characterized as a function in the following manner: O = f ( R, P * S ) ,

Introduction  7 where O =​development outcomes R =​resources P =​policy S =​state capability. P can be further broken down into two components—​knowledge or expertise and experience (k) and political will to implement the available knowledge (p). Thus the function can be re-​written as O = f ( R, k * p * S ) . This book focusses on that last determinant of development outcomes—​ state capability. Clearly, state capability is not the only determinant of development outcomes. However, it is the one that has received the least attention and, in India, is arguably the most important cause of governance failures. Across sectors, public systems do not appear to have the capability to translate even well-​conceived and laid out policies and agendas into the desired outcomes. Despite its critical importance, state capability gets disproportionately less attention in the design and implementation of policies and programs. This is as true of academic research and public debates as of discussions within the government. State capability is usually assumed away as a given, and everything else is built on the existing foundation. But weak foundations beget weak super-​structures. State capability has two dimensions—​policymaking and policy (or program) implementation. Policies define problems, lay down the government’s intent, and outline the guiding principles to achieve it and may include specific actions (‘programs’) to address the problem. Policymaking involves taking a ‘big picture’ view of the problem and committing to specific solutions. In terms of our definition, it relates to choosing the right thing to do. Policy or program implementation is about carrying out policies successfully. The realization of the policy or program’s objectives depends on its execution by the frontline functionaries of the government. It relates to actually doing what has been chosen as the right thing to do.

8 Introduction While we make the analytical distinction between policymaking and implementation, we are not suggesting that they should always be seen in separation. As we discuss later, good policies emerge also from an appreciation of the implementation context. There is an endogenous relationship between the two. Poor state capability may be caused by weaknesses in policymaking, weaknesses in policy implementation, or both. (Note: In this book the term ‘execution’ is often used to denote implementation of both policies and programs and the term ‘policymaking’ often includes design of programs, wherever the context so indicates.)

1.3  Weakness in Policymaking and Program Design Poorly conceived and tailored policies invariably fail. Political considerations are admittedly an important contributor to policy design deficiencies. There are limits to what can be done to mitigate such failures in a democracy, and indeed a degree of policy dysfunction may be a worthwhile price to pay for democracy and the invaluable freedoms that come with it. This book takes politics as a given. But, as already pointed out, there is a significant part of policymaking that is not political and policy dysfunction in India is often not political but merely administrative. Examples of causes include fragmented organization structures, weak and grossly inadequate personnel, and deficient technical competence. There are also distinct features of the Indian bureaucracy, and the career progression and incentives of its bureaucrats, which come in the way of equipping policymakers with the skills to design effective policies. All of these policymaking weaknesses can theoretically be remedied without political implications. Another cause of policymaking weakness is the institutional power shifts that have had the unintended effect of affecting decision-​making at the higher levels of government. Crusading information activists, auditors, investigators, and judges have sought to shine some light on political corruption and bureaucratic complicity and cleanse public life,

Introduction  9 and this has coincided with the growth of 24/​7 electronic news media. Given the high levels of corruption in India, these trends have undoubtedly had a beneficial aspect. However, they have not been without debilitating side effects: they have shifted the expectations on transparency and accountability towards a utopian level which may perhaps be unattainable in any real-​life administration. This has shrunk the decision-​making space available for public servants, including political representatives. Jurisdictional and other excesses by the Courts and other institutions, including the Comptroller and Auditor General (CAG) of India, have created a deep reluctance among senior officials to exercise judgement, accept responsibility, and take decisions. The result is a ‘decision paralysis’ that has gripped many officials at the higher levels of the country’s bureaucracy, both at the state and central government levels. This structural incentive problem is independent of the government in power and is not amenable to being addressed easily through executive or legislative actions. Paradoxically, and contrary to many other countries, this dysfunction is not a result of a lack of institutional independence: it is the unwitting consequence of well-​intentioned actions by strongly independent institutions in areas where they have limited expertise. (To be fair to them, they have often stepped in after the executive has manifestly failed.)

1.4  Weakness in Implementation The second form of capability failure is in implementation. This failure is pervasive and is a major cause of the public perception of weak governance. Poor learning levels; hand pumps in rural areas and water taps in urban areas running dry; dispensaries lacking medicines; potholed roads; delays in fixing a leaking pipeline; piles of garbage; un-​enforced social protection regulations—​all these are common frustrations of Indian citizens. In each of these cases, everyone agrees that ‘it needs to be done’, but it remains undone.

10 Introduction

A word cloud of our infrastructure sector is certain to prominently highlight ‘time over-​run’ and ‘cost escalation’. The capability to design and document projects, manage procurement, mobilize financial resources, finalize contracts, and manage them effectively is often inadequate. Corruption, execution delays, and renegotiations are perceived as widespread in several sectors. There are several reasons for the erosion in state capability at the implementation level. They include politicized and compromised bureaucracies, widespread (though by no means universal) corruption, weakened supervisory monitoring systems, a decline in professional standards, inadequate training, and over-​burdened officials. Falling ethical standards

Introduction  11 and work culture and demotivated and often discourteous employees amplify the downward spiral. In matters of personnel management, well-​intentioned but ill-​informed judicial intervention has contributed to unfilled vacancies and staff indiscipline. Thousands of posts across government are left unfilled because of litigation on issues of seniority among public servants, where courts have ordered stays on filling vacancies. Courts have shown considerable leniency to errant public servants in disciplinary matters. The protection of the rights of bureaucrats is affecting the citizen whom they are expected to serve.

12 Introduction The problem of decision paralysis through excessive activism by watchdog agencies, referred to earlier as a cause of policymaking weakness, is also a cause of implementation weakness. The implementation problem has worsened in the last couple of decades, as there has been a dramatic expansion of the role of government as well as the nature of its focus and expectations thereon. It has moved beyond regulation, public goods, and some welfare services, to cover a much broader range of activities. They include provision of private goods, market development, and facilitation of economic growth itself. In line with the stage of development and expectations, the focus too has shifted from provision of just inputs or outputs to the delivery of outcomes and citizen’s rights and to change human behaviours. Further, instead of directly doing things, government agencies have to manage contracts that involve private production of goods and services. This expansion of mandate and responsibilities has not been accompanied by a commensurate addition of resources, professional expertise, and changes in civil service processes. The result is field bureaucracies that seek to address today’s problems with yesterday’s toolkits and resources.

1.5  Public Policy Design and Implementation in Perspective There are several examples of successful policies. Such successes span across countries and levels of government within countries, and sectors and types of activities. Are there any ‘general’ lessons from them? There are two lessons that are critical to the success of any public policy pursuit. They concern about the WHAT and HOW of the intervention. WHAT should be the basis of the design of a policy? HOW should that policy be pursued? A critical difference between the public and private sectors is the far higher number of constraints facing activities in the former. These constraints emerge from the nature of the systems in which these respective sectors operate. Primarily, many public sector activities are by nature provided only by the government and are universal and cannot be exclusionary. Besides, their design is most often a compromise involving

Introduction  13 competing and contradicting interests, and accountability mechanisms are fewer, more diffuse, and less easily enforceable than in the private sector. In terms of design, a private company can focus on the most technically robust solution. This is, by nature, an exercise that is mostly about technical expertise. Once the problem is identified and diagnosis done, the challenge is confined to finding the best technical solution given the resource constraints. The resource constraints are mostly about the funds and personnel available, and that too mostly in quantitative terms. But in the case of public sector activities, technical robustness is only a part of the design space. There are two other equally important considerations to be kept in mind. They involve political acceptability and administrative feasibility. Would the policy pass the test of being acceptable to the various political constituents whose interests are likely to be affected by it? Would the bureaucratic system have the capacity to deliver effectively the policy being proposed? Complicating matters are the diversity within the political and administrative space, especially so when we are talking about a large country. Importantly, unlike with the private sector, the answers to the two questions mentioned earlier are less technical and less about logical thinking. Instead, they are mostly subjective and practical and involve the exercise of judgement. This element of subjectivity and judgement also means that the likelihood of policy and program design being flawed is also higher. The failure rate with public policy, arising from its very design, is therefore higher. Furthermore, the challenges with policy design are matched by the concerns with its implementation. Here too the contrast with our dominant narrative from the typical private sector activities is stark. In the latter, once a decision is taken, its execution follows a largely predictable trajectory. This is also because the execution can be scripted and is time bound, roles can be prescribed, and outputs can be uniformly defined and monitored. However, with public policy, the implementation is complex and long drawn, involves considerable agent discretion, and is differentially paced, and its outputs are often intangible and diffuse. Added to all this are the problems of implementation in politically challenging and bureaucratically weak environments.

14 Introduction In the circumstances, implementation calls for a different approach: one which elevates organizational attributes like patience, persistence, attention to detail, and openness and iteration in implementation. Besides, it calls for a much higher degree of comfort with messiness and ambiguity and tolerance for the second best. Needless to say, the preceding discussion is based on the typical illustrative examples from public and private sectors and tends to exaggerate the differences to make the point. But the point is valid. This book views policymaking and program implementation from this perspective. The issues discussed and suggestions are made from that perspective. Especially but not only for readers from outside the government, we believe that this perspective can be helpful in appreciating the rationale behind many of our arguments.

1.6  Scope of the Book This book discusses the subject of state capability at a conceptual level, applicable across countries, and then assesses state capability in India. It draws upon theoretical as well as practical insights from around the world. For India, it seeks to identify weaknesses in state capability and their causes and then propose measures to remedy them while recognizing political economy constraints. To enable a clear understanding, it also contains a suitably detailed description of the Indian state wherever relevant. Many of the issues and solutions pertaining to India are also relevant to other developing countries. Issues of state capability can affect all three ‘branches’ of the government—​executive, legislature, and judiciary. However, this book concentrates mainly on the executive branch.2 While the effects of capability deficiencies in the other two branches (particularly the judiciary) on the capability of the executive branch are covered, issues of capability within the judicial and legislative branches—​for example, how to make the courts faster or more efficient or how to increase the quality of parliamentary accountability of the executive—​are beyond the scope of this book. 2 Within the executive branch, coverage of police and prosecution is not as detailed as their importance would warrant, reflecting the authors’ limitations of knowledge.

Introduction  15 Indeed, the term state capability in this book is often used as though it refers to executive capability. This reflects partly the fact that the authors have greater knowledge of the executive state and partly the fact that the executive component of state capability is undoubtedly the most important in terms of development effectiveness. In practice, judicial state capability and legislative state capability too have important effects on the overall capability of a state, albeit to a lesser degree than the executive. The book examines the capability of public systems to design effective policies and implement them, with a particular emphasis on India. Its focus is predominantly on the administrative contributors to state capability deficiencies. It is an account that combines the academic literature and a practitioner’s perspective on the topic. With the authors being two civil servants, it naturally reflects that perspective. There are few works that address this subject, let alone works that attempt a combination of academic rigour and practical insights. The book straddles the fine line between analytical research scholarship and an ethnographic account of actors, processes, and institutions within different levels of government. Being participants and observers in the bureaucratic system, the authors describe reality without always seeking to locate it in the framework of existing academic literature. However, care has been taken to incorporate the available research on bureaucratic systems and especially Indian ones.

1.7  Structure of the Book The book is organized into four parts containing a total of 16 chapters. Part I has two introductory chapters. The second chapter is an introduction to state capability, the distinction between policymaking and its implementation, and its importance in meeting India’s development and economic growth objectives, as well as the role of the civil services and line departments in addressing state capability. Part II examines the institutional design of the Indian state. The third and fourth chapters look at the constitutional provisions and organizational structure, and personnel of the Indian state. The fifth chapter is an assessment of the effectiveness of the Indian civil service.

16 Introduction Part III explores state capability challenges and problems in terms of the size of the state, policymaking, and implementation. The sixth chapter highlights the thin nature of the Indian state and the reasons for it. The seventh chapter looks at the major challenges faced in effective policymaking while the eighth covers the issue of ‘decision paralysis’ which affects the working of the state. The ninth chapter discusses the challenges of execution. Part IV of the book proposes remedies to the problems identified in Part III. The 10th and 11th chapters discuss possible ways of addressing policymaking weaknesses; the 12th chapter offers suggestions to help civil servants become better policymakers. The 13th, 14th, and 15th chapters address execution challenges. The final (16th) chapter looks at some practical ways of initiating systemic improvement and at expectations management.

2 State Capability A Conceptual Overview

2.1  What Is State Capability? State capability is analogous to a hose used to deliver water; if the pipe is of good quality and free of kinks and leaks, the volume pumped through it reaches the delivery point. If the pipe is leaky or has sharp kinks obstructing the flow, only a small part of the water will be delivered; increasing the volume of water pumped may have little effect on the delivered quantity. Similarly, programs designed and implemented by a weak state tend to fall far short of expectations even with enhanced resource allocations. It is an intriguing feature of development discourse this millennium that the entity central to development, the state, has largely been a marginal concern.1 For instance, while the 17 Sustainable Development Goals (SDGs) cover the full spectrum of development objectives, there is no reference to the entity that has the primary responsibility for realization of the majority of those goals. In fact, almost the entire focus of the international aid community has revolved around the provision of physical infrastructure and poverty alleviation interventions. Capability building assistance has been confined to technical assistance to support the implementation of specific aid projects or public financial management (PFM) systems. One could scan the entire portfolio of projects of the multilateral organizations and struggle to find examples of projects which seek primarily to address state capability weakness. (To be fair, this is partly because the charter of international institutions is to work 1 Francis Fukuyama, ‘The Strange Absence of State in Political Science’, The American Interest, 2 October 2012 (https://​www.the-​ameri​can-​inter​est.com/​2012/​10/​02/​the-​stra​nge-​abse​nce-​of-​ the-​state-​in-​politi​cal-​scie​nce/​). State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0002

18 Introduction with their member governments and does not extend to questioning or studying the capability of the government itself, unless explicitly invited to do so. With a long history of colonialism, it is perhaps not surprising that developing country governments have been reluctant to let international institutions, dominated by erstwhile colonial masters, advise on internal governance structures.) However, there is a growing recognition that the Sustainable Development Goal 1 of ending extreme poverty by 2030 cannot happen just by economic growth and private sector development. There have been calls for ‘substantial recalibration of efforts’ to provide the support that ‘the poor need for economic mobility’ by ‘investing in—​not circumventing—​domestic states’.2 Discussions about the state among economists or investors often start with the assumption that it is an institution that hinders growth or uses coercive power. What can be done to get the state out of the market’s way? How can state power be restrained? Or just, how can we ‘shrink the state’? Even when occasionally framed differently, it is confined to maintaining law and order and social stability or concerned mainly with addressing the threat posed by weak or failing states.3 But it was not always so. The developmental state and its capacity were a central feature of the post-​colonial debates on national economic growth. Underlining the importance of state capacity, Samuel Huntington wrote in the late sixties4: The most important political distinction among countries concerns not their form of government but their degree of government. The differences between democracy and dictatorship are less than the differences between those countries whose politics embodies consensus, community, legitimacy, organization, effectiveness, stability, and those countries whose politics is deficient in these qualities . . . The United States, Great Britain, and the Soviet Union have different forms of government,

2 Lucy Page and Rohini Pande, ‘Ending Global Poverty, Why Money Isn’t Enough?’, Journal of Economic Perspectives, 32, no. 4 (Fall 2018): 173–​200 (https://​pubs.aea​web.org/​doi/​pdfp​lus/​ 10.1257/​jep.32.4.173) (Accessed on 12 December 2018). 3 Francis Fukuyama, State Building: Governance and World Order in the 21st Century, Profile Books, 2004. 4 Samuel P. Huntington, Political Order in Changing Societies, Yale University Press, 1968.

State Capability  19 but in all three systems the government governs . . . If the Politburo, the Cabinet, or the President makes a decision, the probability is high that it will be implemented through the government machinery.

In the eighties, works by the likes of Theda Skocpol drew on the framework of Max Weber (the German philosopher and father of bureaucratic theory). She highlighted ‘bureaucratic capacity’ or ‘the mechanisms by which the autonomous power of the state operates’ to realize development goals, including effective intervention to promote economic growth.5 Skocpol defines state capacity as whether a state is able to ‘implement official goals, especially over the actual or potential opposition of powerful social groups or in the face of recalcitrant socioeconomic circumstances’.6 In the 1990s, Charles Tilly highlighted the importance of states’ capacity to effectively raise tax revenues to support the country’s military objectives.7 But even these examinations remained confined to the extractive (ability to raise revenues) and coercive (ability to maintain law and order) powers of the state. In the context of developing countries, Joel Migdal highlights the importance of state autonomy and defines state capacity as ‘the abilities of state leaders to use the agencies of the state to get the people in the society to do what they want them to do’.8 And this in turn depends on the ‘degree of control that state agents exercise over persons, activities, and resources within their government’s territorial jurisdiction’.9 Autonomy becomes important to ensure that the state can pursue its objectives without being a captive of specific interest groups. This also harks back to Weber, for whom the bureaucracy is an instrument to exercise control and domination. This is primarily in the context of the state’s ability to extract resources, regulate social relationships, and penetrate its territories and logistically implement decisions. 5 Luciana Cingolani, The state of state capacity: A review of concepts, evidence, and measures, UNU-​MERIT Working Paper Series, #2013-​053, 10 October 2013. 6 Theda Skocpol, ‘Bringing the State Back in: Strategies of Analysis in Current Research’, in P.B. Evans, D. Rueschemayer, and T. Skocpol (eds.), Bringing the State Back in, Cambridge University Press, 1985. 7 Charles Tilly, The Formation of National States in Europe, Princeton University Press, 1975. 8 Joel Migdal, Strong Societies and Weak States: State-​Society Relations and State Capabilities in the Third World, Princeton University Press, 1988. 9 Doug McAdam, Sidney Tarrow, and Charles Tilly, Dynamics of Contention, Cambridge University Press, 2001.

20 Introduction The success of the East Asian economies sparked interest in the role of the state in the facilitation of structural transformation, industrialization, and economic growth. It spawned a rich body of work that highlighted the important role played by the state in the rapid development of countries like South Korea,10 Japan,11 and Taiwan.12 Adherence to Weberian bureaucratic norms on a ‘Weberianness scale’ (based on adherence to meritocratic recruitment and predictable and long-​term career) was found to closely correlate with state effectiveness.13 Taking this line of professional competence further, Skocpol and Finegold added technical expertise to bureaucratic autonomy as an important requirement for state effectiveness.14 Other studies have explored the role of the strength of state capability on a variety of development and economic growth outcomes.15 They have consistently found strong positive correlations between them. Unlike conventional development outcomes, state capability cannot be seen and thus cannot be directly measured. Therefore assessment of state capability runs the risk of ‘circular explanations and tautological reasoning’.16 Researchers have used proxy measures like tax extraction (or government revenue)17 and productivity (or quality) of public expenditure to assess the strength of state capability.18 Organizations like the World Bank have used composite measures like the Government Effectiveness (GE) within the World Governance Indicators (WGIs) to assess state capability. But there is no consensus about which of these 10 Alice H. Amsden, Asia’s Next Giant, Oxford University Press, 1992. 11 Chalmers Johnson, Japan: Who Governs?: The Rise of the Developmental State, WW Norton and Company, 1996. 12 Robert Wade, Governing the Market: Economic Theory and the Role of Government in East Asian Industrialisation, Princeton University Press, 2004. 13 J.E. Rauch and P.B. Evans, ‘Bureaucratic Structure and Bureaucratic Performance in Less Developed Countries’, Journal of Public Economics, 75, no. 1 (2000): 49–​71. 14 Theda Skocpol and Kenneth Finegold, ‘State Capacity and Economic Intervention in the Early New Deal’, Political Science Quarterly, 97, no. 2 (1982): 255–​278. 15 Luciana Cingolani, The state of state capacity: A review of concepts, evidence, and measures, UNU-​MERIT Working Paper Series, #2013-​053, 10 October 2013. 16 Alexandre A. Gomide et al., The Concept of State Capacity and Its Operationalisation in Empirical Research’, Conference Paper, International Public Policy Association, June 2018. 17 Mattias Ottervik, ‘Conceptualising and Measuring State Capacity’, University of Gothenburg Working Paper Series 2013:20 (https://​www.qog.pol.gu.se/​digita​lAss​ets/​1468/​ 14688​14_​2​013_​20_​o​tter​vik.pdf). 18 M. Dinceco and G. Katz, ‘State Capacity and Long Run Performance’, The Economic Journal, 126, no. 590 (February 2016) 189–​218 (https://​acade​mic.oup.com/​ej/​arti​cle/​126/​590/​189/​ 5077​805).

State Capability  21 measures of state capability best fit the requirement. This dissonance creates its set of challenges in assessing state capability. In more recent years, coming predominantly from the development perspective, Acemoglu et al. argue that ‘societies with limited state capacity are those that invest relatively little in public goods and do not adopt policies that redistribute resources to the poor’. They consider state effectiveness to be related to the abilities of a central authority to monitor bureaucrats.19 Tim Besley and Torsten Persson define four dimensions of state capacity as taxing, contract enforcing (called productive capacity), coercive (peace and order), and regulatory/​distributive (public goods provision).20 In Chapter 1, state capability is defined as the ‘ability of a government to get the right things done’ and a measure of the state’s ability to effectively design, implement, and deliver what is promised to its citizens. This depends on the bureaucracy’s autonomy and its technical expertise. A few examples may help. Take the case of affordable housing in cities. At the heart of the matter is the delivery of the stock of different categories of housing required to meet the pent-​up as well as rapidly increasing demand. It may require many measures—​developing vertically given the acute land scarcity; attracting developers into the affordable housing market; keeping construction costs down, public housing programs, etc. All these need to be brought together as part of a comprehensive package, with the required inter-​ departmental coordination and monitored for effective execution. This demands a state capable of policy formulation and monitoring implementation. Or take the example of the instances of mid-​day meal poisoning. If the headmaster follows the minimum protocols for procurement of items and management of the kitchen and school inspectors do basic supervision, and all these processes are embedded in an institutionalized system of accountability, such incidents could easily be averted. The ability to adhere to these protocols and processes is a measure of the state’s capability

19 D. Acemoglu, D. Ticchi, and A. Vindigni, ‘Emergence and Persistence of Inefficient States’, Journal of European Economic Association, 9, no. 2 (2011): 177–​208. 20 Timothy Besley and Torsten Persson, Pillars of Prosperity, the Political Economics of Development Clusters, Princeton University Press, 2011.

22 Introduction to execute effectively. In weak states, even a mundane task like running a school kitchen, when done at scale, seems fraught with complications. State capability weakness is not unique to India. It is pervasive across most developing countries and is perhaps the most important global development challenge. (As an aside, there are those who would argue that state capability to conduct elections is arguably weaker in some developed democracies than in India!) Lant Pritchett, Matt Andrews, and Michael Woolcock describe many developing countries as being caught in a ‘capability trap’. In a paper, they argue that ‘state capability stagnates, or even deteriorates, over long periods of time despite governments remaining engaged in developmental rhetoric and continuing to receive development resources.’21 They use several indicators to show that in most developing countries, including India, state capability has declined in the past decade. In a now-​famous illustration, researchers examined the differences in the actual performances of post office systems across countries, developing and developed.22 More specifically, they studied how different postal systems responded to undeliverable international mail, which has to be returned to the origin country within 30 days as per the provisions of the Universal Postal Union convention to which most countries are signatories and theoretically adhere. The researchers sought to capture the difference between de jure requirements and de facto performance for each country. So they deliberately sent 10 pieces of mail to non-​existent addresses in each of the 159 members and studied how many got returned and how long it took for them to return. They found, in general, a strong positive correlation between the number and time and national incomes. In the lowest income quartile and the bottom half by schooling, respectively, on average, just 0.92 and 2.2 letters out of 10 were returned.23 21 Lant Pritchett, Matt Andrews, and Michael Woolcock, ‘Escaping Capability Traps Through Problem Driven Iterative Adaptation’, Center for Global Development Working Paper No 299, 22 June 2012, (https://​www.cgdev.org/​publ​icat​ion/​escap​ing-​cap​abil​ity-​traps-​thro​ugh-​prob​lem-​ dri​ven-​iterat​ive-​ada​ptat​ion-​pdia-​work​ing-​paper) (Accessed on 12 December 2018). 22 Alberto Chong, Rafael La Porta, Florencio Lopez-​de-​Silanes, and Andrei Shleifer, ‘Letter Grading Government Efficiency’, Journal of European Economic Association, 12, no. 2 (April 2014): 277–​ 299, (https://​scho​lar.harv​ard.edu/​files/​shlei​fer/​files/​pub​lica​tion​_​032​8201​4_​0.pdf) (Accessed on 12 December 2018). 23 Lant Pritchett, Matt Andrews, and Michael Woolcock, Building State Capability, Oxford University Press, 2017.

State Capability  23 The example above measured the state’s ability to get postal mail delivered and acknowledged. Despite the standardization and apparent simplicity of the processes, the ability of the post office systems in this regard varied widely. The same applies to all activities involving public agencies. And, as discussed in Chapter 1, it covers both policy design and program implementation.

2.2  Policy vs. Execution Opinion makers often carry the impression that an announcement of a policy is the important thing, and once done, outcomes invariably follow. The difficulty with policy execution is much under-​estimated. Take the case of one of the major issues of public debate, privatization of banks. The legal (legislative, regulatory), technical (valuation, asset–​ liability apportionment, equity sale, personnel re-​ deployment), and political economy (political consensus, reasonable returns process transparency) steps required to operationalize privatization are extremely challenging and demand high-​quality engagement. In fact, the struggles of the government with the privatization of Air India are an excellent example of the limits and utility of just a policy mandate. Much the same can be said about many of the major development challenges—​making housing affordable, reducing urban pollution, lowering traffic congestion, improving student learning outcomes, lowering child malnutrition, addressing non-​communicable diseases, promoting small enterprises, and so on. Opinion makers who speak with great confidence about what needs to be done are unlikely to be so confident when asked to get down to details. The European Commission President, Jean Claude Juncker, famously said about the challenge faced by politicians, ‘We all know what to do, we just don’t know how to get re-​elected after we’ve done it24!’ In fact, one could rephrase Juncker from the perspective of bureaucrats, “We all know what we are supposed to do, but struggle to do it” 24 Marco Buti et al., ‘Defying the “Juncker Curse”: Can Reformist Governments Be Re-​ Elected?’, Directorate-​General for Economic and Financial Affairs, European Commission, May 2008, (https://​ec.eur​opa.eu/​econ​omy_​fina​nce/​publi​cati​ons/​pages/​publ​icat​ion_​summ​ary1​2588​ _​en.htm) (Accessed on 10 April 2019).

24 Introduction This raises the distinction between policymaking and its execution.

2.3  The Difference Between Policymaking, Program Formulation, and Program Implementation So how does program implementation differ from policymaking or program formulation? Fundamentally, policymaking and program design have a wider perspective but less detail than program implementation. A very rough example may help. Suppose a state is trying to improve road connectivity within the capital. The policy may be to provide enhanced budgetary support for constructing flyovers and creating a new implementation agency to do the work. This may involve considering alternatives—​ should money be spent on flyovers or on a new public transport system that would reduce car traffic? A program of flyover construction may then be announced. This involves consolidating and analysing data from traffic, topography, economic, and other growth forecasts to decide the locations of the flyovers and finalize the design features and strength of various civil engineering options. Once these variables are fixed, the next stage is executing the construction work based on the design. This is program implementation which throws up its own set of practical problems—​the soil conditions encountered may be different from the plan assumptions, there may be difficulties in acquiring land, courts could grant stay orders, contractors may default on work or become bankrupt, and so on. The span of influence is much smaller with program implementation than with policymaking. District Collectors have smaller jurisdictions and therefore have the mental and physical bandwidth to exercise far greater control than a policymaker can over an entire state or nation. A program may have to accommodate the requirements of widely varying implementation environments of a state or country. This necessitates the incorporation of adequate flexibility to allow context-​specific adaptation. Local initiative is essential to customize the program or policy to suit local requirements. Besides, a policy or program may span many layers of bureaucracy, and even external stakeholders. A strategy that is applicable for

State Capability  25 administration at the lowest bureaucratic tier (district) is not appropriate for administering multiple layers. Implementation, on the other hand, requires much greater attention to detail. Finally, course corrections and abrupt shifts are easier and more likely with program implementation than in program or policy design.

2.4  Flexibility Is Essential—​but May Be Impossible The previous section made the point that while policy is made in the aggregate or at the general level, successful implementation demands attention to the specific context and its particular details. This, in turn, demands that policies themselves be built with adequate flexibility in implementation. In simple terms, since implementation has to be tailored to fit the context, policy has to offer the greatest flexibility consistent with control over the realization of objectives. But while implementation flexibility may be desirable, successful use of flexibility requires state capability. Customization to meet contextual requirements, howsoever marginal, demands significant quality of engagement by officials at the lower levels. It also requires a good monitoring system for officials at various higher levels. If there is a lack of capacity at lower levels and/​or an inability to monitor well, then there is a danger that flexibility may actually produce a result worse than a rigid ‘one size fits all’ policy. This is the Catch-​22 situation that policymakers often face.

2.5  The Overlap Between Policy Design and Execution While making an analytical distinction between policymaking and program implementation, there are practical problems. For a start, design to implementation is (in practice) a continuum, with no distinct separation. Program implementers, whether the District Collector at the top or the Block Development Officer or Block Agriculture Officer at the block-​level or even the Police Constable or

26 Introduction Panchayat Secretary at the frontline, are far from passive recipients with no operational discretion. While their actions will be governed by the program guidelines, the nature of most services delivered by the government confers significant operational autonomy. Consider the example of an agriculture extension officer or an industrial promotion officer. His or her mandate is to support the farmer and the business overcome constraints and enhance productivity and, in the process, also help them access all the relevant government schemes. Even with all the constraints of resources and workload, the operational discretion enjoyed by the officer is enormous. From the manner in which the official engages with the farmer or entrepreneur to their work organization and quality of service delivery, there is a very large scope for innovation and value addition. While not to the same extent, even regulatory officials enjoy significant operational autonomy. Consider the factory inspector or the town planner. While stereotyped as officials with top-​down regulatory mandates, both also have a very important facilitative role in helping the factory or the building owners conform to the extant guidelines. A benign regulatory official is also a facilitator. The officials have a lot of operational freedom in organizing their engagements with their stakeholders. This view of operational autonomy is also borne out by a very rich body of literature. Michael Lipsky wrote four decades back about how street-​level bureaucrats have significant discretion, which leads them to function as also policy decision makers. In particular, he pointed to their autonomy to adopt coping mechanisms like managing client demand either by limiting them or by creaming or cherry-​picking.25 Others have pointed out how, when faced with constraints of resources and workload, street-​level bureaucrats end up reducing demand for output, rationing output, and automating output.26 In a study of street-​level bureaucrats in the United States, Bernardo Zacka points out that far from being mechanistic and soulless

25 Michael Lipsky, Street-​Level Bureaucracy: Dilemmas of the Individual in Public Services, Russel Sage Foundation, 1980. 26 Evert Vedung, ‘Autonomy and Street-​Level Bureaucrats’ Coping Strategies’, Nordic Journal of Studies in Educational Policy, 2015, no. 2 (3 July 2015) (https://​www.tand​fonl​ine.com/​doi/​full/​ 10.3402/​nstep.v1.28643) (Accessed on 7 March 2020).

State Capability  27 operators, these bureaucrats possess a significant margin of discretion.27 He describes how bureaucratic rules and regulations are often vague and without important operational details, leaving implementing officials to form their own interpretations and exercise judgement. Further, beyond the technical side, frontline officials are also faced with moral choices and dilemmas. Public servants are supposed to be efficient while also being fair, responsive, and respectful. This forces trade-​offs across these different dimensions of moral value. Another point for consideration is that policies and programs are not cast in stone. They evolve continuously, and that too based on feedback from implementation. An important point in this regard is that, unlike the policymakers, whose engagement is mainly limited to making the policy, the implementers are constantly engaged with the program. From this engagement emerges innovation and insights, some of which get adopted in refining the design of the program itself. Finally, program implementers also have a direct program design role. The commonest example is those of best practice or positive deviance, which show the way forward for scaling up entirely new programs. It is no surprise that the origins of most, if not all, the current flagship development schemes of the Government of India can be traced back to small experiments in some district or another. In Chapter 11, we discuss the Chinese approach of reform by allowing low-​stakes experimentation, which, if successful, is scaled up. In such cases, as Yuen Yuen Ang has written, policies and programs co-​evolve with the pilot implementation.28 This book touches on some of these points. However, while acknowledging the overlap, it maintains a distinction between policymaking and program implementation. It is a useful analytical distinction to diagnose and propose some solutions to addressing the state capability problems in India.

27 Bernardo Zacka, When the State Meets the Street: Public Service and Moral Agency, Harvard University Press, 2017, (https://​www.hup.harv​ard.edu/​cata​log.php?isbn=​978067​4545​540) (Accessed on 7 March 2020). 28 Yuen Yuen Ang, ‘How China Escaped the Poverty Trap’, Cornell University Press, 6 September, 2016.

28 Introduction

2.6  Addressing State Capability Failures Most of the important public policy challenges—​poor learning outcomes, poor sanitation and hygiene, dysfunctional healthcare, traffic congestion, unaffordable housing, stagnant farm productivity, and so on—​involve integrating interconnected dimensions, building partnerships within and outside government, and empowering stakeholders at different levels. They require diligent and persistent work over a long period of time to make any significant dent in the problem. The standard response to governance failures is to address them through one of four approaches. The commonest refrain, especially among the urban middle class, predominantly employed in the formal private sector, is an increased role for the private sector. This feeling is at least partially fuelled by the representativeness bias arising from contrasting high-​profile examples of ‘successful’ private businesses with pervasive government failures and the internalization of their own workplace norms. And it is reinforced by opinion makers and the mainstream media. (This perception is not always backed up by facts—​the collapse of Jet Airways contradicts the popular narrative of ‘efficient’ private airlines vs an inefficient Air India.) The second common response is to embrace participatory governance, which promotes the empowerment of communities through decentralised decision-​making. Women’s self-​help groups and micro-​loans are the commonest examples of this approach. It is argued that such ‘bottom-​up’ strategies alone can ensure accountability and transparency, critical to the achievement of sustainable and inclusive development. The third approach places faith in innovative approaches to address governance failures. Ideas like conditional cash transfers, unlocking the energy of social enterprises, establishment of a strong anti-​corruption ombudsman, extensive use of the latest in information technology, process re-​engineering within public systems, and applying insights from behavioural psychology are paths to improving governance and eliminating poverty. The belief is that we can leap-​frog deep-​rooted structural deficiencies by these innovations. Finally, influential academics believe that programme design and evaluation, using rigorous field experiments, can be an antidote to governance

State Capability  29 failures. Accordingly, they advocate an evidence-​based, incentive-​compatible design of policies and their evaluation. All the aforementioned approaches are valuable and important. But they are not substitutes for state capability. Provision of basic public goods, an effective welfare state, delivery of statutory services, and the protection of basic civil and property rights are the responsibilities of governments everywhere and will continue to be so. The private sector can, at best, be a marginal contributor to alleviating failures in these. The limitation of the bottom-​up community-​driven approach, as economist Lant Pritchett provocatively argues, is that it is ‘all bottoms and no ups’—​more effective as social empowerment than as an economic growth strategy. The attractions of innovations lie in their deceptive ‘quick-​fix’ appeal. But many complex socio-​economic and political problems are not amenable to such ‘magic-​pill’ solutions. In many cases, there is more value in just executing existing ideas and innovations. Academic researchers often construct interventions on randomly selected ideas using theoretical insights and, in the process, may miss the broader perspective and overlook implementational challenges. The short point is just that no extent of private participation or decentralisation or innovation or alignment of incentives can make up for basic administrative deficiencies. While action is required on all these dimensions, they are not substitutes to improving state capability. In fact, the absence of institutional capability can be an insurmountable obstacle to any of these strategies, just as how a capable state can amplify their effectiveness. This book discusses several of the commonly proposed solutions to improving policy design and program implementation. It also suggests other—​less common—​solutions.

PART II

THE IN ST IT U T IONA L DE SIG N OF T HE IN DIA N STAT E

3 Constitutional Provisions and Structure of Government The constitutional position of the administrative and regulatory state draws partly from the Constitution itself (as amended) but even more from judicial decisions, especially in the last 25 years when judicial decisions have been far-​reaching and, in some cases, gone to the extent of overturning even express provisions of the Constitution. This chapter provides a very brief introduction. Interested readers can refer to standard works on the subject for more detail.1

3.1  Constitutional Provisions on the Executive Articles2 53 and 154 provide that executive authority can be exercised by officers subordinate to the President (in the case of the Union) and the Governor (in the case of states). These articles also allow Parliament/​State Legislature to confer functions on authorities other than the President/​ Governor and are thus the basic source allowing the creation of independent regulatory agencies. Articles 77/​ 166 provide that the President/​ Governor shall ‘make rules for the more convenient transaction of business’ and for allocation of work among Ministers. The rules framed under this Article are called the ‘Business Rules’ or ‘Rules of Business’. These rules provide for the allocation of work of government between different ministries

1 See for instance, S. Choudhry, M. Khosla, P.B. Mehta (eds.), The Oxford Handbook of the Indian Constitution, Oxford University Press, 2016. This chapter draws from the chapter therein on the Administrative & Regulatory State by T.V. Somanathan. 2 All references to ‘Articles’ are to Articles of the Constitution of India. When a figure in parentheses follows the numbered Article, it denotes a clause of that Article, e.g. Article 77 (3) means clause 3 of Article 77. State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0003

34  The Institutional Design of the Indian State and departments and they also enable delegation of authority from the Council of Ministers (cabinet) to individual Ministers and to officers subordinate to those Ministers. They specify which kinds of decisions need to come before the cabinet and the procedure in the event of inter-​departmental disagreement. Therefore, while all executive action is taken in the name of the President/​Governor, actual authority to take those decisions may vest in lower levels of the executive. Under Articles 73 and 162 the executive power of the Union and state governments, respectively, are co-​extensive with their legislative powers. Article 248 allocates legislative power on all residuary matters (i.e. subjects not listed in the State or Concurrent List) to Parliament, and hence executive power thereon belongs to the Union Government. The Constitution does not explicitly mention or require ‘separation of powers’. This is particularly true of the separation between the Executive and Legislature, given that it is a parliamentary form of government. On the contrary, it has chapters on the ‘Legislative Powers of the President /​Governor’, which confers the power to promulgate ordinances on the executive. It also implicitly allowed the executive to exercise, to a limited extent, judicial powers insofar as the (subordinate) judicial services were part of the public services of the state; the separation of the subordinate judiciary was a ‘Directive Principle of State Policy’ and thus non-​justiciable. However, the separation of powers between the judiciary and executive has been held by the Supreme Court to be part of the (unamendable) ‘basic structure’ of the Constitution. (To be more precise, the courts have been vigilant in ensuring that the executive does not encroach on the judicial space; the same is not necessarily the case when it comes to judicial interventions in the executive space.) Executive power is the residual of all powers that are not legislative or judicial (and thus inclusive of quasi-​judicial powers). Mukherjea, C.J., stated in Ram Jawaya Kapur v. State of Punjab3 that: It may not be possible to frame an exhaustive definition of what executive function means and implies. Ordinarily the executive function

3 AIR 1955 SC 549.

Constitutional Provisions  35 connotes the residue of governmental functions that remain after legislative and judicial functions are taken away.

A feature of the modern-​day state is the presence of regulatory agencies outside the traditional politically controlled ‘executive branch’, which are discussed later in this chapter.

3.2  Ministerial Responsibility Under Articles 75 (3)/​164 (2), Ministers of the Union/​state government are collectively responsible to Parliament/​Legislature for actions of the executive. The effect of Article 75(3)/​164(2) read with Articles 77 (2)/​166 (3) is that individual Ministers are personally accountable to Parliament/​Legislature for actions taken within their areas of responsibility, whether those actions were taken by them or by their subordinates. The conduct of ministries and departments is often the subject of comment and criticism in Parliament/​Legislature. The Supreme Court has stated that The Cabinet is responsible to the legislature for every action in any of the ministries . . . Similarly an individual minister is responsible to the legislature for every action taken or omitted to be taken in his ministry.4

The constitutional requirement of ministerial responsibility has been the basis for the control of the administrative Ministry over public sector corporations and undertakings. It has also been the source of a provision (in the special Acts creating regulatory authorities, public corporations, and boards—​including the Reserve Bank—​and in the articles of association of ‘Government companies’5) enabling the government to issue directions to such organizations. The principle is that since the Minister (and thus the Ministry) remains accountable to the legislature, a commensurate degree of authority is necessary. No doubt, Ministers and 4 A. Sanjeevi vs. State of Madras, AIR 1970 SC 1102: (1970) 1 SCC 443. 5 Broadly, companies where the central and/​or state government(s) hold a majority of the shares or otherwise control the composition of the Board of Directors.

36  The Institutional Design of the Indian State officers in Ministries are perceived to have used this authority (without necessarily invoking the formal power to issue directions) to exercise direct control over public undertakings to the detriment of efficiency and public good. It is the principle of Ministerial responsibility that also underpins the relationship between regulatory bodies and the Ministry in charge of them.

3.3  Organizational Structure—​Who Rules India? The Government of India (‘the Union Government’ or the ‘Central Government’) consists of Ministries and Departments. A Ministry may consist of one or more Departments. Ministries are headed by Cabinet Ministers who may be assisted by Ministers of States, Deputy Ministers, and Parliamentary Secretaries, all of whom are part of the political executive and have to be members of Parliament. Some Ministers of State may have ‘independent charge’, meaning that they have independent executive authority over a department or Ministry without reporting to a Cabinet Minister, even though their status is below that of a Cabinet Minister. Below the political executive, each Department is headed by a Secretary (equivalent to ‘Permanent Secretary’ in many Commonwealth countries) who is typically a career civil servant, most often from the Indian Administrative Service (IAS). Coordination between Ministries is the role of the Cabinet Secretary, who is the Secretary to the Union Cabinet and head of the civil service, and his office (known as the Cabinet Secretariat). There is also a Prime Minister’s Office (PMO) that assists the Prime Minister and which has a combination of civil servants and political appointees (who may be retired civil servants). Administrative authority in each Ministry rests with the Cabinet Minister (or, where relevant, the Minister of State with independent charge), though it may in practice be based on the advice of the civil service and may also be delegated to subordinate levels.

Constitutional Provisions  37

As of May 2019, the Government of India had 51 Ministries and 58 Departments. There were 24 Cabinet Ministers and 33 Ministers of State.6 The government can by executive order vary the number and sectoral jurisdiction of ministries and departments. India has 28 state governments and 9 governments of Union Territories (one of them being the National Capital Territory of Delhi). State governments have a similar structure of Cabinet Ministers, though it is 6 Government of India Web Directory (http://​goidi​rect​ory.nic.in/​union​_​cat​egor​ies.php?ct=​ E002) (Accessed on 14 November 2019). In recent years, Deputy Ministers and Parliamentary Secretaries have not been appointed in the Union Government.

38  The Institutional Design of the Indian State less common for them to have Ministers of State, Deputy Ministers, or Parliamentary Secretaries. State governments have multiple Departments (not called Ministries) and each Department is typically headed by a Minister at the political level with a Secretary who is normally a career civil servant (most often from the IAS). The structure and number of departments in the state governments vary from state to state and are not necessarily aligned with the structure of Ministries and Departments at the central level. For example, in May 2014 the Government of Tamil Nadu had 36 departments while the Government of West Bengal had 61 departments. Union Territories vary in size and structure but typically have a similar structure to state governments, albeit with fewer departments. The state-​level equivalent of the Cabinet Secretary is the Chief Secretary, who is the Secretary to the State Cabinet and the head of the civil service. There is usually also a Chief Minister’s Office, which may have both civil servants and political appointees, though typically far smaller in size than the PMO. The third tier of the administration in India, forming part of the state government, is the district, almost always headed by an IAS officer known variously as Collector, District Magistrate, or Deputy Commissioner. The latest figures put the number of districts at 727.7 In some states, groups of districts constitute zones or ‘commissionerates’ headed by a senior IAS officer who is known as Commissioner. Below the District Collector are typically Sub Divisional officers heading a collection of ‘taluks’ or ‘tehsils’, each of which is headed by a Tehsildar. Below the Tehsildar are Revenue Inspectors for groups of villages, and finally Village Officers for individual villages. The Seventh Schedule of the Constitution provides for division of subjects into Union, State, and Concurrent Lists in terms of legislative and executive jurisdiction. However, in the case of the subjects in the Concurrent List, the law passed by the Parliament shall ordinarily prevail over the laws passed by individual state legislatures to the extent of any repugnancy. Apart from, and smaller than, districts are urban and rural local bodies. There are 4041 statutory towns8 as per the 2011 census. The municipalities 7 GOI Web Directory, Districts (Retrieved on 10 December 2019) (http://​www.goidi​rect​ory. gov.in/​distr​ict.php). 8 Census of India 2011, Urban agglomeration and cities (http://​cens​usin​dia.gov.in/​2011-​ prov-​resu​lts/​pap​er2/​dat​a_​fi​les/​Ind​ia2/​1.%20D​ata%20Hi​ghli​ght.pdf) (Accessed on 11 November 2019).

Constitutional Provisions  39 (called Corporations in respect of big cities) typically have a Mayor and a Council consisting of directly elected ward councillors. The administrative responsibilities at the municipal level are with the Municipal Commissioner, who is appointed by the state government for tenure, and often belongs to the IAS in case of the largest cities in the state. Similarly, the rural self-​government typically consists of elected Zilla Parishads and Gram Panchayats. The Gram Panchayats have a Sarpanch with a Council of directly elected Panchayat members. The executive responsibilities are with the Village Secretary, who is appointed from the District level for a tenure. In both cases, there is a delegation of powers between the Council and the executive, though the administrative matters generally vest with the executive. The 73rd and 74th amendments to the Constitution vest municipalities and panchayats with sufficient executive powers to function as effective democratic units of local self-​government. The Indian Judiciary is largely a unitary system. The lower courts are tightly administratively and functionally subordinate to the respective High Courts, and they in turn are subordinate to the Supreme Court. The personnel management decisions, including appointments and transfers, too follow this subordination. In this sense, the Judiciary is considerably more centralized than the other two arms of the government. The central and state executive are autonomous in their respective spheres, inasmuch as the Centre cannot direct the state on what to do within its constitutionally demarcated space in the State List of the Constitution and vice versa. Both exercise powers on matters within the Concurrent List. Thus the Chief Minister and the State Cabinet and the Prime Minister and the Union Cabinet are supreme in their respective spheres. The bureaucracies in their respective spheres are under their exclusive control. At the local government level, while there are elected representatives, they do not have the power to appoint their own executive officers (for example, Municipal Commissioner and Panchayat Secretary) who are appointed by the state government. The state legislatures and Union Parliament are supreme in their respective spheres. The executive, by way of possessing the majority and therefore enjoying the confidence of the Legislature or Parliament, exercises in practice considerable control over their law-​making and oversight powers. The local government entities, municipalities and panchayats, do

40  The Institutional Design of the Indian State Table 3.1  Who Rules India?

Central government

State government

Local government

Executive

Legislature

Judiciary

Prime Minister and Union Cabinet (Supreme on matters in the Union List; also can prevail on matters in the Concurrent List. Appoints its own officers)

Lok Sabha and Rajya Sabha (Supreme on Union List; prevails on Concurrent List)

Chief Justice and Justices of Supreme Court

Chief Minister and Legislative Cabinet Assembly (and (Autonomous in its in some states functional domain, i.e. Legislative Council) State List subjects, and in appointment of its own (Autonomous in its officers. But some posts functional domain, are reserved for All India i.e. State List Service officers) subjects) District Collector/​ elected local bodies (Key officers appointed by state government)

(No legislative powers)

Chief Justice and Justices of High Court

District & Sessions Judges and lower levels

not possess any legislative powers. In these cases, as mentioned earlier, the appointed executive exercises the administrative powers. The structure of government is summarized in Table 3.1. Levels separated by lines indicate that lower tiers are independent of the upper tier; where there is no line, it indicates subordination of the lower tier to the upper tier. In terms of overlap of responsibilities in each tier, the judiciary is perhaps the most distinct and unitary in its features. Its administration, including that of court staff at all levels, is completely unitary. In terms of appointments of judges to the higher courts, the role of the executive is limited. However, though the salaries and pensions of Supreme Court and High Court judges are charged to the Consolidated Fund of India, the administrative budgets of all courts are part of the Executive Budget.9 9 National Commission to Review the Working of the Constitution, Financial Autonomy of the Indian Judiciary, 26 September 2001 (http://​legal​affa​irs.gov.in/​sites/​defa​ult/​files/​Financ​ ial%20A​uton​omy%20of%20the%20Ind​ian%20Ju​dici​ary.pdf) (Accessed on 11 November 2019).

Constitutional Provisions  41

3.4  Regulatory Bodies The earlier sections of this chapter were focused on the ‘administrative’ state. This section examines the regulatory bodies that operate with some degree of independence from the conventional (politically controlled) administrative state. The term ‘regulatory state’ is often used to denote institutions operating at arm’s length from the government, insulated from day-​to-​day political pressures and using technical expertise in reaching decisions.10 However, viewed in terms of the economic functions performed, the ‘regulatory’ state does not necessarily mean independent regulators. Many aspects of financial, industrial, environmental, and labour regulation are implemented in India and elsewhere (including developed countries) through ministries and departments of the executive without an independent regulator; indeed, independent regulators are not necessarily or always ‘better’ than traditional administrative structures. Thus, the ‘administrative’ state is much wider than the ‘regulatory’ state. Many functions and sectors now falling under ‘independent’ agencies outside the ‘traditional’ politically controlled executive branch remain executive/​ administrative functions in constitutional/​ legal terms and were earlier part of the traditional executive. The same functions in other sectors continue to be part of the ‘traditional’ administrative state. In recent years, as the role of markets has grown and the scope of government activities expanded, several new regulatory institutions have been created. The degree of independence of these institutions varies. While many of the regulatory bodies are outside the traditional executive, there are now a number of Tribunals that perform judicial functions outside the traditional judiciary. Several of these are specifically linked to regulatory bodies, while some hear disputes relating to other administrative matters. Some of them handle both judicial and quasi-​judicial functions.11

10 N.K. Dubash and B. Morgan, ‘Understanding the Regulatory State of the South’, Regulation & Governance, 6 (2012): 262–​281. 11 This chapter draws heavily on the ‘The Regulatory State’ chapter by T.V. Somanathan in the Oxford Handbook of the Indian Constitution.

42  The Institutional Design of the Indian State

3.4.1  Constitutional Position of Regulatory Bodies The various judgements of the Supreme Court contain a number of inconsistencies, but the following are the key principles that appear to govern the constitutionality and legality of regulatory bodies: i) To the extent that regulation comprises executive functions, Parliament and the State legislature may (in their respective areas of legislative competence) create independent regulators. All administrative decisions, and thus all decisions of regulators, are subject to judicial review in the High Court and Supreme Court through their writ jurisdiction provided for in the Constitution. Parliament also has the power to create tribunals under its residuary powers. ii) If a regulatory body or tribunal (set up outside the traditional executive or traditional judiciary, respectively) involves the performance of judicial functions, it cannot be comprised entirely of, or headed by, non-​Judges. In hearing matters having a judicial aspect, the bench must include a judicial member. iii) Whether a function is judicial is determined by the specific circumstances, wording, and substance of the legislation. iv) It should be noted that the reference above is to judicial and not to quasi-​judicial functions. Many quasi-​judicial functions are in fact performed by the executive; not all adjudicatory functions are judicial and not all tribunals are judicial tribunals. No clear tests have been laid down as to the borderline between quasi-​judicial and judicial but a function previously discharged by a High Court will almost certainly be considered judicial. v) Even if a statute explicitly provides for an appeal from the orders of a regulator to a tribunal and then directly to the Supreme Court, the High Courts will assert, and the Supreme Court will almost certainly uphold, their (High Courts’) writ jurisdiction as being part of the basic structure of the Constitution. Thus review of decisions of a tribunal (effectively, an appellate jurisdiction though not termed as such) by High Courts will always remain available to aggrieved parties even if the tribunal itself is headed by High Court/​ Supreme Court-​level judges.

Constitutional Provisions  43

3.4.2  The Need for Regulatory Bodies: A Conceptual Overview It is useful to briefly look at the conceptual underpinnings of, and justification for, the regulatory state. The economic arrangements of modern times often require a more complex approach than the state was traditionally capable of. In particular, good decisions on many modern regulatory issues may require a degree of scientific, technical, or economic expertise that may not be found in the generalist civil servants and generalist judges who typify the traditional executive and judiciary, respectively. The case for the involvement of sectoral experts in regulation is thus self-​evident. Of course, sectoral expertise can be brought into the administration by employing or consulting experts without necessarily changing institutional structure—​in a sense this objective can be achieved by ‘modernising’ the administrative state. By itself, the need for expertise does not explain the creation of the regulatory state. The primary rationale for independent regulators in India was and still is the prevalent opinion that some kinds of economic decisions need to be insulated from the political process.12 This is based on the notion—​particularly among economists—​that ‘economic’ decisions should be made ‘rationally’ without being ‘distorted’ by political considerations. Indeed, the initial push for independent regulators came in India through transplantation of Anglo-​American models by lending agencies like the World Bank and then was replicated through ‘copying’.13 In theory, such unelected bodies would play a major role in achieving economic efficiency objectives, with the political process then stepping in through separate income re-​distribution to tackle any normative inequities in the outcomes.14 There is ample evidence in many sectors that even though optimal (or, at least, more efficient) solutions are known, those solutions are not implemented for political reasons.15 12 K. Hochstetler, ‘Civil Society and the Regulatory South: A Commentary’, Regulation & Governance, 6 (2012): 362–​370 (1/​9). 13 N.K. Dubash and B. Morgan, ‘Understanding the Regulatory State of the South’, Regulation & Governance, 6 (2012): 261–​281. (4/​21) 14 Ibid. (5/​21) 15 O.P. Agarwal and T.V. Somanathan, Public Policy-​Making in India: Issues and Remedies, Centre for Policy Research Working Paper, August 2005.

44  The Institutional Design of the Indian State An additional impetus for the creation of regulatory bodies has been the higher degree of operational flexibility they have in personnel and other routine managerial matters vis-​à-​vis ministries; thus independent regulators may internally be more efficient than ministerial regulators.16 However, improving the overall level of economic welfare does not automatically improve the welfare of all participants and a solution that makes some people better off often makes others worse off. Questions of distribution of gains and losses essentially are political. Resolving them may require negotiation and persuasion and the trading of gains and losses rather than legalistic adjudication or scientific determination. Thus it is not really possible to ‘exclude’ politics as many economists and technocrats might wish to. As Dubash put it, much of the clamour for independent regulation has been ‘based on the somewhat questionable premise that it is possible to create an apolitical regulatory sphere simply by legislating one’.17 A consequence of the push towards independent regulation is not only that the political aspect is, so to say, brushed under the carpet but also that it is implicitly regarded as illegitimate. On the other hand, it should be noted that regulatory bodies are often required to follow procedures allowing for the consultation of the public, or more particularly, those affected by regulation; this can result in a better articulation of public reason and a better-​informed process of regulation. In addition to a rational-​decision-​taking dimension and a political-​ sharing-​of-​gains-​and-​losses dimension, regulatory decisions often affect the legal rights and responsibilities of citizens, going beyond voluntary agreements by contract. In essence, therefore, regulation involves three aspects: • a political aspect, relating to trading off benefits and costs to different parties that in turn may require or benefit from consultation, negotiation, collaboration, and cooperation with and among them; • a technical aspect, relating to finding the best technical and economic solutions to a problem;

16 K.P. Krishnan, Remarks at Conference on the Indian Constitution, 18 July 2014. 17 N.K. Dubash, ‘Independent Regulatory Agencies: A Theoretical Review with Reference to Electricity and Water’, Economic & Political Weekly, 43, no. 40 (4–​10 October 2008): 43–​54.

Constitutional Provisions  45 • a legal aspect, relating to protecting legal and constitutional rights of various parties. For a regulatory structure to earn the confidence of the regulated and of the broader citizenry, it needs to have democratic legitimacy, substantive competence, and legal legitimacy, respectively. The essential design issue in regulatory structure is the relative balance between these. A consensus involving political, technical, and legal interests is necessary for any regulation to be both welfare-​enhancing and sustainable.

3.4.3  Regulation Through the Traditional State Regulation through the traditional state apparatus gives primacy to the political aspect since decisions are ultimately taken by, or under the authority of, the democratically elected political executive with advice from the civil service. Over the years, the relative importance of the political executive vis-​à-​vis the civil service has increased and indeed political interference (politicians taking or influencing decisions that by law and/​ or procedure ought to be taken neutrally by civil servants) is perceived to have increased. In India, such decisions are always subject to judicial review and this provides legal protection. However, the traditional state is weak in factoring in technical expertise. As already pointed out, there is no intrinsic reason why this is so, and in theory, a modernized public administration would be able to bring in expert inputs. However, current civil service rules and pay structures in India may make this more difficult to do within the administration than under an independent regulator.

3.4.4  Independent Regulators When independent regulators are created, the technical aspect is accorded primacy in two ways. First, these structures explicitly provide for and incorporate the dimension of sectoral expertise. Second, and more importantly, decision-​taking is made independently of the political executive and therefore is more likely to be economically efficient. In India, legal protection is also safeguarded either through the inherent power of

46  The Institutional Design of the Indian State judicial review or through the creation of an appellate tribunal. By design, the political executive—​depending on the degree of independence of the regulator—​has only limited influence.

3.4.5  Increasing Influence of the Legal Aspect Rose-​Ackerman points out that court-​like procedures are very good at protecting individual rights but are poorly designed for resolving policy issues. She adds that the ‘bureaucracy is best placed to balance conflicting interests’, though ‘not to preserve rights or discover scientific truths’. In most cases, the task is to strike a balance between the obligation to make technically competent choices and the need to respond to concerns of citizens and organized groups. Politically expedient choices are not per se illegitimate but should be described as such rather than masked as ‘scientific’ or ‘legal’ necessities.18 Parker and Braithwaite point out that regulation in modern times often requires ‘experimentalism’, whereby new arrangements are tried as needs and technologies change, and this kind of experimentalism is best done through participatory or democratic processes rather than a legalistic process. The central task of the new regulatory state is to ‘connect private capacity and practice of pluralized regulation to public dialogue and justice’. Yet, ‘courts mostly reject experimentalism as threats to consistency of justice and scotch most kinds of collaboration as threats to the independence of the judiciary.’19 For these reasons in the United States and most western countries, the involvement of the courts in the regulatory process is broad but shallow and courts will usually not go into the details de novo.20 The position

18 S. Rose-​Ackerman, ‘Law and regulation’, in Gregory A. Caldeira, R. Daniel Kelemen, and Keith E. Whittington (eds.), Oxford Handbook of Law & Politics, Oxford University Press, August 2008. 19 C. Parker and J. Braithwaite, ‘Regulation’, in Mark Tushnet and Peter Crane (eds.), Oxford Handbook of Legal Studies, Oxford University Press, January 2005. 20 A.K. Thiruvengadam and P. Joshi, ‘Judiciaries as Crucial Actors in Southern Regulatory Systems: A Case Study of Telecom Regulation’, Regulation and Governance, 6 (2012): 327–​343, 2/​17.

Constitutional Provisions  47 in India is different, especially but not only in environmental matters. Indian courts’ involvement is broad and deep. In India, over the years, and through judicial pronouncements, the legal aspect of regulation has come to acquire a very large influence over the regulatory process, even overshadowing the technical aspect.

3.4.6  Trade-​Off Between Consultative Processes and Conventional Political Representation Independent regulators—​while free from direct political control—​are often required to follow consultative procedures before making regulatory decisions. Depending on the specific procedures, this allows all affected parties to make their views known before decisions are taken. In one sense, this process allows for a deeper and higher quality participation in decision-​making than the processes of the administration, where the political representatives (Ministers, MPs, MLAs, or local councillors) are expected to reflect the views of constituents and where structured consultation may not take place. When this happens, independent regulators may be able to achieve the best of both worlds—​better substantive decisions (through the use of technical expertise and rational decision making) but informed by a deeper participatory process than the administration. If so, their decisions may not only be substantively better but also more ‘legitimate’ and thus have greater public support.21 This is the optimistic scenario for independent regulation. However, consultative procedures have disadvantages too. They tend to be used more by parties who are willing to invest time and resources in responding to the consultation. The logic of collective action indicates that the general public—​where a large number of people may face a small gain or loss from a decision—​may not participate effectively, while a small number of people (regulated entities or well-​organized interest



21 N.K. Dubash, op. cit., p. 51.

48  The Institutional Design of the Indian State groups) who have a lot to gain or lose may invest time and effort to make their case to the regulator. This can tilt regulatory decisions in favour of well-​organized interest groups. Hochstetler argues that civil society in developing countries is more likely to be skewed in favour of the interests of relatively well-​off groups. In contrast to such processes, Rose-​ Ackerman points out, the elected political executive—​because of its need to get re-​elected from the population as a whole—​acts as a check against the dominance of particular narrow interest groups. Thus, as Levy and Spiller point out, independent regulators have both strengths and weaknesses vis-​à-​vis the regulation by (or operation of economic entities by) the administrative state, and independent regulators are only appropriate in some contexts.

3.4.7  Advantages and Disadvantages of Independent Regulators In the Indian constitutional context, the pros and cons of independent vs. administrative regulators on the political, technical, and legal dimensions are summarized in Table 3.2. Given that India is a country where political corruption is widespread, it is sometimes thought that an advantage of independent regulators and judicial tribunals would be reduced scope for corruption. However, such a conclusion is difficult to support in the face of the fact that not only the civil services but even the judiciary are perceived to also suffer from corruption.22

22 See, for instance: http://​arch​ive.indian​expr​ess.com/​news/​rs-​impeac​hes-​just​ice-​soumi​ tro-​sen-​for-​misap​prop​riat​ing-​funds/​833​874/​; http://​arch​ive.indian​expr​ess.com/​news/​sc-​dire​ cts-​cen​tre-​to-​probe-​graft-​char​ges-​agai​nst-​excji-​balak​rish​nan/​947​658/​0; http://​times​ofin​dia. ind​iati​mes.com/​india/​Eight-​chief-​justi​ces-​were-​corr​upt-​Ex-​law-​minis​ter/​arti​cles​how/​6568​ 723.cms.

Table 3.2  Advantages and Disadvantages of Independent vs. Administrative Regulators Criterion Political aspect

Technical aspect

Legal aspect

Independent regulator

Administrative regulator

Accountability to Low—​no direct Parliament and accountability but general public government may have a formal power to give directions. Quality of High if required under consultation statute. Ability to reach Medium to low negotiated depending on compromises flexibility/​rigidity of among interest policy framework and groups procedures prescribed in the legislation. Extent of compromise and innovation depends on extent of procedural restrictions. Failure to follow prescribed approach can lead to decisions being challenged. Access to High—​independent expertise status and different service conditions enable easier recruitment of technical talent.

High—​directly accountable to Parliament/​legislature for policies and decisions. Usually low though not necessarily so. High since the process is explicitly political rather than ‘rational’.

Ability to take decisions in an economically rational manner

Low: Short term political needs acquire much greater salience.

Protection of legal rights

Speed with which legal finality is reached

High/​medium: High in theory, usually medium in practice, as seen from studies of actual behaviour. High; either through High Court/​Supreme Court or through tribunal/​High Court/​ Supreme Court. Can be lower if special tribunal is involved due to increased levels (and sometimes channels) of appeal.

Medium—​theoretically, can access in-​house expertise in many fields and outside experts through consultation. In practice, bureaucratic process and tendency for generalist domination may reduce it.

High; through High Court/​Supreme Court.

Usually higher than for independent regulator as no tribunal is usually involved.

50  The Institutional Design of the Indian State

3.4.8  Regulatory Institutions: An Economic Analysis In terms of an economic framework of analysis, Dani Rodrik and Arvind Subramanian describe four types of institutions—​regulating, stabilizing, legitimizing, and promoting—​that improve the effectiveness of markets.23 Efficient and fair markets are characterized by effective regulation. 23 Dani Rodrik and Arvind Subramanian, ‘The Primacy of Institutions’, IMF Finance & Development, June 2003 (https://​www.imf.org/​exter​nal/​pubs/​ft/​fandd/​2003/​06/​pdf/​rod​rik.pdf) (Accessed on 28 December 2018).

Constitutional Provisions  51 The literature sometimes betrays a confusion between core government functions including provision of public goods and services (schools, hospitals, etc.) and the concept of ‘regulating’ provision of goods and services by the private sector. For the purpose of this book, regulatory bodies are those in the latter category. The market failures that demand regulation commonly arise either due to information asymmetry (between the sellers and the buyers, which causes problems like adverse selection and moral hazard), monopoly characteristics (as with railway or electricity distribution which promotes anti-​competitive practices), externalities (like pollution whose supply can be unlimited), high transaction costs (in settling contractual disputes whose resolution requires institutional mechanisms), and problems of common property resources (like allocation of mines or petroleum blocks). In terms of the nature of failure mitigating actions, regulatory objectives can be broadly classified into consumer protection, limiting state discretion, ensuring level-​playing field, sustainable development, and safeguarding public interest. The regulatory mandates, in turn, can be classified into four functional categories: • accreditation (mandating of entry-​level standards); • standards certification (quality of goods and services being delivered); • price setting (tariff determination in case of monopoly and public goods); and • dispute resolution. While the first is an entry-​level function, the other three are post-​entry functions. These functional categories can be thought of as the spectrum of regulatory choices available to mitigate the market failures described earlier. The effectiveness of these four regulatory interventions is critically dependent on the effectiveness of its monitoring for service quality or contract compliance. A classification of the various regulatory institutions based on the type of market failure and the mitigating regulatory mandates reveals a useful framework for analysis. Information asymmetry engenders various

52  The Institutional Design of the Indian State Table 3.3  Framework of Analysis (Theoretical Principles) Accreditation Information asymmetry Common pool resources Externalities Monopoly characteristics Transaction costs

Standards Price Dispute certification determination resolution

✓

✓

x

x

✓

x

x

x

✓ ✓

x x

✓ ✓

x x

x

x

x

✓

incentive distortions, which are best mitigated by accreditation norms, standards certification, and service monitoring. Again, apart from constant monitoring, problems arising from common-​pool resources and monopoly practices are most effectively addressed by regulating entry and price. Negative externalities are primarily controlled by rigorous entry norms, strict operational standards, internalization of external costs through the pricing mechanism, and continuous monitoring. Finally, non-​judicial dispute settlement mechanisms lower transaction costs and ease contracting. From these, we derive the framework for analysis given in Table 3.3.

4 Personnel of the Indian State The central government alone employs over three million civilians, as can be seen from Table 4.1. In addition, there are the employees of the state and local governments and the various public sector authorities and corporations. Within this number, it is useful to look separately at the ‘higher’ civil services and at the intermediate and lower services.

4.1  The Higher Civil Services The ‘higher’ civil services in India, also known as the ‘Class I’ or ‘Group A’ services, are of three main types. The members of All India Services (AIS) serve both Union and state governments. The Union government is also known as the central government or the Government of India (GoI). Central Services are those whose members work under the Union government only. State civil services are the higher services of state governments. The Indian Administrative Service (IAS) and the Indian Foreign Service (IFS), a central service, are considered the most prestigious as judged by preferences of examination candidates and general public perceptions. Their pre-​eminence is derived partly from a slight edge in pay scales and slightly faster promotions, but more from their job content—​the glamour and travel opportunities of the foreign service and the breathtaking variety of job content of the administrative service. The IAS, the Indian Police Service (IPS), and the Indian Forest Service (IFS) are AIS, most of whose members are recruited through rigorous competitive examinations by the independent Union Public Service Commission (UPSC). A minority of them are promoted from state civil services. Members of these services work in both the Union and state governments. The central government and the state governments also State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0004

54  The Institutional Design of the Indian State Table 4.1  Number of Civilian Employees in Position as on 1 April 2014 Department Railways MHA incl. CAPFs Defence (Civilian) Posts Revenue Others Total

Group A 8,493 16,090 17,160 527 5,827 43,404 91,501

Group B 7,283 60,162 59,415 6,826 40,907 106,299 280,892

Group C 1,299,912 903,862 321,847 182,418 49,078 172,026 2,929,143

Total 1,315,688 980,114* 398,422 189,771 95,812 321,729 3,301,536

Note: This includes 140,294 employees of Union Territories (including Delhi Police personnel). Source: Report of 7th Central Pay Commission, New Delhi: Government of India, 2015.

have their own civil services. The Secretary is the administrative head of each department who exercises supervision over all staff and financial resources. The IAS is a ‘mandarin’-​type career civil service, comparable very broadly to the examination-​recruited higher services in Britain, France, or Japan. This is contrasted with ‘position-​based services’, where persons are recruited to a specific post.1 Where it differs from many other mandarin services is in the vertical and lateral mobility of IAS officers. Not only do they work at several different levels of government—​Union, state, and local and with public sector corporations—​but they also move across ministries and functions to a much larger extent than in most other mandarin civil service systems. In this latter respect, the IAS is closer to the career component of the United States Senior Executive Service (SES).2 IAS officers do, however, specialize by geographical area—​a necessity in a country with such diversity of language, culture, religion, and custom. Recruitment is primarily through a very stiff competitive examination at a young age, typical of mandarin-​style services. Candidates coming through the competitive examination are referred to as ‘regular recruits’ or ‘direct recruits’. Two other streams of entry exist (known as promotion 1 R. Mukherjee, Senior Public Service: High Performing Managers of Government, World Bank Working Paper, Washington, DC: World Bank, 2004. 2 B. Nunberg, Managing the Civil Service: Reform Lessons from Advanced Industrialized Countries, World Bank Discussion Paper 204, Washington, DC: World Bank, 1995.

Personnel of the Indian State  55 and selection), which follow less rigorous procedures; these allow state civil servants to be promoted into the IAS. The recruitment examination is conducted by the UPSC, an apolitical body, and is common across the IAS, IFS, IPS (but not Forest Service), and most non-​technical Central Services. Successful candidates are allotted services on the basis of their preferences and examination ranks. Most candidates choose the IAS/​ IFS as their first two preferences (not necessarily in that order), and thus those selected for these services are usually the highest ranked. IAS officers are allotted to state cadres in which they specialize, learning the state language, customs, laws, and so forth, and clearing examinations in these subjects. They undergo a two-​year training, about half of which is spent in the National Academy of Administration, Mussoorie, involving a rigorous multidisciplinary course of examinations and assignments. One year is spent in on-​the-​job training in the state of allotment, including long stints in rural areas. Part of their career is usually spent in the Union government. The Union government’s Department of Personnel and Training (DoPT) is in charge of the administration of the IAS. Though the IAS came into existence in 1948, it inherited the mantle of the colonial Indian Civil Service (ICS) and members, especially in the early years, saw themselves as successors of the ICS and its traditions. Key administrative and police positions in state governments are designated as ‘cadre posts’, signifying that they may only be held by officers of the IAS/​IPS. This is a deliberate feature of the AIS system intended to promote quality, impartiality, integrity, and an all-​India outlook. Thus, the Secretaries (equivalent to ‘permanent secretaries’ in most Commonwealth countries) of most state ministries and the head of the civil service (the Chief Secretary) are always IAS officers, and the senior police officers (Inspectors-​General) and the head of the state police (Director-​General of Police) are always IPS officers. The institution of the civil service occupies an important place in India’s governance, a fact recognized by the Indian Constitution. The Central Civil Services also play a crucial role in the administration of the GoI. Most of them are recruited through the same common civil services examination as the IAS. The IFS is equivalent in pay, prestige, and promotion opportunities to the IAS. The Indian Revenue Service has two separate components—​one for income tax and the other for Customs and GST. The Indian Audit and Accounts Service works under the Comptroller and Auditor General. A list of the main Central Services can be found in Table 4.2.

Table 4.2  Distribution by Service of Senior Officers in Central Government Ministries, 2012 Service (those in italics are recruited through generalist Civil Services examination) IAS IPS IFS Indian Forest Service Indian Audit and Accounts Service Indian Civil Accounts Service Indian Defence Accounts Service Indian Economic Service Indian Statistical Service Indian Revenue Service (Customs and Central Excise) Indian Revenue Service (Income Tax) Indian Postal Service

Secretary level

Additional secretary level

Joint secretary level

Director level Deputy secretary level

Total

81 18 3 0 0 0 0 0 0 0

100 10 0 0 1 0 2 0 0 1

308 37 4 11 20 0 18 6 0 4

136 26 6 59 30 3 28 16 7 31

46 9 2 2 9 4 12 14 12 2

671 100 15 72 60 7 60 36 19 38

0 1

2 0

7 2

49 23

9 4

67 30

Indian Information Service/​Broadcasting Engineering Service/​Indian Broadcasting Programming Service Railway Services Central Power Engineering Service Indian Defence Estates Service Indian Legal Service Central Secretariat Service Indian Telecom. Service Indian Ordnance Factory Service Indian Cost Accounts Service Central Trade Service Indian Post & Telecom. Finance & Accounts Service Others Total Percentage of IAS officers Percentage of AIS officers

0

0

8

5

1

14

1 0 0 3 0 0 0 0 0 0

0 0 0 0 1 0 0 0 0 1

11 0 0 0 9 0 2 0 0 2

49 2 4 0 * 39 26 5 5 18

1 4 1 0 * 1 1 11 0 4

62 6 5 3 * 40 29 16 5 25

6 113 72 88

1 119 84 92

0 449 68 77

21 * * *

6 * * *

34 * * *

Source: K.P. Krishnan and T.V. Somanathan, The Civil Service, in Devesh Kapur, Pratap Bhanu Mehta and Milan Vaishnav (eds.) Rethinking Public Institutions in India, Oxford University Press, 2017.

58  The Institutional Design of the Indian State

4.2  The Design of the AIS The initial design of the AIS had the following key features: All-​India character: The services and its members were designed to be all-​India—​and not local—​in attitude and thinking, for it was intended that the ‘common national interest should always prevail over sectional and local claims’. For this reason, the AIS were to occupy key positions at both the national and state levels and to be a source of objectivity and neutrality rising above regional and local pressures including language, religion, and caste. An explicit objective was that officers should alternate between centre and states so that policymaking at the centre was imbued with a strong sense of practicality gained by actual experience in the field, while policymaking and implementation in the states was informed by a broader national perspective. The cadre allotment rules and central staffing scheme were also designed to serve this objective by ensuring that a proportion of AIS officers in each state were natives of other states and that officers alternated between centre and states. Dual Control: The need for a single service working at the central and state levels, along with democratic accountability to state governments, led to a design based on ‘dual control’ of the AIS by both tiers of government. Merit-​based selection: Selection was to be based on merit through a stiff competitive examination with the intent of attracting the best and the brightest to the public service. Independent Selection: Merit-​based selection was entrusted to an independent constitutional body (the UPSC). Protection against arbitrary punishment: The Constitution, under Article 311, made provisions to protect civil servants from arbitrary punishment to ensure their objectivity and neutrality and enable them to function without fear of political reprisal. Other design elements: These implicitly or explicitly included common training with other services (to build camaraderie), low age of recruitment (for a spirit of idealism), limited number of attempts at the examination (to ensure intellectual calibre), good remuneration

Personnel of the Indian State  59 OBJECTIVE

ACHIEVED THROUGH

Preserving national unity and integrity and uniform standards of administration Neutrality and objectivity – non-political, secular, nonsectarian outlook

Competence, efficiency and professionalism • At entry (by attracting the “best and brightest” • Throughout career Integrity Idealism

DESIGN FEATURES Explicit • All India character of the services (centralized recruitment, dual control) • Cadre regulations • Officers work in Center and States • Meritocratic • Career advancement on ‘merit with due regard to seniority’ • Protection against arbitrary removal, demotion, punishment • Low age at entry Implicit • Good remuneration and service conditions • Social status and prestige Special Considerations • Promotion from State services • Reservations

Figure 4.1  Initial design of the All India Services

and service conditions, and social status and prestige (to attract talent and preserve morale). Special considerations: These were two special considerations in the initial design that modified the other elements described. The first was promotion from the state civil services and the second was reservation for Scheduled Castes (SC) and Scheduled Tribes (ST). The objectives of the initial design and the design features through which they were sought to be achieved are summarized in Figure 4.1.3 Clearly, most features of the design were conducive to institutional effectiveness since the design promoted competence (through meritocratic recruitment), integrity (through good remuneration and job security), political neutrality (through independent selection security of tenure and protection against arbitrary removal), dynamism (through recruitment at a young age), and flexibility and social sensitivity (through the requirement to work in the states and districts, thereby keeping civil servants firmly connected with grassroots reality). The initial design was modified over the years by explicit changes as well as through the effects of various political, economic, and social 3 Figure 4.1 is taken from Krishnan and Somanathan (2005). This figure from that paper was also cited and reproduced by the 2nd Administrative Reforms Commission in its 10th Report.

60  The Institutional Design of the Indian State forces. The main explicit changes were the introduction of 27% reservation for the Other Backward Classes after 1995, an increase in the state service promotion quota from 25% to 33%, the multiple increases in the upper age limit and the number of attempts allowed, and changes in the examination pattern itself. These changes have significantly altered the demographic profile of the successful candidates. The upper age limit increased from 26 and two attempts in 1978 to 32 and six attempts by 2014, with an attendant increase in the average age of successful candidates. In fact, in 1991, 76% of the candidates recommended were below the age of 26, while in 2008 only 33% were so. The average age of successful candidates is higher than the overall candidate population, which may reflect the advantage of multiple attempts. Further, the increased age and attempts coupled with the general national culture surrounding examinations mean that the civil services examination has joined the other competitive examinations in becoming a ‘practised’ race with a large industry of coaching centres supporting it.

4.3  The Intermediate and Lower-​Level Services The intermediate and lower services of the central and state governments present a very heterogeneous picture. Central and state governments classify their employees into four categories. The Group A or Class I officers have already been described above under the ‘higher civil services’. The Group B or Class II category staff are the next level of executive staff. This category of staff in turn are of two types—​ones who hold significant administrative responsibilities and are, or used to be, called gazetted officers, and others who do not and are, or were, called ‘non-​gazetted’. (The gazetted/​non-​gazetted distinction is, however, becoming less common.) The Group C or Class III category staff belong to the clerical establishment and have non-​supervisory roles. The Group D or Class IV category staff are the lowest administrative category and consist of those doing largely unskilled work: for example, messengers, cleaners, etc. The Seventh Pay Commission merged the Group C and D categories, though several state governments continue to maintain them as separate categories for their employees.

Personnel of the Indian State  61

These categories have several modes of recruitment. Some recruitments are done in a centralized manner, and some others at the level of the Departments concerned. In the case of the GoI, recruitments to Group A and B posts are generally done by the Union Public Services Commission (UPSC). In the case of the state governments, its Group A and B recruitments are done by the respective State Public Service Commissions (PSCs). The recruitments of Class C (non-​technical) and sometimes even Class B staff to various Ministries and Departments of the central government and their subordinate offices are done by the Staff Selection Commission (SSC). Railways and Defence, the two largest recruiters, have their internal recruitment boards.

62  The Institutional Design of the Indian State Several state governments have set up specialized bodies for recruiting certain kinds of staff—​for instance, Tamil Nadu has a Teachers Recruitment Board, a Medical Services Recruitment Board, and a Uniformed Services Recruitment Board. Public sector undertakings of the central government generally have their internal recruitment processes, but top-​level appointments are made by a Public Enterprises Selection Board and appointments to nationalized banks by a Banking Boards Bureau. Then there are decentralized recruitments. In the case of the department-​level recruitments, they are done at various levels of the government—​Heads of the Department or Regional Directors, and District Collector or Zilla Panchayat. Most often, these recruitments are done as part of standardized state-​level processes. The credibility associated with these recruitments varies widely. The UPSC is generally accepted as the gold standard, with its selections seen as based on merit and free of corruption. Many (but not all) state PSCs and the central SSC too have a fair degree of credibility and rigour. But the credibility and rigour of recruitments done at sub-​state levels vary widely. One of the largest numbers of recruitments involves teachers, which are done periodically across states. The quality of the process too varies widely across states. An area of debate in recent times around the quality of the recruitment process has been on the issue of interviews. It is often argued that interviews compromise the quality of selections by introducing significant subjectivity and therefore become a source for influence-​peddling in recruitments. The recurrent outbreaks of scandals across states, where interview marks have allegedly been manipulated in return for bribes, lend credence to this argument. Accordingly, in 2015, the GoI decided to dispense with interviews for recruitments for junior-​level posts of central government ministries and their public sector undertakings, Group B (non-​gazetted) and below.4 Some state governments too have followed suit.5 4 No interviews for government jobs at junior level from January 1, The Economic Times, 29 December 2015 (https://​econom​icti​mes.ind​iati​mes.com/​news/​econ​omy/​pol​icy/​no-​int​ervi​ews-​ for-​gov​ernm​ent-​jobs-​at-​jun​ior-​level-​from-​janu​ary-​1-​skill-​test-​may-​conti​nue-​dopt/​arti​cles​ how/​50369​824.cms?from=​mdr). 5 You can now get an Andhra Pradesh government job without an interview, Livemint, 18 October 2019 (https://​www.livem​int.com/​news/​india/​you-​can-​now-​get-​an-​and​hra-​prad​esh-​ gov​ernm​ent-​job-​with​out-​interv​iew-​115​7136​8499​248.html) (Accessed on 12 January 2020).

Personnel of the Indian State  63 There are other ad hoc methods of recruitment. One is compassionate appointments—​appointments to the next of kin of staff who die in service. Then there are one-​off regularizations of long-​serving contractual employees. Such regularizations are usually either political decisions, based on electoral promises or otherwise, or are forced by directions of courts on petitions filed by the contract employees. Governments often end up relaxing the recruitment guidelines to regularize them. Apart from these regular recruitments, there are also contractual appointments at various levels of the government, which are ostensibly temporary and for a particular purpose. Like with the regular appointments, these too vary widely in their nature and rigour. The trend towards regularization of contract employees has led to a reduction in this kind of recruitment as many governments have changed their rules to disallow contractual appointments. As described above, those occupying the leadership positions of various central government departments are mostly recruited through the annual generalist civil services examinations administered by the UPSC. Within the central government, top-​level ‘control’ posts are manned partly by the Central Services and partly by the AIS. Table 4.2 details the share of the different services (including the AIS) in the control posts in the central government.6 As the table shows, the AIS, particularly the IAS, have a pivotal role in terms of its share of top-​level policymaking positions across the range of central ministries; the IAS and IPS together account for 88% of Secretary-​level posts. Furthermore, while the data are not presented here, they occupy an even larger percentage of the control posts in state governments. However, this position has undergone very significant changes in recent years. The share of IAS officers occupying posts below the level of Secretary has declined dramatically and officers from other central services have taken their place. The decline in the batch sizes of the IAS and

6 These are primarily the posts in the ministries (contrasted from posts in the operating ‘Heads of Department’ subordinate to the ministries). However, posts in the Intelligence Bureau and central police organizations are control posts de facto (though within Heads of Department de jure) and hence included in the table in respect of the IPS.

64  The Institutional Design of the Indian State the lower number of officers from the cadres availing deputation to the GoI have been contributory factors. While accurate statistics are not available, it is today well-​known that far less than half, probably even less than a third, of the important Joint Secretary level posts in the GoI are held by IAS officers.

5 Civil Service Effectiveness An Assessment

In Chapter 2, the different tiers of government were described. In every tier of the executive branch of government (which is by far the largest branch), a critical role is played by the civil service. The effectiveness and capability of the civil service is thus a crucial determinant of the effectiveness and capability of the state as a whole at central, state, and local levels. This chapter examines the effectiveness of India’s civil services.1

5.1  Criteria of Effectiveness There are various possible criteria of ‘effectiveness’ of the civil service as an institution. One perspective is from institutional economics, and from that angle civil service effectiveness would be assessed on the basis of reducing transaction costs for economic agents and thereby promoting economic development. However, this approach is too narrowly rooted in economics. Democracy, secularism, and other freedoms have value in themselves regardless of their economic effects. Further, the institutional economics perspective links the effectiveness of the civil service to outcomes that may be the result of deliberate policy choices of others, namely elected politicians. Faithful implementation of policies of an elected executive (insofar as they do not contravene the law or the constitution) is the mark of an effective, not ineffective, civil service even when the policies may increase transaction costs or are perceived as ‘wrong’ by economists. Also, the preservation of the constitutional order is itself an objective for the civil service in a democracy. 1 This chapter draws heavily from the Civil Service chapter (K.P. Krishnan and T.V. Somanathan), in Devesh Kapur, Pratap Bhanu Mehta, and Milan Vaishnav (eds.), Rethinking Public Institutions in India, Oxford University Press, 2017. State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0005

66  The Institutional Design of the Indian State Alex Matheson et al. evolved a framework of four public service behaviours that corresponds quite closely to the criteria discussed above. They argue that the four behaviours (constitutional respect and continuity, impartial and inclusive public service, responsive public service, and performing civil service) are in a hierarchy with constitutional respect and continuity as the most important, impartiality and inclusiveness as the second, and responsiveness and performance (jointly) in third place in the hierarchy.2 Drawing upon that framework, this chapter uses the following criteria to assess civil service effectiveness: 1. Criterion I: Preserving India’s constitutional order, including democracy, secularism, national unity, and the rule of law (being ‘good’ in themselves, regardless of their effect on economic development); 2. Criterion II: Impartial implementation of the rule of law in day-​ to-​day dealings with the citizenry (this is partly connected with Criterion IV because predictability in day-​to-​day working is crucial for economic development). 3. Criterion III: Faithfully translating the will of elected governments into policies and then implementing those policies effectively. 4. Criterion IV: Promoting economic development by reducing transaction costs and by providing effective and efficient public services. Criteria III and IV broadly represent the role of the civil service in what Mark Moore calls ‘creating public value’.3 The traits that civil servants need in order to be effective vary across these criteria. Criterion I is primarily related to political neutrality and a sense of national spirit and larger purpose. For Criterion II, integrity (lack of corruption) is the most vital trait, though it also requires political neutrality. Criterion III depends considerably on political neutrality (so that civil servants are able to implement policies contrary to their private 2 A. Matheson, B. Weber, N. Manning, and E. Arnould, ‘Study on the Political Involvement in Senior Staffing and Delineation of Responsibilities between Ministers and Senior Civil Servants’, OECD Working Papers on Governance 2007/​6, OECD Publishing: France, 2007. 3 M.H. Moore, Creating Public Value: Strategic Management in Government, Cambridge: Harvard University Press, 1995.

Civil Service Effectiveness  67 beliefs and ideologies) and competence and capacity (so that they execute policies effectively). It also requires flexibility and willingness to change, social sensitivity, and idealism and empathy with the public. Criterion IV is primarily an issue of capacity and competence, the ability to regulate economic matters effectively, but is also closely related to levels of integrity. A civil service that is competent but dishonest may not be conducive to reducing transaction costs, though the experience of East Asian countries including China suggests that the actual effect depends on the manner and nature of the dishonesty.4 ‘Effectiveness’ under one parameter does not always mean ‘effectiveness’ under another since there are often trade-​offs between them. A ‘committed’ civil service, for instance, may score well on Criterion III but end up undermining democracy under Criterion I. A politically neutral and honest but rigid and inflexible civil service may do well under Criteria I and II but, by providing ineffective and incompetent administration, fail under Criteria III and IV. The next section presents the authors’ own assessment of the effectiveness of the Indian civil service under these four criteria.

5.2  Assessing Effectiveness of the Indian Civil Service On Criterion I (preserving the constitutional order), the civil service has been quite effective and, by developing country standards, outstandingly so. The preservation of national unity and constitutional rule through many instances of insurgency, civil disorder, political instability, and dozens of impositions of President’s Rule on the one hand and the conduct of free and fair elections (in a country not otherwise known for excellence in governance) on the other, lie at the core of India’s success as a democratic polity. The conduct of a free and fair election in Jammu and Kashmir in December 2020 under trying circumstances, resulting in a majority for parties opposed strongly to the ruling party at the Centre, is simply the most recent in a long series of examples. They are achievements for which the IAS and IPS have been responsible in large measure but for 4 Indeed, corporate executives often seem to suggest that they do not mind corruption so long as ‘services’ paid for corruptly are delivered quickly and without uncertainty.

68  The Institutional Design of the Indian State which the services do not get as much credit as they deserve. It is worth noting that the Election Commission and the UPSC—​both nearly always headed by officers from the civil service and insulated from direct political control—​continue to enjoy a well-​deserved high reputation for neutrality and probity. Persons outside government rarely appreciate the extent to which the ability to hold India together (as a single political and economic unit), and for the constitutional distribution of executive powers to operate, has depended, and continues to depend, on the AISs. In a country of unequalled diversity, the AISs have also played a crucial role in enabling the central government to understand local pressures and dynamics and adapt its policies realistically within the constitutional framework to meet local aspirations without the kind of rigidity often witnessed when federal governments have civil services unconnected with the federating units. Without necessarily attributing causality, it is noteworthy that Pakistan made major changes to the civil service structure and has not been successful in preserving democracy and constitutionality.5 On Criterion II, namely neutrally and fairly implementing rule of law in dealings with individual citizens, the civil service has been far less effective. The powerful and the influential have often received favourable treatment with the poor and the weak often treated shabbily. Enforcement of the law has been erratic. Political interference (see definition in the next section, Causes of Poor Performance) has played a major role, but the civil service is not free of blame. Corruption by civil servants has become rampant. The civil service has often not put up the level of resistance that its many legal and constitutional protections should have enabled it to. Partly, this is because of the incentive structure under which modern-​day civil servants operate. However, to provide perspective, it should be noted that the poor overall record on this issue is mitigated slightly by the impressive work of several distinguished officers who have spent whole careers fighting within the system for the fair treatment of the disadvantaged sections of society. Some like the late (and legendary) S.R. Sankaran became well known, while many others have remained in relative anonymity.6 The number of such

5 World Bank, ‘A Framework for Civil Service Reform in Pakistan’, Report No. 18386-​PAK, Washington, DC: World Bank, 1998. 6 E.A.S. Sarma, ‘S R Sankaran: In Memoriam’, Economic and Political Weekly, 45, no. 3 (23–​29 October 2010): 25–​27.

Civil Service Effectiveness  69 officers, while a small proportion, is large enough that the image of ‘the crusading honest officer’ continues to find a place in popular cinema and motivates many young students to write the civil services examination. On the third criterion of faithfully executing policy, the assessment is mixed. There is some truth in the perception that the civil service has from time to time obstructed, or at least delayed the implementation of, well-​intentioned plans of the political executive. In part, this is an unconscious or indirect result of other problems like corruption and weak law enforcement. In some cases, the resistance to implementation comes not from the senior levels but from the lower levels of civil service.

70  The Institutional Design of the Indian State In the authors’ opinion, deliberate obstruction is less common than may be thought by outsiders. Genuine objections to illegal or improper courses of action can also appear or be characterized as ‘obstruction’. Some delays and obstructions attributed by the press to the civil service are (from the authors’ own experience) the result of lack of cohesion within the political executive, especially in coalition governments, where what appears to be a civil servant’s disagreement is actually a reflection of disagreements among Ministers. Also, the civil service often lies at the intersection between the political executive’s ‘publicly proclaimed lofty intentions and privately expressed specific instructions (usually less lofty)’. There is a lot of evidence to show that where the political executive has a clear and genuinely held policy view and expresses it consistently, the civil service does usually deliver.7 Thus, it is difficult to discern the extent to which the civil service is legitimately to be blamed under this criterion.

The retired diplomat, T.C.A. Rangachari,* recalls the following anecdote: Back in the summer of 1977, soon after the Janata government came to power, the Ministry of External Affairs arranged a Heads of Missions meeting in Delhi for India’s Ambassadors and High Commissioners to get a first-​hand view of the new government’s approach to domestic and foreign policy. Several Ministers addressed them, among them Home Minister (later Prime Minister) Mr. Charan Singh. One of our Heads of Missions asked: ‘Sir, for the first time since Independence there is a non-​Congress government at the Centre. How do you expect a bureaucracy that has served the Congress ideology and programme for thirty years, to now serve the new government’s agenda faithfully?’ Mr. Charan Singh responded in Hindi: Afsarshahi to ghode jaisi hoti hai. Agar ghud sawaar nipun hai to ghode ko wahaan le jaayega jahaan woh jaana chaahata hai. Agar ghode ko andaaz ho gaya ki ghud sawaar anaadi hai to ghoda use avsar milte hi patak dega. 7 There are several examples at different points of time, including the successful implementation of the noon meal scheme in Tamil Nadu schools, protection of agricultural tenants under Operation Barga in West Bengal, reliable rural power supply in Gujarat, and the conduct of Kumbh Melas in various states.

Civil Service Effectiveness  71 Loosely translated: ‘The bureaucracy is like a horse. If the rider is an expert, it takes him where he wants to go. If it detects that the rider is incompetent, it throws him off when it gets an opportunity.’ * Email to Indiceconomic Network, 7 May 2021.

On Criterion IV (effectiveness in promoting economic development/​reducing transaction costs and providing effective public services), the civil service has not performed well. There is a vast body of facts and evidence which shows that policy uncertainty, inconsistent or arbitrary application of known policy, delays in decision-​taking, ‘transaction costs’ in money and time due to bribery, ineffectiveness in law enforcement, and regional or communal parochialism (impeding the development of all-​India institutions and markets) have been widespread characteristics of civil service functioning in the last three decades. In addition, lack of competence has been a problem quite often, though far less so than lack of integrity or neutrality. All of these have reduced the effectiveness of the civil service in promoting economic development. In sum, the civil service has done well in preserving the overall constitutional order but performed poorly in impartially implementing laws and policies at the individual level. Its record on the third criterion of responsiveness to political will is mixed but on the whole reasonably good, considering that political will has often been unclear or contradictory. It has performed badly in promoting economic growth or providing good public services.

5.3  Causes of Poor Performance The overall conclusion is that the civil service has not met the expectations of the design or the needs of the country. The main causes of the poor performance are discussed here. The biggest single weakness of the civil service today is its perceived inability to function with integrity and political neutrality. Political

72  The Institutional Design of the Indian State interference has become very common over the last three to four decades. Political interference is defined as: Acts of politicians—​whether formally part of the Executive or not—​intended to compel a civil servant, by means of threats or blandishments or both, to follow the course desired by them on decisions which ought to be taken by the civil servant impartially under government policy. It does not refer to the legitimate role of the political executive (ministers and politically-​appointed heads of agencies) in exercising powers duly vested in them—​a role which may indeed lead them to legitimately overrule advice tendered by civil servants. The distinction is quite easy to observe in practice; political interference is almost always oral, with the civil servant usually pretending that he took the decision of his own volition.

Political interference is currently perceived to occur at all levels: a local Member of the Legislative Assembly (MLA, or state legislator) may interfere with a petty official in the issue of a land title document, a state Minister may interfere with a District Collector in a personnel decision, a central Minister may interfere in the award of a contract or in the enforcement of a penal provision of the law, and so on. Political interference is also felt to be politically and geographically neutral; members of all political parties (national and regional) in all states are perceived to indulge in it, though there may be variations in degree. Often, political interference takes the form of telling a civil servant not to tender advice in the manner the civil servant intended to tender it, which is reflected in a failure to record dissenting views (as legally required under the All India Services [Conduct] Rules). Popular commentary on the civil service often compares the contemporary civil service adversely with the ‘steel frame’ of the ICS. It needs to be remembered that the ICS did not function under a democratic framework. Except for a very short and partial interlude, it did not have any elected political executive to deal with; thus a comparison of the IAS with the ICS is not apt. However, a reasonable comparison would be with the civil service in the first three decades after Independence. On that basis,

Civil Service Effectiveness  73 the evidence is clear that the levels of integrity and neutrality were high then and are much lower now.

Many critics (and critiques) of the civil service dwell on the view that the civil service has only itself to blame for the problem of political interference. ‘Spineless’, ‘crawling when asked to bend’, and ‘corrupt and self-​ serving’ are among the epithets commonly used. If every officer said ‘no’ to an improper instruction and boldly accepted all penal transfers (after all transfers do not affect salaries), political interference would end—​so goes a common critique.

74  The Institutional Design of the Indian State

Since civil service recruits do not represent an anthropologically different group from the rest of university-​educated India, the more interesting analytical question is why civil servants behave as they do. The answer to the question lies in a basic change in the structure of career incentives. The design principle was that civil servants would be secure in terms of pay and perquisites, which could not be arbitrarily varied, favourably or unfavourably, by the political executive, so that they would act without fear or favour in accordance with the law. Article 311 of the Constitution provides protection to civil servants against arbitrary dismissal and demotion through a set of strong procedural safeguards. These safeguards have been considerably widened through judicial interpretation. As a result, arbitrary de jure punishment of civil servants is extremely difficult. Promotion is mainly seniority-​based and administered largely by the

Civil Service Effectiveness  75 civil service itself, with little political involvement. The All India Service Rules prescribe a fixed age of mandatory retirement and exceptions were extremely rare.8 Together, they meant that the political executive could not offer special rewards to favoured civil servants. However, through a combination of the wielding of a stick (at the state level) and the dangling of a carrot (at the central level), the political executive has gradually acquired the power to punish those they dislike and confer special rewards on those they favour. The ‘sticks’ of personally disruptive transfer and arbitrary demotion have been acquired and used to telling effect, largely but not exclusively at the state level. The former kind of ‘stick’—​frequent transfer to different duty stations, thereby disrupting the personal life of the officer as well as the employment of spouses and the education of children—​is well known. Such transfers operate as de facto punishments due to the adverse personal consequences for the officer. The lesser known but equally widespread problem is arbitrary demotion. This has happened through the power of transfer combined with the ability to upgrade and downgrade particular posts at will, without any semblance of public interest. Thus, by creating a new post with the same pay scale but with clearly lower responsibility or upgrading an existing junior post and making it senior, and then transferring a senior officer to that (junior) post, the government is able to effectively demote an officer. This has primarily been done by the state political executive but the central government and the judiciary have effectively endorsed the right of state governments to behave in this manner. Therefore, while the constitutional protection against arbitrary dismissal remains, the protection against arbitrary demotion does not exist in practice. AIS officers today can be effectively punished by the political executive with no procedural formalities, let alone procedural fairness. The ‘carrot’ of conferment of special reward beyond normal conditions of service has become widespread, largely but not exclusively at the central level. This is through resort to extensions of service and to post-​retirement employment in regulatory or quasi-​government bodies. The power of extension (for more than six months) rests exclusively with the centre, and hence this is more common at the central level. 8 The principle has been diluted in recent years with provisions for limited extensions beyond the age of 60 for a small number of posts: six months for Chief Secretaries of States, two years for five Secretary-​level posts in the Union (two held by the IAS and three by the IPS), and up to four years for the Cabinet Secretary.

76  The Institutional Design of the Indian State Apart from de jure extensions, de facto extensions occur through the appointment of retiring officers to regulatory bodies, statutory commissions, and the like. This practice has never been absent and, realistically, can never be completely proscribed in any civil service. However, in the early decades, the number of such posts was very small. Das has argued that the recent proliferation of regulatory bodies, tribunals, and training institutions has made the practice widespread and the manner of exercise has been perceived in recent decades to be geared to reward pliability.9 The dangling carrot of an extension or post-​retirement job means that officers in the most powerful posts of their careers (Secretaries) are often more pliable than their juniors, since acquiescence or support to political interference may be rewarded with post-​retirement ‘after-​life’ and perquisites to boot. Further, whereas those on the extension of service cannot draw more than their pensions, those in post-​retirement jobs in autonomous bodies may be able to double-​dip by collecting both pension and the new salaries. Another facet of political interference is the politicization of promotion of state service officers into the IAS; officers close to particular political parties or personalities are perceived to have an advantage in getting promoted to the IAS/​IPS. Officers from the promotion and selection categories form approximately a third of the service. In addition to these, there have been some other factors that have diminished civil service effectiveness. First, dual control has weakened especially in the last two decades and driven by the compulsions of politics in a coalition era, the centre was perceived as not having protected AIS officers even where it had the power to do so. Second, arbitrary transfers have led to short posting tenures, which in turn greatly diminishes effectiveness and also leaves officers vulnerable to seeking favours for good postings or longer tenures. Lakshmi Iyer and Anandi Mani have found empirical evidence that ‘suggest(s) that the cost of political transfers in terms of longer-​term (development) outcomes can be quite high.’10 Third, the non-​transparency and the importance of annual appraisals in the system of ‘empanelment’ through which only a minority of IAS officers are made eligible for senior postings at the Centre leaves officers vulnerable to the discretion of political executive, besides also being harsh on frequently transferred officers who are likely to miss a non-​trivial share of appraisals. 9 S.K. Das, Building a World Class Civil Service for 21st Century India, New Delhi: Oxford University Press, 2010, Chapter 17. 10 L. Iyer and A. Mani, ‘Traveling Agents: Political Change and Bureaucratic Turnover in India’, Review of Economics and Statistics 94, no. 3 (August 2012): 723–​739.

Civil Service Effectiveness  77 Fourth, judicial interpretation in widening the protections of Article 311 and considerable judicial leniency to civil servants who perform inefficiently or negligently has reduced efficiency.11 Fifth, the dearth of successful prosecutions of the corrupt has made corruption a ‘financially efficient’ strategy when risks are set against rewards. Sixth, the drastic pay compression (by way of reduction in the ratio of senior officers’ pay vis-​à-​vis the junior ranks of the government service) until the 1990s accelerated the loss of morale and created an excuse for corruption.12

Finally, socio-​economic trends, especially the explosion in private sector salaries and the culture of conspicuous consumption, have increased corruption, reduced morale, and damaged the social prestige of a civil service career, removing a major ‘non-​monetary’ perquisite. 11 For details, see the section entitled ‘The civil service: An institutional perspective’ by K.P. Krishnan and T.V. Somanathan, in Devesh Kapur, Pratap Bhanu Mehta, and Milan Vaishnav (eds.), Rethinking Public Institutions in India: Performance and Design, Oxford University Press, 2017. 12 K.P. Krishnan and T.V. Somanathan, ‘The Civil Service’, in Devesh Kapur, Pratap Bhanu Mehta, and Milan Vaishnav (eds.), Rethinking Public Institutions in India, Oxford University Press, 2017, Chapter 9.

78  The Institutional Design of the Indian State A few of the factors mentioned above are indeed attributable to the civil service itself, for instance, the creation of post-​retirement posts and the dysfunctions of the empanelment system; most, however, have arisen from factors originating outside the civil services.

5.4  Misaligned Incentive Structure The current incentive structure of the civil service is summarized in Table 5.1.13 Using the analogy of (financial) portfolio theory, the different behaviours can be divided into ‘efficient’ and ‘inefficient’ ones in terms of the risk/​reward balance. The following kinds of behaviour are ‘efficient’ and thus encouraged:

1. Acquiring/​retaining professional competence; 2. Passive political neutrality14; 3. Political alignment; 4. Passive honesty; and 5. Bribery.

The following types of behaviour are ‘inefficient’ and therefore discouraged: 1 Active neutrality15; 2 Enforcing the law independently; and 3 Innovation. This assumes that each of these types of behaviour can be pursued independently of the others. In practice, there are ‘synergies’ between some of them—​for example, between passive neutrality and bribery. An officer 13 As with the rest of the chapter, the analysis is as of 2013. 14 Defined in Krishnan and Somanathan, ‘The Civil Service’, as ‘total submission to whoever is in power without any attachment to that party per se. . . . [carrying] out the bidding of the ruling party even if it is contrary to established rules or procedures but [making] it clear to the “losers” that [one] is simply carrying out “orders” and is not really “committed” to the current ruling group; should the ruling party change, [one] would equally happily carry out the bidding of the new group’. 15 Defined in Krishnan and Somanathan, ‘The Civil Service’, as ‘when taking decisions on individual matters, [the officer] does not favour those belonging to, or recommended by the ruling party, and applies rules and norms impartially.’

Table 5.1  The Incentive Structure of the Civil Service Behaviour

Consequences

Risk/​reward balance

Rewards

Punishments

Career

Social

From illegal activity

Career

Social

From illegal activity

1. Acquiring/​retaining professional competence 2. Political neutrality —​Active neutrality

Low

Low

N/​A

Nil

Nil

N/​A

Low risk/​low reward

Nil

High

N/​A

High

Nil

N/​A

—​Passive neutrality —​Political alignment 3. Acting independently in enforcing the law and due process 4. Honesty and integrity —​Active honesty —​Passive honesty —​Bribery

Low High Nil

Nil Low High

N/​A N/​A N/​A

Low High High

Nil Low Nil

N/​A N/​A N/​A

High risk/​moderate reward Low risk/​low reward High risk/​high reward High risk/​moderate reward

Nil Low Nil

Low Nil Low

N/​A N/​A High

High Nil Mod.

Nil Nil Low

N/​A N/​A Mod.

5. Innovation

Nil

Low

N/​A

Mod.

Nil

N/​A

High risk/​low reward Low risk/​low reward Moderate risk/​high reward Moderate risk/​low reward

80  The Institutional Design of the Indian State who is passively neutral has a better chance of maximizing corrupt opportunities. Likewise, an officer who keeps professionally up-​to-​date is more likely to be innovative. Active honesty and active political neutrality go together. There are negative synergies too: a corrupt officer has little to gain from acquiring professional skills. Going by this analysis, the theoretical prediction would be that the average IAS officer would tend to be reasonably competent but politically spineless, incapable of enforcing the law fairly, and corrupt. Iyer and Mani’s study provides empirical corroboration to several of these inferences.16 This analysis has its limitations, and fortunately (because behaviour is not merely a product of incentives and factors like esprit de corps do matter) there are still a large number of honest and efficient officers. However, it does show that there is a major misalignment between the actual incentives facing IAS officers and the objectives of the service according to the constitutional design. Without changing the incentive structure, it is futile to expect the majority of civil servants to conform to the behaviours expected in the original design.

5.5  Recent Trends Affecting the Civil Service In the first few decades after Independence, the lower ranks of the civil service (i.e. Group B and Group C levels) held a lot of attraction for graduates of the universities. Good private-​sector jobs were relatively few and the stability and decent pay and conditions of the public service made them the first choice for those not able to compete at the Group A (IAS, etc.) level. Thus, the calibre of entrants was good. At the central level, selection was also largely merit based. At the state level, arbitrary selection was more common but generally not based on outright or widespread corruption. Within the service, the systems of performance appraisal left behind by the British still operated, and so clerical and supervisory levels were also subjected to annual performance appraisals written by their immediate supervisors. The appraisals had some effect on career progression, which meant a fairly high level of discipline. The performance of the lower rungs of the government (central and state) are perceived as having deteriorated in recent years. The lower ranks are the points of public interface, so poor performance at this level is a significant contributor to poor civil service performance overall.

16 Iyer and Mani, ‘Traveling Agents: Political Change and Bureaucratic Turnover in India’

Civil Service Effectiveness  81

Compounding the challenges, the record of State Public Service Commissions is now very poor17 (unlike the Union Public Service Commission, which retains a reputation for independence and integrity). Staff associations have begun to behave in a trade union-​like manner and also have grouped themselves across caste categories. They have successfully lobbied for dilution or dismantling of performance appraisal systems, and in many states, most of the non-​Group A civil servants have no formal annual appraisals. The combination of security of tenure with non-​appraisal of performance has had a strongly negative effect on discipline, quality of work, and the ability of senior officers to get work done. These negative trends have accelerated in the last decade and a half. 17 ‘DVAC Raids 73 Premises of Candidates; TNPSC Chairman’, Indian Express, 13 January 2012; ‘PPSC Former Chairman Ravi Sidhu Handed Down 6-​Year Jail’, Times of India, 16 July 2013.

82  The Institutional Design of the Indian State

The growth of employment opportunities for young graduates in the information technology sector has adversely affected the attractiveness of the lower-​level civil service jobs and, anecdotally, appears to have reduced the quality of entrants. Even then, the relative insecurity and lack of work–​ family balance in the private sector has meant that government jobs still hold some attraction for good graduates. For central government posts, fairly good selection mechanisms have ensured that there is a reasonably fair route for entry, and thus some good candidates are still attracted into the Group B and C levels of the central government. In many state governments, for reasons already mentioned, the good candidates rarely enter the fray. Finally, there is the twin perception effect of increased corruption and decreased competence.18 There are several possible reasons for the perception of increased corruption. For a start, the greater politicization discussed

18 Krishnan and Somanathan, ‘The Civil Service’.

Civil Service Effectiveness  83 earlier has undoubtedly increased the opportunities and compulsions for corruption. The increased political involvement in both appointments and decision-​making has brought civil servants closer to politicians in culture and approach (despite the benefits of security of tenure and lack of need to spend on elections). The social pressures arising from high private-​sector pay enjoyed by comparators with similar educational backgrounds, the trend towards conspicuous consumption and away from simple living, weakening of public service values, etc., have also had an effect.

Further, the stakes involved in policymaking and regulatory administration have increased dramatically with economic growth and the increasing role of the private sector in the economy. There has been a sharp increase in the new regulatory powers of the state in areas ranging from utilities to the environment (reversing the trend of the 1990s when the licensing system was discontinued).

84  The Institutional Design of the Indian State The other perception, especially outside the civil service, has been that senior civil servants do not possess the level of competence needed to take good policy decisions or tender good policy advice at the state and central levels. On the one hand, exposure of civil servants to modern management concepts and to training in highly regarded international academic institutions has definitely increased. The familiarity of senior IAS officers with the latest jargon and buzzwords like ‘results frameworks’ or ‘gender budgeting’ and their ability to talk the talk of modern public management has definitely increased. On the other hand, there are many examples of poor civil service advice to Ministers. The ability of civil servants to coordinate policymaking between Ministries also appears to have deteriorated.

Civil Service Effectiveness  85 A difficulty with this issue is that the outside perception coincides with a period when public opinion has decisively turned against government in general to a mood of disenchantment. It is also difficult to estimate the extent of political interference in many questionable policy actions and therefore to assess whether there was a lack of civil service competence or whether competent advice was over-​ruled or whether advice was not tendered due to political interference. Governance is very often a ‘joint product’ of the civil service and the political executive, and attribution of competence or incompetence to one or the other is very difficult. One dimension of ‘competence’ or ‘capacity’ is the ability to decide or make specific policy recommendations. This has declined in recent years. It is not clear whether this is because of any change in ability or more due to a change in the incentives. It appears to be partly related to the effects of increased public scrutiny and a feeling that while a bad decision may lead to adverse consequences, indecision is safe. Notwithstanding the caveats, the authors, as insiders, do feel that there is a lack of competence in several areas of government and that many civil servants in key positions do not have the necessary subject-​matter and decision-​analysis skills. Capacity building is relatively non-​controversial and thus an area where reforms may be easier to implement. Some suggestions are offered in Chapter 12.

PART III

STAT E C A PA BILI T Y CHA L L E NGE S A ND WE A KN E SSE S

6 The Size of the Indian State 6.1  The Bloated State? The conventional wisdom paints the Indian state as overstaffed, apathetic and inefficient. This chapter explores the facts in terms of the personnel strength of the Indian state at various levels of administration. How does reality stack up against the perception? The perception of ‘bloated’ government has many origins. Right at the top of the government, at both the state and central government levels, the number of Ministries and Departments has undergone a dramatic expansion over the decades. The first central government under Jawaharlal Nehru had 21 Ministers in 16 Ministries,1 compared to 77 Ministers (including Ministers of State) in 55 Ministries today. State government Ministries too have undergone dramatic expansion over the decades. There were eight Ministers in the 1962 Tamil Nadu government under K. Kamaraj,2 compared to 34 today.3 This has been accompanied by similar increases in the administrative units at sub-​state levels. The number of districts has more than doubled from 356 in the 1971 census to 727 today.4 Many sub-​divisions have become full-​fledged districts. Similarly, the number of statutory towns has increased over the decades, rising from 2,987 in the 1991 census5 to 4,041 as per the 2011 census.6 New functional units have been created at all 1 Wikipedia, First Nehru Ministry, Retrieved on 8 December 2019 (https://​en.wikipe​dia.org/​ wiki/​First​_​Neh​ru_​m​inis​try). 2 Wikipedia, Third Assembly of Madras State, Retrieved on 8 December 2019 (https://​ en.wikipe​dia.org/​wiki/​Thi​rd_​A​ssem​bly_​of_​M​adra​s_​St​ate). 3 List of Ministers, Government of Tamil Nadu, Retrieved on 24 July 2021 (http://​www.tn.gov. in/​minist​ersl​ist). 4 Hemanshu Kumar and Rohini Somanathan, Mapping Indian Districts Across Census Years 1971–​2001, Working Paper No 176, Centre for Development Economics, Delhi School of Economics (http://​www.cde​dse.org/​pdf/​work​176.pdf). 5 1991 Census Handbook (http://​cens​usin​dia.gov.in/​Dig​ital​Libr​ary/​data/​Cens​us_​1​991/​Publ​ icat​ion/​India/​459​69_​1​991_​CHN.pdf). 6 Census of India 2011 (http://​cens​usin​dia.gov.in/​2011-​prov-​resu​lts/​pap​er2/​dat​a_​fi​les/​Ind​ ia2/​1.%20D​ata%20Hi​ghli​ght.pdf). State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0006

90  State Capability Challenges and Weaknesses levels to accommodate the expanding scope of government’s responsibilities. This has naturally been accompanied by a multiplication of posts and personnel. There has been a dramatic expansion of the welfare role of the government. The number of government schools, colleges, and universities and their associated bureaucracies has multiplied. There is a similar trend with health centres, hospitals and medical colleges, and their respective bureaucracies. On a per-​capita basis, the availability of teachers, nurses, and doctors has increased sharply in the last few decades. The developmental state has come to occupy the top priority in governance.

The Size of the Indian State  91 Further, for various political and other considerations, there has been a dramatic expansion of personnel at the lower levels of the government. Some of it has been a natural consequence of the proliferation of administrative units. But, as observed in multiple Pay Commission and Administrative Reforms Commission Reports, this has resulted in a very bottom-​heavy personnel dispensation across central, state, and sub-​state governments. This expansion of administrative and service delivery units has naturally increased access for citizens to public services. It has also scaled down the geographic spread of administrative units, thereby enhancing the ability of government to improve service standards.

6.2  How ‘Bloated’ Is the State: A Reality Check However, the perception of ‘bloated’ government has to be viewed against the reality of the dramatic expansion in the scope of government. From basic regulatory functions, the government today has assumed developmental and economic growth responsibilities. Its interface with the private sector and associated enabling and regulation of markets have expanded significantly. Further, the public expectations on service quality and standards too have risen. This scope expansion has exposed the limitations of government’s reach and bandwidth in the effective management of these responsibilities. This problem manifests acutely at the top positions of the bureaucracy, across levels and functional areas. Incidentally the problem starts with the higher tiers of the administration. While the sanctioned strength of IAS officers has increased continuously, the actual number of officers working has failed to keep up. The retirements of the large cohorts of the seventies and early eighties without commensurate balancing increase in recruitments has been a major contributor. In 2018, the shortfall was 23% (Figure 6.1). This deficiency applies just as much to other services. Against an authorized strength of 4,940, there were 3,972 Indian Police Service (IPS) officers.7 7 Ministry of Home Affairs, Authorised Cadre Strength as on 01.01.2018 (https://​ips.gov.in/​ pdfs/​Cad​rest​reng​th_​3​1082​018.pdf) (Accessed on 22 December 2018).

92  State Capability Challenges and Weaknesses

Number of Officers

6300 5800 5300 4800

Authorised

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

4300

Actual

Figure 6.1  IAS Officers (Cadre)—​authorized vs. actual working Source: Annual Report 2017–​2018, Department of Personnel and Training, Government of India (https://​dopt.gov.in/​sites/​defa​ult/​files/​DOPT-​PG-​AND-​TRAIN​ING-​ANN​UAL-​REP​ ORT-​2017-​18.pdf) (Accessed on 28 December 2018)

As regards the foreign policy establishment8: India Has 1.3 Billion People, But Fewer Than 1,000 Diplomats . . . The nation of 1.3 billion people only deploys around as many diplomats as New Zealand, which has a population of around 5 million . . . With roughly 940 foreign service officers, India has one of the most understaffed diplomatic corps of any major country—​just slightly higher than New Zealand’s 885 officers, or Singapore’s 850. It’s vastly outnumbered by the Japanese and Australian services of around 6,000 people, the estimated 7,500 diplomats of rival China and the U.S. State Department’s service of nearly 14,000 . . . In 2016, the ministry said it was still 140 officers short of its sanctioned strength of 912 diplomats.

Indian foreign policy capability in terms of the number of diplomats is very low compared to peers (Figure 6.2). In fact, as of January 2014, the entire Ministry of External Affairs, at Group A, B, and C levels, had 4,577 posts, of which 34% were vacant.9 The 8 Iain Marlow, ‘India Has 1.3 Billion People, but Fewer Than 1,000 Diplomats’, Bloomberg, 31 August 2018 (https://​www.bloomb​erg.com/​news/​artic​les/​2018-​08-​30/​india-​s-​diplo​mat-​short​ age-​lea​ves-​it-​far-​beh​ind-​asian-​rival-​china) (Accessed on 20 December 2018). 9 Report of the Seventh Central Pay Commission Report, November 2015 (https://​www. fin​min.nic.in/​sites/​defa​ult/​files/​7cpc​_​rep​ort_​eng.pdf?downl​oad=​1) (Accessed on 21 December 2018).

The Size of the Indian State  93 Despite 1.3 billion people, India has barely more diplomats than New Zealand Singapore 850 New Zealand 885 India 900 UK 4.2K Japan 6.2K Australia 5.3K China 7.5K United States 13.9K

Figure 6.2  Cross-​country comparison of strength of diplomatic cores

thinly staffed Indian delegations often struggle to keep up with the massive United States or British or Chinese delegations with their specialized staff and experts in trade and diplomatic negotiations. These trends were even more amplified at the lower levels. In fact, nowhere is the Indian state thinner than at its frontline, or ‘limbs’ as Lant Pritchett so evocatively captured with his description of the ‘flailing state’. At the most basic level, in terms of its ability to functionally engage in a meaningful enough manner, the frontline state is seriously handicapped by chronically deficient manpower. Consider a few examples. In terms of the personnel requirement on law and order management, India’s state capability problem is acute by the standards of the G-​20 economies (Figure 6.3).10

10 Pramit Bhattacharya, ‘Kathua, Unnao Rape Cases: What It Will Take to Get Justice for India’s Daughters?’, Livemint, 17 April 2018 (https://​www.livem​int.com/​Polit​ics/​KtG​Z7f6​zjfT​ gzmY​vcdi​vdL/​Kat​hua-​Unnao-​rape-​cases-​What-​it-​will-​take-​to-​get-​just​ice-​f.html) (Accessed on 12 December 2018).

94  State Capability Challenges and Weaknesses The chart shows the data for the 10 largest emerging markets belonging to the G-20 club Judges per100,000 population

Russia

20 China

Turkey 10

Brazil South Korea Mexico Indonesia India South Africa 100

200

300

400

Argentina Police per 100,000 population 500

600

700

800

900

Data might not be strictly comparable across countries due to difference in definition of police and judges. Data for China (police) refers to 2007, all other data refer to years between 2010 and 2015

Figure 6.3  Personnel strength in law and order management

As against a global average of one policeman per 450 people, India has one for 709 people, with the numbers being 1,298 and 1,282 for Bihar and UP, respectively.11 A report in the Times of India had this summary of the personnel deficiency problem in India’s judicial system12: In 2014, the judge-​population ratio (sanctioned strength) was 17 judges per million . . . The current sanctioned strength of the subordinate judiciary is 20,214 judges while that of the 24 high courts is 1,056 and the pendency of cases has remained abnormally high at 3.10 crore. On the sanctioned strength, there are 4,600 vacancies of judges in the subordinate judiciary which is more than 23% of the strength. The situation in the high courts is worse with almost 44% (462) judges’ posts vacant. The Supreme Court too has six vacancies on a sanctioned strength of 31 . . . The Law Commission study found at the current rate of disposal, HCs require an additional 56 judges to break even and an additional 942 judges to clear the backlog. 11 Devanik Saha, ‘India Short of Half a Million Policemen’, IndiaSpend, 29 December 2014 (https://​arch​ive.ind​iasp​end.com/​viznom​ics/​india-​short-​of-​half-​a-​mill​ion-​police​men-​27070). 12 Pradeep Thakur, ‘India Has 17 Judges for a Million People, 5000 Posts Vacant’, Times of India, 17 April 2016 (https://​times​ofin​dia.ind​iati​mes.com/​india/​India-​has-​17-​jud​ges-​for-​a-​mill​ ion-​peo​ple-​5000-​posts-​vac​ant/​arti​cles​how/​51860​142.cms) (Accessed on 12 December 2018).

The Size of the Indian State  95 Unsurprisingly, nearly 58% of the 21.9 million cases pending in district/​subordinate courts were pending for more than two years.13 In the aftermath of Mumbai’s Kamala Mills fire accident which claimed 14 lives, one report had this to say about India’s fire safety preparedness14: A ministry of home affairs-​sponsored study found that of the minimum 8,559 fire stations needed in the country, only 2,087 are in place, a shortage of 65%. Urban areas alone require an additional 4,200 fire stations just to meet the minimum standard for response time . . . As many as 17,700 Indians died—​48 people every day—​due to fire accidents in 2015, of which 10,925 (62%) were women.

Even where fire stations and equipment are available, they are insufficient and also inadequate to fight fires in high-​rise buildings, etc. The situation is similar across departments. Take the case of planning for cities, the engines of the country’s economic growth. The then Ministry of Urban Development estimated a requirement of 40,000 planners, whereas the actual availability of registered planners was close to 3,000.15 The UNESCO Global Education Monitoring Report 2019 writes about India’s shortage of urban planners,16 In India, only 21 universities offer postgraduate town planning programmes; only 5 offer them at the undergraduate level. as a result, the country had only about 4,500 qualified town and country planners, the 2011 census showed. In the same year, the government estimated needing 300,000 town and country planners by 2031 for city development plans . . . A review of institutions accredited by India’s Institute of town planners found that many planning students lacked exposure to

13 Howindialives.com, ‘The Case of Missing Judges’, Livemint, 26 April 2016 (https://​www. livem​int.com/​Polit​ics/​RKF​cGxy​8QLX​RLgr​FRYW​AoN/​The-​numeri​cal-​defi​cit-​in-​Ind​ian-​cou​ rts-​Chief-​Just​ice-​is-​plea.html) (Accessed on 12 December 2018). 14 Ajai Sreevatsan, ‘Why Kamala Mills Tragedy Is a Warning for Urban India’, Livemint, 2 January 2018 (https://​www.livem​int.com/​Polit​ics/​h3l​Fty4​e0ny​JlND​Q43q​2hI/​Why-​Kam​ala-​ Mills-​fire-​trag​edy-​is-​a-​warn​ing-​for-​urban-​India.html) (Accessed on 12 December 2018). 15 McKinsey Global Institute, India’s Urban Awakening, April 2010 (https://​www.mckin​ sey.com/​featu​red-​insig​hts/​urban​izat​ion/​urban-​awaken​ing-​in-​india) (Accessed on 12 December 2018). 16 Global Education Monitoring Report 2019: Migration, Displacement, and Education, UNESCO (https://​unes​doc.une​sco.org/​ark:/​48223/​pf000​0265​866) (Accessed on 23 December 2018).

96  State Capability Challenges and Weaknesses town and city problems. they were unfamiliar with the functioning of urban development institutions, had limited contact with slums and urban infrastructure projects, and were not up to date on urban development programmes and technical tools, leaving them ill-​prepared to take up planning jobs upon graduation.

India’s pharmaceuticals industry, one of the true success stories, has in recent times faced several cases of punitive actions by the US Food and Drug Administration (FDA) for inadequate safety standards in manufacturing facilities and falsification of test results. This is unsurprising considering that the country’s drugs regulator, Central Drugs Standard Control Organisation (CDSCO), under the Ministry of Health and Family Welfare, which supervised a sector with $14.6 billion of exports17 in 2012–​2013 and over 370 FDA-​approved drug manufacturing facilities, the largest number outside the United States, has a total staff of just 650, a 20th of the US FDA’s 13,000 staff.18 In fact, a Task Force of the Department of Pharmaceuticals in 2015 recommended one drug inspector for 50 manufacturing units and one inspector for every 200 sales and distribution outlets for effectively discharging their regulatory responsibilities.19 This was a reiteration of the finding of the 59th Parliamentary Standing Committee report of May 2012. Given the 600,000 retail sales outlets and 10,500 manufacturing units, this requires 3,200 drug inspectors. But there were only 846 drug inspectors against 1,349 sanctioned posts in the states. The gap has only grown over the years. In 2010, Maharashtra, with 29% of India’s drugs manufacturing units and 38% of its medicines exports, had just 161 drug inspectors, of whom 55% were vacant. The vacancy rate would surge to 83% if we take the same Task Force’s standard and have the inspector requirement to be 433.20 17 ‘India, US Agree on Pharma Regulation’, Livemint, 11 February 2014 (https://​www.livem​int. com/​Indus​try/​HgB​ddBU​pRhN​l44U​yiWa​d3O/​India-​rai​ses-​conce​rns-​over-​FDA-​acti​ons-​on-​ domes​tic-​drug-​firm.html) (Accessed on 12 December 2018). 18 Nivedita Mookerji, ‘Why Do Indian Health Authorities Keep Quiet on Pharma Firms’ Failings?’, Business Standard, 5 December 2014 (https://​www.busin​ess-​stand​ard.com/​arti​cle/​ opin​ion/​why-​do-​ind​ian-​hea​lth-​auth​orit​ies-​keep-​quiet-​on-​pha​rma-​firms-​faili​ngs-​114​1205​ 0039​0_​1.html) (Accessed on 12 December 2018). 19 Department of Pharmaceuticals, Recommendations of the Task Force to Enable Private Sector to Lead the Growth Of Pharmaceutical Industry, 2015 (https://​phar​mace​utic​als.gov.in/​ sites/​defa​ult/​files/​Pharma​task​forc​e_​1.pdf) (Accessed on 22 December 2018). 20 Abhay B. Kadam et al., ‘Correcting India’s Chronic Shortage of Drug Inspectors to Ensure the Production and Distribution of Safe, High-​Quality Medicines’, International Journal of Health Policy and Management, 5, no. 9 (September 2016): 535–​542 (https://​www.ncbi.nlm.nih. gov/​pmc/​artic​les/​PMC​5010​656/​) (Accessed on 23 December 2018).

The Size of the Indian State  97 In early 2014, the US Federal Aviation Authority (FAA) downgraded India’s aviation safety ranking to Category II, citing several safety-​related inadequacies.21 Not surprising since the country’s aviation safety regulator, Director General of Civil Aviation (DGCA), had at the time a whopping 528 vacancies out of a sanctioned strength of 924.22 The task facing the country’s food safety regulatory, Food Safety and Security Standards Authority of India (FSSAI), is no less daunting.23 It had just 356 sanctioned technical posts. Fortunately, the government recently sanctioned an additional 493 posts.24 Hyderabad, where one of us worked, had just three.25 In stark contrast, New York City had 180 food inspectors in 2010.26 Thus the problem is two-​fold: not enough posts created (mainly because of fiscal constraints) and many existing posts remaining unfilled (mainly because of practical issues arising from litigation surrounding promotions, administrative complications, and so on). How does all this stack up in the aggregate? As of 2011, India had 18.5 million governments employees27—​3.4 million with central government, 7.2 million with state governments, 5.8 million with Public Sector Units (PSUs), and 2.1 million with local governments.28 As a share of total employment, of around 400 million, it was just 4.6%.29 However, as 21 Tarun Shukla, ‘FAA Downgrades India’s Aviation Safety Rankings’, Livemint, 1 February 2014 (https://​www.livem​int.com/​Compan​ies/​VBs​85pY​9lO5​OD6n​TSX5​akJ/​US-​dow​ngra​des-​ Ind​ias-​air-​saf​ety-​ranki​ngs.html) (Accessed on 12 December 2018). 22 Firstpost, ‘Government Working Towards Filling up 500 Vacancies in DGCA’, 20 December 2014 (https://​www.firstp​ost.com/​busin​ess/​econ​omy/​govt-​work​ing-​towa​rds-​fill​ing-​up-​500-​ vacanc​ies-​in-​dgca-​553​218.html) (Accessed on 12 December 2018). 23 Amy Kazmin, ‘India’s Food Watchdog Leaves Foreign Groups Wary’, Financial Times, 27 August 2015 (https://​www.ft.com/​cont​ent/​88d39​19e-​4a33-​11e5-​9b5d-​89a02​6fda​5c9) (Accessed on 12 December 2018). 24 PTI, ‘Government Sanctions Nearly 500 Additional Posts for Food Regulator FSSAI’, Times of India, 15 August 2018 (https://​times​ofin​dia.ind​iati​mes.com/​busin​ess/​india-​busin​ess/​govt-​ sancti​ons-​nea​rly-​500-​add​itio​nal-​posts-​for-​food-​regula​tor-​fssai/​arti​cles​how/​65412​787.cms) (Accessed on 23 December 2018). 25 ‘Shortage of Food Safety Inspectors in Hyderabad’, Agro and Food Processing (http://​agro​ nfoo​dpro​cess​ing.com/​short​age-​of-​food-​saf​ety-​ins​pect​ors-​in-​hydera​bad/​) (Accessed on 23 December 2018). 26 Glen Collins, ‘Restaurant Grading Begins in New York’, The New York Times, July 27, 2010 (https://​www.nyti​mes.com/​2010/​07/​28/​nyreg​ion/​28insp​ect.html) (Accessed on 23 December 2018). 27 Rohan Venkataramakrishnan, ‘UPA Government to Earmark 1.74 Lakh Crore for Salaries of Employees’, India Today, 8 September 2013 (https://​www.ind​iato​day.in/​india/​north/​story/​ econ​omy-​limps-​upa-​gov​ernm​ent-​earm​ark-​cro​res-​for-​salar​ies-​of-​its-​employ​ees-​210​272-​2013-​ 09-​08) (Accessed on 12 December 2018). 28 Indian Audit and Accounts Association, ‘Census of Central Government Employees’, 15 August 2012 (https://​auditf​l ag.blogs​pot.com/​2012/​08/​cen​sus-​of-​cent​ral-​govt-​employ​ees.html) (Accessed on 12 December 2018). 29 Planning Commission of India, Estimates of Labour Force, Work Opportunities, and Unemployment in Tenth and Eleventh Plans, Table 107.

98  State Capability Challenges and Weaknesses a share of organized sector employment, it stood at 62.2% in 2010.30 In contrast, in the United States, governments now employed 15.9% of all Americans with jobs.31 Consider this summary of the status32: Data compiled from multiple sources, including a 2008 official survey, Right to Information applications, media reports and the 2011 census show, India has 1,622.8 government servants for every 100,000 residents. In stark contrast, the U.S. has 7,681. The Central government, with 3.1 million employees, thus has 257 serving every 100,000 population, against the U.S. federal government’s 840. This figure dips further if the 1,394,418 people working for the Railways, accounting for 44.81 per cent of the entire Central government workforce, are removed. Then, there are only about 125 central employees serving every 100,000 people. Information technology and communications services account for another 7.25 per cent of the Central government’s staff . . . For the most part though, India’s relatively backward States have low numbers of public servants . . . Bihar has just 457.60 per 100,000, Madhya Pradesh 826.47, Uttar Pradesh has 801.67, Orissa 1,191.97 and Chhattisgarh 1,174.62. This is not to suggest there is a causal link between poverty and low levels of public servants: Gujarat has just 826.47 per 100,000 and Punjab 1,263.34. The data could explain, though, why even well-​off States like these have found it tough to ensure universal primary education and eradicating poverty.

Even these numbers do not convey the full picture,33 Total government sector employment declined by two million to 17.6 million during the period 1995-​2011. Today, close to half or 43 per cent of government employees are temporary and as many as 30 Planning Commission of India, Labour Force and Growth Rate, Data Table 108. 31 Floyd Norris, ‘Shrinking Governments: Good or Bad?’, Economix Blog, The New York Times, 7 March 2014 (https://​econo​mix.blogs.nyti​mes.com/​2014/​03/​07/​shrink​ing-​gove​rnme​ nts-​good-​or-​bad/​) (Accessed on 12 December 2018). 32 Praveen Swami, ‘Figures Bust Myth India’s Bureaucracy Is “Bloated” ’, The Hindu, 30 January 2012 (https://​www.thehi​ndu.com/​news/​natio​nal/​figu​res-​bust-​myth-​ind​ias-​bure​aucr​acy-​is-​ bloa​ted/​art​icle​2843​014.ece) (Accessed on 12 December 2018). 33 Subir Roy, ‘Hire and Fire Needs Safety Net’, Business Standard, 16 June 2015 (https://​www. busin​ess-​stand​ard.com/​arti​cle/​opin​ion/​subir-​roy-​hire-​and-​fire-​needs-​saf​ety-​net-​115​0616​0126​ 0_​1.html) (Accessed on 12 December 2018).

The Size of the Indian State  99 3.5 million government jobs have been outsourced to the private sector during 2000-​12. The government sector accounts for 58 per cent of formal sector jobs, but a good 44 per cent (for the private sector this figure is 75 per cent) of these are temporary in nature.

During the 1984–​2014 period, when the economy grew nearly ten-​fold, the total sanctioned strength of the central government hardly changed.34 As of January 2014, the central government employed 3.302 million people against a sanctioned strength of 4.049 million. Of them, there were just 91,501 Grade A-​level officials. Excluding the 2.694 million people employed by defence, paramilitary, and the railways, the total remaining personnel of the central government was just 0.608 million. As a percentage of GDP, central government employees’ salaries formed just 1.3% in 2012–​2013. In fact, this is what the Seventh Central Pay Commission Report concludes: The total Central Government employment other than those in security related or commercial departments, as on 01.01.2014 is 4.18 lakh. In fact the number of personnel working in the Secretariat of ministries/​departments, after excluding independent/​statutory entities, attached and subordinate offices will add up to less than thirty thousand. The ‘core’ of the government, so to say, is actually very small for the Government of India, taken as a whole.

The Commission finds the total number of central government personnel per million of population in India and the United States to be 1,390 and 6,680, respectively. Chronic manpower deficiency is complemented with similar scarcity of financial and other physical resources. The limited personnel available face numerous resources constraints. Inspectors (or supervisory officials who do field inspections), for example, are by definition field personnel and have to be supported with mobility logistics. Never mind the lack of vehicles, itself a serious constraint, it is common to have inspectors not being reimbursed their travel allowances in time. 34 Report of the Seventh Central Pay Commission Report, November 2015 (https://​ www.fin​min.nic.in/​sites/​defa​ult/​files/​7cpc​_​rep​ort_​eng.pdf?downl​oad=​1) (Accessed on 21 December 2018).

100  State Capability Challenges and Weaknesses Many field offices struggle with insufficient budgets for basic requirements like office stationery. Anecdotal stories of police stations demanding victims bring their paper to write their complaints are not uncommon and highlight the point being made.

This is common across cutting-​edge activities—​school inspectors, outreach nurses, agriculture extension officers, town planning inspectors, engineering supervisors, and so on. It is also common in many states, for a number of reasons, to have important departments like Health and Education without regular district-​level departmental heads and being managed by in-​charge (i.e. acting) officials. The situation is no better

The Size of the Indian State  101 at the state level, with several vacancies of Deputy Directors and Joint Directors across Departments. While the administrative units have proliferated, the structure of the district-​level bureaucracy, administered by the District Collectors, has remained virtually the same for decades. During the same time, there has been a dramatic expansion in the government’s responsibilities—​ development and welfare interventions as well as its market regulation mandate. District government agencies today manage several forms of contracting with private agencies. The scale of resources being transacted at different levels of the district government has risen exponentially. Governments have sought to bridge the gap by hiring consultants and advisors and contracting services. Without fundamental reforms to strengthen district governments, including deploying additional personnel, such measures are likely to remain at best palliatives and more likely to worsen the crisis. This squeeze in hiring manifests in the dire state of affairs in terms of people deployed to perform specific tasks. A recent paper described that ‘local bureaucracies are chronically under-​resourced relative to their responsibilities.’35 Devesh Kapur and Aditya Dasgupta state: Drawing on regressions which look at how bureaucratic resources affect the performance of two major rural development programs, we have provided evidence that: i) inadequate personnel and resources force rural development officers to multi-​task excessively; ii) this inability to specialize has an adverse impact on the performance of development programs. The flip side of this coin is that additional resources have large, positive impacts on public service delivery. The implied returns to additional investments in bureaucratic resources are so large that they clearly indicate that politicians in India are failing to make rational investments in local state capability . . . To cope with staff and physical resource shortages, local officials must spread their finite time thinly across a variety of tasks which they would otherwise delegate. Excessive

35 Aditya Dasgupta and Devesh Kapur, ‘The Political Economy of Bureaucratic Overload: Evidence from Rural Development Officials in India.’ American Political Science Review, 114 no. 4 (2020): 1316–​1334.

102  State Capability Challenges and Weaknesses multi-​tasking tends to crowd out the managerial responsibilities of local officials. In contexts where implementing development programs often requires significant office-​ based planning, management, and form-​filing, this crowding out can slow down the flow of official funds and harm the performance of development programs . . . Politicians and ruling parties are inattentive to investing in local state capability because the electoral returns to these investments are diffuse and uncertain. By contrast, they possess strong incentives to announce and inaugurate new and ambitious rural development programs, as well as paperwork and monitoring requirements which facilitate the claiming of credit for service delivery.

This state of affairs in personnel strength is not exclusive to India and has its resonance in global debates on the issue. For a start, in developing countries, the state has historically been ‘thin’ relative to developed countries.36 The origins of the perception to the contrary can be traced to the New Public Management (NPM) school’s analysis of bureaucracy using the framework of public choice theory.37 The ideology really took wings in the early 80s with the Thatcher–​Reagan governments on both sides of the Atlantic. Reagan famously said, ‘Government is the problem’.38 The NPM literature on public choice theory highlights the importance of pervasive misaligned incentives within public bureaucracies.39 Unlike the private employee who has to ‘perform or perish’, the salary and tenure of career bureaucrats are protected irrespective of outcomes (note this is not necessarily wrong, as explained elsewhere). The resultant distortion of incentives causes bureaucrats to shirk and lower their accountability to deliver on the politician’s quest for outcomes.

36 General Government Final Consumption Expenditure, World Development Indicators, World Bank (https://​data.worldb​ank.org/​indica​tor/​NE.CON.GOVT.ZS?locati​ons=​IN) (Accessed on 10 October 2019). 37 Gernod Gruening, ‘Origin and Theoretical Basis of New Public Management’, International Public Management Journal, 4 (2001): 1–​25 (https://​www.pravo.unizg.hr/​_​downl​oad/​rep​osit​ ory/​4_​1_​O​rigi​n_​an​d_​Th​eore​tica​l_​Ba​sis%5B1%5D.pdf) (Accessed on 23 December 2018). 38 How ‘ “Government” Became a Dirty Word’, npr, 1 September 2012 (https://​www.npr. org/​2012/​09/​01/​160438​753/​how-​gov​ernm​ent-​bec​ame-​a-​dirty-​word) (Accessed on 23 December 2018). 39 William Niskanen, Bureaucracy and Public Economics, Edward Elgar Publishing, 1996.

The Size of the Indian State  103 However, this ideological belief went contrary to several empirical findings on the size of governments in the historical trajectory of development as well as the existence of the so-​called Armey curve, which finds an inverted U-​shape between government size and economic growth.40 But that did not prevent commentators worldwide from embracing the narrative tightly. Post-​liberalization era India was naturally not immune. Thus, the post-​ liberalization narrative that considered government as being ‘bloated’ and ‘inefficient’ has played no small part in exacerbating the state of affairs. The economist Joan Robinson famously said that ‘whatever you can rightly say about India, the opposite is also true.’ So far this section has been trying to argue that the state is not bloated. But the opposite is also true: there are major parts of the state that may in fact be very bloated. The Ordnance Factory Board, set up centuries ago to provide the British Indian army with ordnance, had 98,892 employees while the Indian Railways employed 1.315 million as of 1 April 2014.41 Any benchmarking with other arms manufacturers or railways in the world, in physical or financial terms, would suggest huge over-​staffing. What then is the right conclusion? On the one hand, the pervasive belief that government as a whole is ‘bloated’ and ‘inefficient’ is questionable. In fact, far from being ‘bloated’, the Indian state is, in many areas, paper thin! And given the workload and resource availability, the typical bureaucrat is perhaps more efficient than is being credited. On the other hand, there are indeed parts of the central and state government that have far more people than they need for today. It is this combination that makes it difficult to take a government-​wide approach to staffing levels.

40 K.P.K.S. Lahirushan and W.G.V. Gunasekara, ‘The Impact of Government Expenditure on Economic Growth: A Study of Asian Countries’, International Journal of Humanities and Social Sciences, 9, no. 9 (2015) (https://​waset.org/​publi​cati​ons/​10002​402/​the-​imp​act-​of-​gov​ ernm​ent-​expe​ndit​ure-​on-​econo​mic-​gro​wth-​a-​study-​of-​asian-​countr​ies) (Accessed on 23 December 2018). 41 Seventh Central Pay Commission Report, November 2015 (https://​www.fin​min.nic.in/​ sites/​defa​ult/​files/​7cpc​_​rep​ort_​eng.pdf?downl​oad=​1) (Accessed on 23 December 2018).

104  State Capability Challenges and Weaknesses The Sixth Central Pay Commission Report wrote,42 It is, however, noted that while rightsizing in government is necessary given the changes in work process due to technology and consequent reduction of layers, a blanket ban on filling of vacant posts across the board can impact effective functioning. More flexibility is required in this policy for effective service delivery and care has to be taken that administrative delivery structures do not become hollow or thin in critical areas. It is, therefore, essential that the government revise the Annual Direct Recruitment Plan (ADRP) in terms of which only one third of the vacancies can be filled up. This instruction has resulted in an aging bureaucracy which does not easily adapt itself to technology. An active and younger profile in government employment is the need of the hour. Further, strengthening of the cutting edge for efficient delivery is required. New guidelines, where reduction in manpower and levels of fresh manpower intake would be assessed and prioritized by the individual ministry or department keeping its work processes, service delivery and functional requirements and budgetary savings at centre stage should, therefore, be issued.

The approach suggested by the Commission is conceptually rational. The practical challenge that Personnel and Finance functions in government face is that few departments come forward and volunteer to reduce staff. Even the most inefficient ones claim they need more staff, and the adversarial process of identifying surpluses is quite difficult. Added to this is the fact that government employment is permanent (till retirement

42 Sixth Central Pay Commission Report, March 2008 (https://​doe.gov.in/​sixth-​cpc-​pay-​com​ miss​ion) (Accessed on 23 December 2018).

The Size of the Indian State  105 age) and that people from areas of surplus (e.g. industrial workers in an ordnance factory) often cannot be effectively redeployed to other purposes. It then becomes possible to sympathize with seemingly incorrect solutions (like blanket bans on recruitment or post creation), as they may sometimes be the only practical instrument through which surpluses are gradually weeded out.

6.3  Why Does Much of the State Remain Thin? If much of the state is thin, what prevents governments from expanding their workforce? The answer is that there are many rational reasons from the point of view of both fiscal and personnel management to minimize the staff strength of government. Creating more posts and filling them with government staff has costs—​direct and indirect, financial and non-​financial—​ that far exceed the apparent annual cost of hiring a person.

6.3.1  Litigation on Personnel Management A major contributor to the reluctance to hire has been the role played by the courts in personnel-​related decision-​making. Its impact on cadre management, bureaucratic discipline, and administrative efficiency does not receive the level of attention it deserves. Article 311 of the Constitution of India provides some basic safeguards to protect the independence of public officials. It mandates that an official can be dismissed or removed or reduced in rank only ‘after an inquiry in which he has been informed of the charges against him and given a reasonable opportunity of being heard in respect of those charges’. This provision was intended to protect the civil service from arbitrary, politically motivated punishment.

106  State Capability Challenges and Weaknesses

However, several higher court judgements on disciplinary cases have expanded the standard of justiciability to a much higher level, far beyond the common-​sense interpretation of ‘reasonable opportunity of being heard’. In fact, it may not be incorrect to argue that disciplinary cases have come to assume features of criminal case trials. This has forced its procedures to become elaborate and lengthy and made disciplinary cases very complicated to administer. This, in turn, contributes to making them long-​drawn affairs, which adversely affects both government efficiency as well as the interests of employees. Arguably, employees now have an unreasonable opportunity of being heard and the administration has no reasonable opportunity to impose minimum administrative standards or

The Size of the Indian State  107 discipline! The level of protection bestowed by Courts on public servants appears much higher than in other countries. Judicial intervention goes well beyond disciplinary cases and extends to increments, promotions, pension determination, comparative fairness between people in one stream vis-​à-​vis another, and so on. While judicial intervention to safeguard the independence of government officials is understandable, even if not its current expansive scope, the same cannot be said about its role in other service matters. The consideration of ‘perfect fairness to individual employees’ has come at the cost of fairness to the collective public good. The sympathy for the employee has arguably come in the way of the government being able to get work done by its employees, contributed to employee indiscipline, weakened cadre management, and adversely impacted effective delivery of public services. Further, it has led to all types of service matters getting caught up in litigation for long durations. The cascading effect of the piling up of such litigation on personnel management within government departments has been debilitating. Yet another feature of judicial intervention has been the willingness of courts to intervene with even the procedures of cases involving service matters. Courts have not been averse to entertaining writ petitions challenging even the processes associated with a disciplinary enquiry. For example, it is not uncommon to have courts pass orders staying even the framing of charges or appointment of an enquiry officer. The interim stay orders passed by courts often do not get vacated for years. Originally, writ petitions on personnel cases were being filed in the High Courts and then the Supreme Court. The Administrative Tribunals were then introduced to expedite cases on service matters. Appeals on the decisions of the Tribunals were to be made only to the Supreme Court. However, in 1997 in the L Chandrakumar vs. Union of India and Others case, the Supreme Court reversed its earlier decision in the Sampath Kumar case, struck down this restriction as being violative of the basic structure of the Constitution, and allowed writ petitions against orders of Tribunals to be made to the High Court. This effectively meant that there were three judicial layers instead of two before the tribunals were created. Instead of speeding things up, tribunals now take up additional time. It has become standard practice, rather than an exceptional one, for employees to litigate cases up to higher courts. Invariably, these cases

108  State Capability Challenges and Weaknesses drag on for years, and even if concluded in favour of the government, it would have defeated the purpose of the administrative sanction or decision. Besides, they also contribute to disrupting cadre management and weakening administrative discipline. Further, the ever-​present threat of judicial challenge and falling afoul of the courts also makes officials risk-​averse, politically correct, and opaque in decision-​making. It is now widely accepted that performance assessments, including at the level of AIS officers, have become proforma exercises, with the award of the highest-​grade band being the norm. Assessing officers have realized the costs of honest feedback in terms of the very strong likelihood of their assessments being disputed and even litigated upon. In fact, the vital (and inevitably subjective) personal judgement associated with performance assessments has been so minimized as to make them practically meaningless. Paradoxically, judicial intervention may also have reduced transparency in personnel management issues: transparency and recording of reasons for decisions create the material for disputation. For example, in order to limit the likelihood of a cause or evidence for litigation, officers have become reluctant to offer honest formal opinions in personnel management decisions. In some cases, informal and off-​the-​record feedback sources have become more important. It encourages the adoption of stealthy procedures to manage promotions and postings. Personnel litigation takes up a huge share of the administrative bandwidth. In some ways, cases involving service matters are like a tapeworm inside the administrative system, continually enfeebling it. The growing pile of cases takes up an inordinate time and effort of officials, detracting attention from their substantive activities. It is not uncommon in some of the departments like Education, Health, Commercial Taxes, and Engineering for cases involving service matters to take up an outsized share of the time of the Heads of the Department and Secretaries. Top-​ level officers are frequently summoned to courts. It is often stated (correctly) that government is the largest litigant in the country. But this does not reveal that a large proportion of these cases involve government personnel litigating their personal cases with their employers. The expansive justiciability standard and the liberal entertaining of writ petitions have been critical factors in the pile-​up of cases involving government.

The Size of the Indian State  109 An illustration of this fact comes from the Commercial Taxes Department of the Government of Tamil Nadu. The Department services more than 800,000 taxpayers and has nearly 7,000 employees. However, among the cases pending against the Department in the High Court, there are more cases filed by its own employees than by taxpayers! And one particular litigation on the ratio of promotion between those recruited as Assistant Commissioners and those recruited as Commercial Tax Officers took two decades to resolve, during which promotions were held up for many. In short, the huge administrative burden that falls on senior managers is an important reason why they are reluctant to recruit more people or fill posts—​more staff means more administrative work and more litigation. The diligent manager who tries to fill vacancies is often ‘punished’: he or she faces the burden of litigation but usually does not last long enough in the job to see the posts getting filled—​because notions of fairness to the candidates become the court’s priority rather than service provision to the public. Working with fewer staff may well be rational!

6.3.2  Long Delays in Recruitment A related but distinct issue is the time taken and procedural difficulty of recruiting people. Huge numbers of people typically apply and the requirement of fairness means that selection procedures are elaborate and time consuming. Those who are not selected often go to court alleging unfairness and courts often grant stays of the entire recruitment.

6.3.3  Unpredictable and Hidden Fiscal Costs of Personnel Salaries and benefits of government have often increased unpredictably and unreasonably because of administrative, political, and judicial decisions. Examples are given below: • The traditional defined benefit pension system inherited from the colonial days had a number of safeguards—​a full pension based on

110  State Capability Challenges and Weaknesses 33 years of service with reduction if actual service was less, pension based on last 3 years’ salary, and commutation of a part of the pension and cost of living indexation based on the pension but no other upward revisions of pension. Over the years, the system was gradually replaced by far more generous (and fiscally ruinous) terms through administrative and judicial orders (and those passing the orders were ultimately beneficiaries). Pro-​rata reduction for less than 33 years of service was removed, pension was based on the last day’s salary, commutation was restored after a period (thus giving a double benefit), and instead of mere indexation for inflation, the pension itself was revised every time salaries were revised. This last and most costly change was on the principle (not followed in any other country) that if a Deputy Secretary who retired in 1980 got a basic pension less than one who retired in 2020, then it is a violation of ‘equality before law’. Overall a fiscally sustainable defined benefit pension was converted into a fiscally unsustainable ‘undefined guaranteed increasing benefit’ pension. Eventually the pension system was changed to defined contribution for new civilian entrants after 2004. • The socialistic pattern has meant that the lowest levels in government, which are also numerically the largest, are typically paid wages far higher than the market wage. This is often supplemented by a bonus that was initially supposed to be productively linked. But the productivity link was soon given up and today the ‘bonus’ is simply an additional wage. • Any attempt to create a special lower wage structure for posts that are ‘above market’ is soon dismantled by political or judicial intervention. For instance, Regional Rural Banks were created on the understanding that their staff would have lower salaries than nationalized banks but would not suffer the risk of nationwide transfers. Eventually courts ordered that they get the same benefits—​resulting in a set of staff with the same wages but without the transferability, which (from an efficiency perspective) was a lose-​ lose outcome. • Many judicial decisions have raised pensions and/​or salaries on the grounds of fairness between some group and another, but this has

The Size of the Indian State  111 resulted in fiscal stringency that is arguably unfair to the poor and the unemployed. • As mentioned earlier, many kinds of staff, especially those who work in specialized departments, cannot easily be redeployed to other uses, and if they become redundant, they still have to be carried on the payroll till they retire. As a result the unit cost of employing people in the Government is often much higher than initially appears. This has led to an understandably high reluctance of Finance Departments to create posts.

6.4  What Are Possible Solutions to the Thin State? The need for new recruitments to both fill up the vacancies and for the creation of new posts at certain levels cannot be denied. But given the pervasive inefficiencies and other problems within the existing bureaucracy, this will have to be carefully structured to align incentives and mitigate the concerns. For a start, it may be useful to identify all the posts to be filled. Which of them necessarily demand regular recruitments and which can be contracted or outsourced? Can some of the tasks be shed? Can those functions be rationalized, and posts merged? Is the function itself relevant now given the stage of the country’s development? Are the regulations being administered themselves unnecessary? Can compliances be reduced? Another requirement may be to manage these recruitments cost-​effectively. As multiple Pay Commission reports show, the central and state government bureaucracies have come to be characterized by pay compression at the top and above market wages at the bottom. It may therefore be necessary to reduce the unit costs at the bottom. This may require rationalizing the functions associated with some of the posts. It may also require the creation of newer, lower pay scales and abolition and mergers of certain existing ones. The creation of the post of Multi-​Tasking Staff (MTS) by the 6th Pay Commission and the merger of Group C and D services by the 7th Pay Commission are the right steps in this direction.

112  State Capability Challenges and Weaknesses

This demands a comprehensive re-​examination of the cadres43 (hierarchies of posts) within each functional unit across departments. Many of these cadres were designed for the requirements of a bygone era. Digitization and changes in priorities necessitate shedding many cadres and adding a few new ones. For example, the secondary healthcare institutions in Andhra Pradesh have 364 cadres, which can be radically

43 A ‘cadre’ consists of ‘posts’ at different levels of pay and seniority. For instance, a state may have a ‘Public Health Radiology Cadre’ with posts of Radiology Assistant, Radiologist, Senior Radiologist, and so on. The post of Radiologist might have some people directly recruited to that post, some who are promoted from the post of Radiology Assistant, and so on.

The Size of the Indian State  113 downsized to about 60. The revamp of these cadres should be with a staffing pattern that responds to present-​day requirements. This will help right-​size the government and ensure personnel are deployed to optimize effectiveness. The cadre rationalization and revised staffing patterns of functional units within each Department should be followed with finalization of recruitment channels for each cadre. Those activities that are core to the organization’s functions could be reserved for regular employees. The non-​core or peripheral activities could be contracted or outsourced. This is also important since many posts, especially at the lower levels and on the technical side, have limited career progression avenues which can distort incentives and damage organizational effectiveness. It is appropriate that all such posts be recruited on a contract basis, thereby keeping incentives aligned and also allowing them to seek out opportunities elsewhere once they gather some experience. Another idea is to introduce internships for junior levels of the bureaucracy. The interns can be recruited following a process and then can work with the government for, say, 2–​3 years. The interns can be evaluated for their performance during this period, and the internships of those found inadequate can be terminated. Those performing well can be accorded some preferential treatment in the regular recruitments. This has resonance with the increasing popularity of initiatives like Teach for All.44 Young graduates, especially those with interest in pursuing public service careers, increasingly see them as an opportunity to learn more about the field before making their career choices. For some others this experience can change their perspectives and make them interested in public service careers. Still others view it as a good addition to their resumes—​after all, there are few comparable experiences to that of working in large systems like the government. Besides, such opportunities will allow youth to engage meaningfully with the civil society and the government and make them sensitive about public issues as they start out their respective careers. In short they have the potential to cultivate public-​spirited citizens.



44 Teach for All (https://​teac​hfor​all.org/​) (Accessed on 23 December 2018).

114  State Capability Challenges and Weaknesses If designed well, an internship with government has the potential to attract high-​quality talent to work with governments across sectors, and in the process generating all the aforementioned externalities. This would help governments procure young talent at a lower cost. It could provide an on-​the-​job-​training pathway for regular recruits. It could also expand the pool and quality of talent available for these posts. Besides, it could also bring in fresh ideas and new practices into the bureaucratic system at the lower levels. In this regard, there is the example of the widely acclaimed internship program of the Prime Minister’s Rural Development Fellows (PMRDF). This approach can be adopted for the recruitments of teachers as well as for positions in engineering departments, local governments, and so on. Students who have graduated from college could work in schools or as engineers to both get an experience to burnish their resumes as well as stepping-​stones in their career choices. New graduates from the public policy or urban planning schools could get the opportunity to engage with what they learnt by interning with municipalities. These positive features have to be counter-​balanced with formidable challenges. The process of recruitment or selection will be one of the biggest challenges. A full-​fledged selection process, like with regular recruitments, is clearly not appropriate for interns. But anything less also runs the risk of the process being compromised and the objective not being realized. There is also the ever-​present possibility of writ petitions and judicial intervention. Then there is the issue of terminating the internship. Technically an internship is a temporary arrangement. But given the political economy around such issues, such terminations can, over time, become tricky. (Although in a different context, most regular appointments, state and central, come with a probation period of 1–​2 years; but among the millions recruited over the years the number whose probation has ended in termination on performance considerations is minuscule.) There is also the issue of credible performance assessments of such interns, a necessary requirement given the high stakes associated with a possible performance weightage for interns in regular recruitments. The nature of the work being done also makes credible and quantified assessments very difficult. It is also not difficult to imagine this becoming a race

The Size of the Indian State  115 to the top where every intern appearing for the regular recruitment will have more or less the same performance score. Another option to address the issue of the thin state is to avoid the regular recruitments route and outsource some of these functions. The outsourcing market for many of these services is reasonably deep and can immediately provide the supply. A big advantage of the outsourcing approach would be that it would avoid the perennial risk of involvement by the oversight agencies and courts. This risk can be further mitigated by using service contracts instead of labour contracts and initiating procurement afresh at least every couple of years. But perhaps the most important requirement for addressing the problem of the thin state is the need to change the erroneous perceptions on the role of government as well as about the government’s existing capacities. The Covid-​19 pandemic has led to a welcome re-​assessment of the role of governments45 and re-​established the need for a capable state.

6.5  Recognizing the Reality As the previous sections have attempted to show, it is as inaccurate to say that the Indian bureaucracy is top heavy as it is to say it is bottom heavy. Or overstaffed at the centre and understaffed in the states. The most accurate generalization is that it is understaffed at most levels in most governments. The thin nature of the Indian state presents an interesting conundrum. While its ability to deliver on the desired outcome (effectiveness) is undoubtedly weak, the same cannot be said about its efficiency. Given the available resources and the outcomes realized, the Indian state perhaps is remarkably efficient. The grossly over-​burdened state delivers a reasonably satisfactory level of performance in the aggregate. In terms of value for money, one could perhaps argue that the Indian state delivers bang for the buck. Imagine ensuring a reasonably functional food safety apparatus

45 ‘The Virus Means the Big State Is Back’, The Economist, 21 March 2020 (https://​www. econom​ist.com/​brit​ain/​2020/​03/​21/​the-​virus-​means-​the-​big-​state-​is-​back) (Accessed on 4 April 2020).

116  State Capability Challenges and Weaknesses in Delhi with just three food inspectors or a working urban planning system with just 4,500 town planners for 4,041 notified towns! Ironically, this efficiency of the state surfaces when governments decide to outsource public services. The conventional wisdom would have it that public systems are ineffective and bleed money, and outsourcing and privatization of the same services would save money for the government besides improving the quality of service delivery. But the resource requirements for the effective delivery of services, when captured in outsourcing tenders, are invariably much larger than the prevailing public expenditures. So often one ends up with much higher expenditures when the same activity is outsourced. Partly this may be because of better specification and enforcement of service levels, but, as we shall see in Chapter 15, the outcomes are rarely proportionately better in efficiency terms. This assumes significance since the popular discontent at public service delivery, and support for private service delivery, is as much based on the perception of the lower aggregate cost associated with private providers as with the belief in their relatively greater effectiveness. Both authors have had to manage circumstances where outsourcing and privatization contracts ran into problems due to significantly larger aggregate financial expenditures from the reform. This in turn engenders a perception that the private provider is profiting from public contracts (which often may be so, but for different reasons) by charging excessively; the associated political economy factors have been a major factor in derailing many privatization and outsourcing contracts. Further, it is reasonable to assert that, if faced with anything remotely close to the same work burden and scope, public systems in developed countries or even private systems anywhere would have thrown up their arms. In fact, there are numerous examples of deeply dysfunctional public facilities in developed countries, which, though resource constrained, are far better endowed than the median facility in India.46 None of this is to suggest ‘go ahead and hire people and things will be fine’. Doubtless, there are other contributors to weak state capability. It is 46 Campbell Robertson, ‘An Alabama Prison’s Unrelenting Descent into Violence’, The New York Times, 28 March 2017 (https://​www.nyti​mes.com/​2017/​03/​28/​us/​alab​ama-​pri​son-​ viole​nce.html) (Accessed on 27 December 2018).

The Size of the Indian State  117 not a good use of money to just hire people without structural reforms of the bureaucratic system—​process re-​engineering, delegation of authority, right level of regulatory engagement, use of technology to make transactions more transparent and efficient, measures to contain politicization, etc. But what the facts show is that hiring more people should be right at the top of the agenda!

7 Challenges in Policymaking In Chapter 2, a distinction was made between the two sides of public policy, viz. design and implementation. This chapter explores the policymaking process and the trade-​offs involved in making policies and discusses the weaknesses in India’s policymaking process. It then talks about some critical issues related to policymaking in the Indian context. One is the argument that policymaking requires specialist expertise and generalist bureaucrats are not well equipped in this endeavour. Another is the reluctance of officials to take decisions for fear of post-​facto investigations and prosecution by activist oversight agencies and populist media trials. The chapter concludes with a discussion on issues that are important for the regulatory state.

7.1  Why Does India Underperform in Policymaking? If it is taken as given that India is an under-​performer in policymaking, why is this the case1? A priori, under-​performance vis-​a-​vis potential could be due to • adopting the wrong public policies • poorly implementing the right public policies. There can, of course, be valid disagreements as to what is the ‘right’ policy in a given sector, in a given situation. It can be argued that merely because

1 This section draws substantially from O.P. Agarwal and T.V. Somanathan, Public Policy Making in India: Issues and Remedies, New Delhi: Centre for Policy Research Working Paper, 2005. State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0007

120  State Capability Challenges and Weaknesses there are errors, changes, or postponements in policies, one cannot conclude that policymaking suffers from weakness. Success is often the result of trial and error. Disagreements, often strong ones, are common and, in a democratic society, both inevitable and healthy. Vigorous debate prior to policymaking and adaptation in response to debate is good, not bad. Flexibility in policymaking to respond to evolving exogenous factors is good, not bad. And the phenomenon of political considerations intervening in decisions otherwise well taken is inevitable in a fractious but genuinely democratic polity like India. A survey of some recent and not-​so-​recent examples of policymaking in India suggests, however, that there may indeed be something wrong with the policymaking process. In most of the cases surveyed: • debate has occurred after policymaking, instead of before • views of one or other important party affected by a decision seem to have not been adequately considered or canvassed before policy was made • considered decisions on relatively apolitical issues have been reversed at the last minute even where no new information or circumstances have arisen • factors that were endogenous to the problem, that were known or could have been foreseen while making policy, appear to have not been anticipated or considered. These features are symptomatic of a poor policymaking process and in particular of ‘executive policy unreliability’.2 Politics is not the prime reason for delays or changes in most cases. Even where politics appears to be the reason, there is often more to it. While it is quite possible for purely political considerations to derail a well-​structured policymaking process, in many cases it is weaknesses in the policymaking process that are exacerbated political interference. Indeed political ‘interference’ is often (though by no means always) just a manifestation of factors ignored or missed in the policymaking process. 2 G. Evan and N. Manning, Helping Governments Keep Their Promises: Making Ministers and Governments More Reliable Through Improved Policy Management, South Asia Working Paper, Washington: World Bank, 2003.

Challenges in Policymaking  121 Good policymaking structures and processes do matter and can overcome political bickering, as apparent from the evidence of other countries, which are democracies. To cite one instance, a good part of the credit for the post-​1945 Japanese economic miracle goes to the processes which enabled Japan to come out with coherent and well-​implemented policies.3 This chapter is not concerned with what is the ‘right’ policy in any sector. Rather it asks a more basic question: Why do India’s policymaking structures have so much difficulty in formulating the “right” policy and then sticking to it? It goes on to ask, and attempt to answer, the question: What can be done to improve the structures involved in the making of public policy in India? It should be noted that there is no attempt here to define an ‘ideal’ process. The field of public policymaking is not one that lends itself easily to such a proposition, without normative rather than positive judgements; any such normatively defined ‘ideal’ would have the infirmity that what is ideal with one set of norms would be sub-​optimal from a different normative perspective.

7.2  Good Policymaking Process: A Conceptual Overview One way of describing a ‘good’ policymaking process is one that ‘is committed to producing a high-​quality decision—​not any particular decision’ and that ‘invests any decision made with a high degree of legitimacy, power and accuracy.’4 What features or characteristics should a policymaking process have which, if present, would lead to high-​quality decisions? First, to start with the most obvious, a good policymaking process would involve due consideration of up-​to-​date available subject-​matter knowledge and relevant data and the use of available analytical tools. 3 Economic & Social Commission for Asia & The Pacific, The Lessons of East /​South-​East Asian Growth Experience, Development Paper No. 17, ESCAP, Bangkok, 1995, pp. 92–​93. 4 Mark H. Moore, Creating Public Value—​ Strategic Management in Government, Cambridge: Harvard University Press, 1998.

122  State Capability Challenges and Weaknesses Second, policies made ostensibly for one sector often have significant impacts on other sectors: a transport policy (e.g. expansion of national highways in lieu of investment in rail) affects the environment; an environmental policy (stricter pollution norms) affects industrial development; and a revenue enhancement measure intended to develop one sector can adversely affect another (e.g. a cess to fund the National Highway Development Project may reduce the competitiveness of road transport). Policymaking therefore nearly always means trade-​offs, the giving up of something to get something else, losses to one group or section in exchange for (hopefully larger) gains for another. Policymaking processes and structures should ensure the gathering of information on such inter-​sectoral impacts, the analysis of trade-​offs, and fully informed choices between alternatives after a proper consideration of effects on different sectors. Many analytical techniques have been evolved to assist policymakers in dealing with these issues, coming broadly under terms like policy analysis, program evaluation, and cost–​benefit analysis. These techniques are not without their critics, and their effect on policymaking has been less than their protagonists would like to think.5,6 Nevertheless, these techniques are generally judged to have a positive effect on the quality of decisions made.7 Third, especially in a democratic polity, such analysis should invariably include an assessment of the “winners” and “losers” from a given policy and a strategy for dealing with likely opposition from losers to what has been determined to be the “right” policy. Fourth, theory and practice both show that decisions which are seen to have ‘legitimacy’ are far more likely to be successfully implemented.8 Legitimacy is both procedural and substantive. • Procedural legitimacy is sometimes narrowly viewed as meaning that the decision is made by an authority legally authorized to make 5 Charles E. Lindblom, Inquiry and Change: The Troubled Attempt to Understand and Shape Society, New Haven: Yale University Press, 1990. 6 Laurence E. Lynn, Knowledge and Power—​The Uncertain Connection, National Academies Press, Washington DC, 1978. 7 Laurence E. Lynn, ‘Policy Analysis in the Bureaucracy: How New? How Effective?’, Journal of Policy Analysis and Management, 8, no. 3, pp 373–​377(1989). 8 Bernardo Kliksberg, ‘Six Unconventional Theses About Participation’, International Review of Administrative Sciences, 66, no. 1, pp 161-​174 (March 2000).

Challenges in Policymaking  123 it, but in practice consultation of those affected is crucial to perceived legitimacy. Procedural legitimacy can often be more important in securing the implementation of a policy than its substantive merits. • Substantive legitimacy is achieved when the persons and groups who have knowledge and expertise in the field affected by a policy are involved in formulating the policy.9 Note that this point is about the legitimacy—​not efficacy—​of a policy. The question is not whether the policy was substantively correct but whether persons who are publicly known or perceived to have subject matter knowledge were involved in making it. Fifth, a good policymaking process should produce policies that can be executed swiftly and successfully. This requires the close involvement, during formulation, of the persons who actually have to implement a policy on the ground and implies a degree of ‘decentralisation’ of policymaking.10 At the same time, a degree of centralized control is necessary so that the priorities and interests of implementers do not supplant the public interest. Whether this central control should be confined to ‘process control’ (i.e. control over how the decision is made) or should extend to ‘quality control’(control over the substance of the decision) is the subject of debate, but the choice is partly a factor of the kind of organization and the kind of policy being made.11 On the whole, while policymaking must remain in touch with reality and be conscious of implementation issues, it should not be a prisoner of the current short-​term priorities, time constraints, and conveniences of implementers. A good policymaking structure should, therefore, provide for appropriate separation between the policy and implementation functions. Finally, in order to make the (often difficult) decisions on trade-​offs and make them without undue delay, information, analysis, and good procedures alone are insufficient. Those charged with making, or advising on, policy must possess certain skills (e.g. in coordination, synthesis, and 9 Mark H. Moore, Creating Public Value—​ Strategic Management in Government, Cambridge: Harvard University Press, 1998, pp. 126–​128. 10 Richard Darman, ‘Note on Policy Development No. 4’, cited in Moore, 1998. 11 Roger Porter, Presidential Decision Making: The Economic Policy Board, Cambridge: Cambridge University Press, 1980.

124  State Capability Challenges and Weaknesses integration) and attributes (such as freedom from bias) that increase the likelihood of quick and sound decisions. To recapitulate, a ‘good policymaking process’ would meet the following criteria:



i. the problems and issues confronting a sector are subjected to expert analysis; ii. information on overlaps and trade-​offs with other sectors is systematically gathered and made available to policymakers; iii. opposing points of view within and between sectors, are properly articulated, analysed, and considered; iv. those likely to benefited or harmed are identified and their reactions anticipated; v. decisions are made with due legal authority, after consultation of those likely to be affected, and with the involvement of knowledgeable persons in the sector(s) concerned; vi. those responsible for implementation are systematically involved in the process but are not allowed to take control of it; and vii. policymakers and/​or their advisers have the honesty, independence, intellectual breadth, and depth to properly consider and integrate multiple perspectives and help arrive at optimal policy choices within a reasonable time.

7.3  Translating Theory into Practice—​Managing the Trade-​Offs Unfortunately the application of these theoretical principles in designing a real-​world structure is not simple. There are trade-​offs. Criterion (i)—​ expert analysis of a given sector—​is usually achieved by specialists in a field. The pursuit of specialized expertise often leads, quite logically, to fragmentation, i.e. the creation of more and more specialized organizations—​ministries, departments, directorates, etc. The narrower the specialization, the greater the potential for depth in knowledge of that field. For example, instead of one Department of Science and Technology, one can have separate departments for Space, Ocean Development, etc.

Challenges in Policymaking  125 Instead of an Education Department one can divide it into Primary Education, Secondary Education, Collegiate Education, Technical Education, etc. In the years since Independence, the Central Government has created an ever-​increasing number of more-​specialized departments in place of more-​generalized departments. However, narrow specialization diminishes knowledge of the larger picture, of overlaps and trade-​offs. Thus, excessive pursuit of criterion (i) reduces the attainment of criterion (ii). Improved analysis may come at the cost of reduced synthesis—​a weakness as prevalent in the private sector as in the public sector.12 Besides, specialists in a real-​world bureaucracy begin to acquire an interest in the pursuit of their specialism or ministry—​expansion of that sector means more departments and hence more top jobs, faster promotion, greater responsibility, more prospects of public recognition, etc.—​thereby diminishing their independence and thus attainment of criteria (iii) (consideration of opposing points of view) and (vii) (independence and lack of bias). Besides, while fragmentation improves specialized knowledge, it • reduces communications between the fragmented units, both formal and informal, and • reduces coordination and integration. Integration of different functions is intrinsically difficult and costly.13 Officers in the same department interact frequently at meetings (formal communication) and may even meet daily for lunch (informal communication). Disagreements between them may be quickly resolved by referral to their common superior. Officers in different departments interact less frequently and more formally, reducing the quantity and quality of information-​ and idea-​sharing. Thus as a corollary of the preceding criteria, a good policymaking structure must neither be so wide as to militate against specialization,

12 Russell L. Ackoff, Re-​Creating the Corporation—​A Design of Organisations for the 21st Century, New York: Oxford University Press, 1999. 13 Robert Chambers, Managing Rural Development, Uppsala: Scandinavian Institute of African Studies, 1974.

126  State Capability Challenges and Weaknesses nor be so fragmented as to affect integration.14 What then is the ‘right’ or ‘optimal’ degree of fragmentation? In the following section, two general principles are suggested. While policy in any sector can theoretically affect any other, in practice the number of interconnections is greatest among related sectors. Thus the interconnections or trade-​offs between road transport and rail transport are greater than between road transport and space technology, while the interconnections between information technology and rail transport are less than between, say, information technology and telecommunications. This leads to the conclusion that: As a general principle, related sectors (meaning sectors with significant policy interactions between them) should be grouped together so as to maximise policy coordination. The lower down the hierarchy one operates, the greater the value of specialized knowledge. Thus the sanitary engineer operating a sewage pumping station needs very specific knowledge about the working of his pumps, a level of detail which his utility’s chief executive does not need to know. By contrast, the chief executive has to have a basic level of awareness of every facet of the utility’s operations. A heart surgeon in a teaching hospital needs highly specialized knowledge—​but the Director of Medical Education needs a very different set of skills and information. The corollary is that, As a general principle, fragmentation needs to diminish as one goes higher up the hierarchy. Finally at the apex, namely the Prime Minister, one person becomes responsible for everything.

7.4  Weaknesses in India’s Public Policymaking A comparison of the reality of policymaking in India with the theoretical framework outlined in the preceding section shows the following shortcomings. 14 Robert Klitgaard, ‘Integrating Public Services—​ a Framework for Policy Analysis’, pp. 158–​159 in Adjusting to Reality—​Beyond State versus Market in Economic Development, San Francisco: International Centre for Economic Growth, 1991.

Challenges in Policymaking  127

7.4.1  Excessive Fragmentation in Thinking and Action One of the main problems with policymaking in India is extreme fragmentation in the structure. For example, the transport sector is dealt with by five departments/​Ministries in the government of India, whereas in the United States and the United Kingdom it is a part of one department (Department of Transport and Public Works in the United States and Department of Environment, Transport and Regions in the United Kingdom). Similar examples exist in the energy, industry, and social welfare sectors as well. Such fragmentation fails to recognize that actions taken in one sector have serious implications on another and may work at cross purposes with the policies of the other sector. Besides, it becomes very difficult, even for closely related sectors, to align their policies in accordance with a common overall agenda.

7.4.2  Excessive Overlap Between Policymaking and Implementation Another problem is the excessive overlap between implementation, program formulation, and policymaking which creates a tendency to focus on operational convenience rather than on public needs. Policymaking in Indian ministries occurs at the levels of Director and above, but the most important level (crucial for consideration of cross-​ cutting impacts) is that of the Secretaries to the Government of India, who are their Ministers’ ‘policy advisers-​in-​chief ’. However, as mentioned earlier, the very same Secretaries spend a large part of their time bogged down on routine day-​to-​day administration of existing policy. Time is spent anticipating and answering parliamentary questions, attending meetings and functions on implementation issues, etc. Partly the problem is symptomatic of over-​centralization—​excessive concentration of implementation powers at the higher levels of the Ministries. Partly, it is also due to such officers being more comfortable with implementation matters than with policymaking. The result is that sub-​ optimal policies, where adequate attention has not been paid to citizen needs, tend to emerge.

128  State Capability Challenges and Weaknesses Figure 7.1 attempts to depict both, the fragmented policymaking structures in India and the low degree of separation between policymaking and implementation. In Figure 7.1, points nearer the centre of the circle represent higher levels of authority, the centre of the circle being the Cabinet. The black radii represent departmental divisions and their thickness denotes the relative lack of communication between departments. The two thin concentric circles denote the separation between implementation and policymaking levels. They have been depicted as thin circles (as opposed to the thick radii) to signify that there is little effective separation between policy and implementation. The principle of gradual diminution in fragmentation as one goes up the hierarchy is not followed, a clear indicator that fragmentation at policy levels is excessive. (This is in contrast to the early years after Independence, when there were far fewer Secretaries.) Fragmentation has often occurred for reasons not directly connected with the design of an optimal structure. Indeed there is a widespread belief that fragmentation was driven more by the compulsion to accommodate a larger council of ministers, as well as the bureaucratic desire for more top-​ level posts. (In Chapter 10, Figure 10.1, an alternative structure is suggested.)

Current structure Cabinet Secretary levels Upto JS Levels

Implementation Levels

Figure 7.1  Policy making structure in GOI (Source: T.V. Somanathan and O.P. Agarwal, Public Policy Making in India: Issues and Remedies, Research Gate, February 2005)

Challenges in Policymaking  129

7.4.3  Inadequate Non-​Governmental Inputs and Informed Debate Often public policy has been made without adequate input from outside government and without adequate debate on the issues involved. The best expertise in many sectors lies outside the government. Yet the policy processes and structures of government have no systematic means for obtaining outside inputs, for involving those affected by policies, or for debating alternatives and their impacts on different groups. Such debates not only enable an assessment of different viewpoints but also help build up a constituency in support of the policy through sound arguments. For long, the only example of fairly systematic consultation of outside expertise in India was in the process of formulating the Central Budget, where there is a long tradition of pre-​budget confabulations with chosen members of industry, labour, and academia. (In recent years, however, the MyGov portal, which puts out consultative papers for feedback, has led to an improvement in this regard.) There are several reasons for a poor pre-​policy consultative process. Firstly, structures for consulting outsiders either do not exist or, if they do, are weak. Secondly, in the absence of good consultative structures, outsiders who do make themselves heard in the policymaking process are often single-​issue advocates. This makes them liable to the charge of having vested interests, and their views lose credibility. Even if a receptive civil servant were to take their views seriously, he would run the risk of appearing to do an illegitimate favour. Thirdly, outsiders involved in policy are usually allowed to make spasmodic or single-​issue inputs but are not required to sustain their interaction, to confront trade-​offs, or to meet the objections of other outsiders with opposite views. This makes it easy for outsiders who were indeed consulted to then disclaim any responsibility for the final decision by protesting that their advice was only partially followed. Fourthly and as a result of the first three, there is a lack of identification of stakeholders with any policy. In countries like the United States, there are often strong advocates on both sides of a policy question—​for example, pro-​and anti-​abortion and pro-​and anti-​capital punishment. In India, judging by the public reaction to many policy announcements, it would appear that almost every new policy announced by government has ‘only opponents’. This is because the ‘winners’ from

130  State Capability Challenges and Weaknesses a government policy rarely feel involved in it, and hence rarely stand up and support it. (To be fair, consultation in India is not easy. And like most ‘remedies’, it has costs as well as benefits. Indeed, some laws passed after extensive discussion in Parliamentary committees have subsequently been felt to have serious flaws—​the Companies Act 2013, for instance. It is also true that for some contentious and highly politicized policy changes, announcing a change unilaterally without consultation can have some practical advantages by changing the starting point of the ‘game’ and improving the final outcome.)

7.4.4  Lack of Systematic Analysis and Integration Prior to Policymaking Policy decisions have often been made without adequate analysis of costs, benefits, trade-​offs, and consequences. There are several underlying causes for this: 1. Excessive fragmentation: This has already been referred to. Fragmentation has led to a widespread prevalence of the ‘blind men and the elephant’ syndrome in policymaking. 2. Inadequate time spent on policymaking, mainly due to excessive overlap of policymaking and implementation and to over-​centralization of implementation authority (discussed above). 3. Inadequate professionalism of policymakers and advisers: Debates have been common in India about the pros and cons of ‘generalists’ vs. ‘specialists’ in government. There is a school of thought which suggests that the excessive involvement of poorly informed generalists is the main cause of poor policymaking and implementation. However, when it comes to the realm of policymaking and the making of trade-​offs, experience in government and the private sector suggests that this is usually best handled by an intelligent, well-​informed person who has a wide rather than narrow perspective. This person could be termed the ‘intelligent and informed generalist’ who, though not a specialist in any one field, is in fact a specialist in analysis, integration, and synthesis—​i.e.

Challenges in Policymaking  131 identifying problems, trade-​ offs, and solutions. His strength and training lie in being well informed about a variety of related subjects, in incisive analysis, and in the intelligent use of information provided by specialists to frame policy options and assess their consequences. Note that many successful businesses in India and abroad are headed by generalists (Master’s in Business Administration [MBA] graduates, for instance), and the Tata conglomerate continues to operate through the generalist ‘Tata Administrative Service’ to man key positions—​an approach regarded as a great success.15 The problem currently encountered is that the civil servants (who act as key policy advisers) often are not sufficiently well informed or trained to act in this manner. This could be described loosely but conveniently as ‘inadequate professionalism’. 4. Inadequate consultation of in-​house specialists: Even conceding that public policymaking might not be improved by insisting on specialists becoming the policymakers, it is nevertheless crucial that specialist knowledge be fully consulted and utilized in arriving at policy. For reasons ranging from ‘generalist arrogance’ to inter-​service rivalries between groups of specialists, the available expertise of specialists within the government is often under-​utilized. 5. Mediocrity of in-​house specialists: While there are many outstanding specialists working for the government, there is a widespread feeling that many in-​house specialists are not on top of their specialisms. This perception of mediocrity vis-​à-​vis outside experts tends to worsen the problem of inadequate consultation of even the good in-​house specialists who get tarred with the same brush. It also promotes undue respect for outside specialists and the error of accepting poorly formulated prescriptions from outsiders simply because they have a more professional or expert image.

15 Arun Maira, The learning factory: How the leaders of Tata became nation builders, Viking, 26 October 2020.

132  State Capability Challenges and Weaknesses

7.5  Specialist vs. Generalist—​or Specialization of the Generalist A reasonable argument can be made that in the modern age, with increased technological sophistication and a knowledge explosion, generalist services are an anachronism. The logical conclusion, if this view is accepted, is to move • Either: to a position-​based system that prevails in countries like the United States and organizations like the World Bank. Staff are hired against specific positions based on qualifications for the particular post. (In the United States, most senior-​level positions are also political appointments.) This would change not only the IAS but also the other Central Services, where, for instance, chartered accountants would form the Indian Audit and Accounts Service, accountants and lawyers would form the Revenue Service, architects would man the Defence Estates service, and so on. • Or: to a ministry-​based system, where generalists start specializing in a ministry or function as soon as they are recruited, as in many countries with mandarin systems. This is a variant of the idea of greater specialization. The key difference is that specialization is by experience, not by qualification. Most Central Civil Services are generalists at the point of recruitment, followed by specialization through experience and thus offer an example of the possible strengths and weaknesses of such an approach. In both cases, this may be accompanied by a generalist Senior Executive Service at higher levels. For leadership positions involving coordination across sectors and for policymaking, the question is complex. Subject matter knowledge is obviously important. The Central Civil Services officers do possess, on average, better subject matter knowledge of their own specialism than a typical IAS officer who enters a new ministry. Thus, a Commissioner of Income Tax is likely to be more knowledgeable on income tax than the typical IAS officer of equal seniority on the subject to which he is assigned. The weakness of the IAS vis-​à-​vis services, which specialize by experience, is lack of contextual knowledge and of specialized skills;

Challenges in Policymaking  133 vis-​à-​vis truly specialized services or a position-​based system, the difference in specialized knowledge is even greater. Yet, there is little evidence that, taken as a whole, say, the Indian Revenue Service (IRS) is more professionally competent in their domain or has higher integrity than the IAS. (The scandal involving a former Union Minister of State for Revenue in 2003 who resigned after his Personal Assistant was arrested for taking bribes from IRS officers highlighted the extent of corruption in, and politicization of, postings in the IRS).16 Taking less corruption-​prone functions than taxation, there is no evidence that the Postal Department or Defence Estates or Ordnance Factories or even Indian embassies are significantly better administered than the average central ministry. As regards the real ‘specialist’ services—​the Forest Service, Economic Service, Statistical Service—​there is little evidence that they have higher levels of coordination or policymaking competence. Selections to the post of, say, Chief Economic Adviser—​where IES officers have had the opportunity to compete through an open and competitive UPSC process—​have rarely gone to officers of that service and rather have gone to outside experts. Looking at practices in the civil services of developed countries with a parliamentary system, the tendency is actually in the opposite direction: the United Kingdom and Australia have both moved from greater specialization to greater generalization in recent years and have unwound some of the changes made earlier. The United Kingdom is consciously requiring senior civil servants to acquire wider experience than in a single sector.17 Australia is moving in the same direction.18 Singapore’s civil service—​generally regarded as both honest and effective—​is a generalist service. Even in the private sector, managerial talent is highly fungible across disciplines, to a greater extent than is generally appreciated. The MBA is a degree in (business) ‘administration’ and MBAs are quintessentially generalists. The Tata Administrative Services and the World Bank’s and

16 ‘Gingee Ramachandran Resigns from Ministry’, Rediff, 23 May 2003. 17 HM Government, Civil Service Reform Plan. 18 PricewaterhouseCoopers, ‘Review of the Senior Executive Service: Report to the Special Minister of State for the Public Service and Integrity’, Australian Public Service Commission, 2011. Available at http://​www.apsc. gov.au/​ data/​assets/​pdf_​file/​0004/​48559/​reviewofses.pdf (Accessed on 22 December 2018).

134  State Capability Challenges and Weaknesses Asian Development Bank’s Young Professional Programs (YPP) are non-​governmental examples of managerial cadres which move across the organization. The Chief Executive Officers of all three major US car companies recruited after the crisis in 2008 and credited with rescuing the companies were rank generalists: none of them had previous experience in the automobile industry. One was a former telecom executive, another was from civil aviation, and the third was from a customs/​trade inspection company. Two other recent Detroit chief executives were also generalists.19 PepsiCo India hired a chief executive from a mobile phone company.20 Agarwal and Somanathan studied this issue in detail, looking at practices across sectors, and concluded that an ‘intelligent generalist’ (as specifically defined—​not necessarily an IAS officer) provided in most situations the best bet. No doubt, they concluded, that currently many civil servants are often ‘not sufficiently well informed or trained to act in this manner’.21 Apart from competence, there are other issues that need consideration. First, a position-​based service is very likely to increase politicization by enabling the appointment of chosen persons to specific positions. Second, both position-​based and (even more so) ministry-​based services have stronger affiliations to their given functions or ‘turf ’ than the IAS, and ceteris paribus, this would tend to have a negative impact on inter-​ sectoral collaboration and coordination.22 Also, the selection of officers for cross-​sectoral coordination posts would be more difficult. Third, a ministry-​based service would lack the multi-​sectoral experience and perspective that the present system has and that a position-​based system could bring (if the right people were chosen for coordinating posts). Fourth, there is the vital issue of the all-​India (federal) dimension and

19 ‘Three Outsiders, Three Styles’, The Economist, 19 January 2013. 20 M. Dalal, S. Tandon, and S. Agarwal, ‘Leadership Change: New Chief Executives to Steer Consumer Goods Firms in 2014’, Mint, 3 January 2014. 21 O.P. Agarwal and T.V. Somanathan, ‘Public Policy-​Making in India—​Issues and Remedies’, Working Paper. New Delhi: Centre for Policy Research, 2005. [Emphasis in the original.] 22 For instance, disagreements between the Departments of Industrial Policy and Promotion and Environment are likely to be even more difficult to resolve if both of them were led by specialists with a strong stake in their respective fields.

Challenges in Policymaking  135 coordination between three tiers of government. Many proposals for civil service reform ignore this aspect. It is practically impossible to retain the common all-​India character of the civil service along with either a position-​based or a ministry-​based system. The absence of common training, common ethos, and long service at different levels would definitely militate against neutral preservation of the constitutional quasi-​ federal order, against coordination between the centre and states, and against the modicum of uniformity in administration that the country now has. Table 7.1 summarizes the strengths and weaknesses of the three approaches. Overall, there can be no easy or a priori conclusion in favour of domain specialization as the predominant method of organization for a federal (all-​India) service. Suffice it to say that the evidence is too equivocal to warrant this being a major plank of reform. However, measures to promote a degree of specialization in the second half of IAS officers’ careers are likely to have a positive effect while retaining the advantages of the current structure. Table 7.1  Strengths and Weaknesses of Different Civil Service Structures in India Rating Under Characteristic

Current system

Ministry-​based system (generalist Position-​based recruitment, system specialization by experience)

Subject matter knowledge

Low

Medium

High

Political neutrality of recruitment

High

High

Low

Multi-​sectoral experience and perspective

High

Low

High

Collaboration/​ coordination across functional boundaries

High

Low

Medium

All-​India (federal) character and coordination among tiers of government

High

Low

Low

8 Decision Paralysis An issue that has emerged as an important constraint on state capability is the perceived reluctance of officials to take decisions for fear of being at the receiving end of what is perceived by them as activism by oversight agencies. The problem is more serious in relation to policymaking, though it also affects implementation.

8.1  The Problem The issue of ‘decision paralysis’ can be traced back to the trends on transparency and accountability that swept public affairs in India in the first decade of this century. Led by the Right to Information (RTI) Act 2005, crusading information activists, auditors, vigilance commissions, investigators, and judges have sought to shine a light on political corruption and bureaucratic complicity and cleanse public life. This, coupled with the explosive growth of 24/​7 electronic news media channels and ubiquitous social media, has shifted the benchmarks on transparency and accountability. While largely welcome, this transformation has not been without its unintended negative effects. The judgements and actions of oversight agencies, most often with the benefit of hindsight, have always been a matter of concern among officials. However, this was, on balance, not bad since it contributed to keeping the system honest. It also helped that the oversight agencies too were incentivized to exercise restraint. A very tenuous balance kept the system right. Once the restraint gave way in the face of activism, the concerns became salient. And it has come to pervade the entire administrative system. Ironically, the problem has largely been due to well-​intentioned actions

State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0008

138  State Capability Challenges and Weaknesses by strong and vibrant institutions, making their satisfactory resolution even more difficult. The cumulative effect of the post-​2000s trend has been to shrink the decision-​making space available for public servants, including political representatives. It has contributed to a reluctance among officials to exercise judgement and take decisions, especially on important but potentially contentious issues: A decision is taken. It is ‘exposed’ as a scandal. Investigations are initiated. In parallel, a public trial begins—​speculation of conspiracy theories abounds, guilt is charged, and media defamation and public humiliation follows. It may be that the ‘guilty’ are finally exonerated by the investigators or courts. But the damage to reputation is already done, irreparably so. Wary of this, officials have been loath to objectively examine issues and take decisions on merits that can potentially benefit private parties, even if such benefit is legitimate and not against public interest. They have preferred to take cover behind mechanistic solutions like auctions or to let courts decide. Given the very high stakes involved and the central role of private participants, infrastructure, and finance ministries are worst affected. The result is a ‘decision paralysis’ that is perceived to have gripped the higher levels of the country’s bureaucracy, both at the state and central government levels. This problem is independent of the government in power and is not easily addressed through executive or legislative actions. At its worst, it may have left indelible scars on a generation of bureaucrats and politicians.

Decision Paralysis  139

In the context of the large pile of non-​performing assets (NPAs) of the country’s banks and the struggles of the public sector–​dominated banking system to resolve them in time by selling it off to the likes of asset reconstruction companies (ARCs), this gets to the heart of the challenge1: Public sector banks are scared to sell to private ARCs for fear that the quantum of hair cut can always be questioned by the government’s auditor, vigilance or at worse be probed by the intelligence agencies.

1 P. Vaidyanathan Iyer, ‘Government Plans to Defuse Ticking Bank Bomb’, Indian Express, 1 February 2016 (https://​indian​expr​ess.com/​arti​cle/​busin​ess/​econ​omy/​govt-​plans-​to-​def​use-​tick​ ing-​bank-​bomb/​) (Accessed on 21 November 2018).

140  State Capability Challenges and Weaknesses Even apex regulatory authorities are not immune from decision paralysis. Some have suggested that an example was the request of the Reserve Bank of India (RBI) to the government to promulgate an ordinance to amend the Banking Regulation Act 1949. This ordinance introduced explicit enabling provisions to allow the RBI to issue specific directions to banks to resolve their stressed assets.2 It is not obvious that there was a need to issue this, given the broad sweep of supervisory powers already available under Section 35A. Indeed, the provision in the new Section 35AB(2) can be interpreted as a subset of possibilities under the already existing Section 35A. In fact, over the last three years, the RBI has issued several different policy directions under the existing Section 35A to resolve bad assets, which conveys its broad sweep of powers in this regard.3 And the actions of the Committees to be established under Section 35AB(2) are advisory in nature, leaving decision-​making power still with the individual banks. There was a view that the RBI insisted on the ordinance as a layer of comfort for the RBI to pursue a more active asset disposal agenda without raising the question of why that was not pursued earlier. This is very relevant in a bureaucratic environment where every decision is likely to be subjected to post-​facto scrutiny, often many years later and completely divorced of its context. Consider the case where an ARC makes windfall gains from one of its assets. A malicious complaint can trigger an enquiry. An auditor, without sufficient appreciation of the decision-​making context and swayed by the information which emerges over time, can attribute presumptive loss and fault the decision by the bank management to take haircuts. An investigating agency without sufficient professional expertise to assess and investigate the issue could find fault with the magnitude of the haircut or with not having resolved the asset in-​house. Construction of such counterfactuals and hindsight information without an understanding of the decision context by oversight agencies 2 Gopika Gopakumar and GC Prasad, ‘Onus on RBI as Government Notifies New Rules to Tackle NPAs of Banks’, Livemint (https://​www.livem​int.com/​Indus​try/​EzC​txzP​19Hm​wi14​g8kO​ mYO/​Onus-​on-​RBI-​as-​govt-​notif​i es-​new-​rules-​to-​tac​kle-​NPAs-​of-​ban.html) (Accessed on 27 December 2018). 3 Raj Kumar Ray, ‘Five Ways Govt, RBI Is Trying to Speed up NPA Recovery’, The Hindustan Times, 8 July 2017 (https://​www.hin​dust​anti​mes.com/​busin​ess-​news/​five-​ways-​govt-​rbi-​try​ ing-​to-​speed-​up-​npa-​recov​ery/​story-​kWK​LDFy​Z2Hb​5Rcy​oMj4​U8J.html)(Acces​sed on 27 December 2018).

Decision Paralysis  141 shapes the expectations of officials, which engenders, at best, sub-​optimal decision making, and, at worst, decision paralysis. In this environment, vitiated further by political acrimony, a large-​scale program of distressed asset sales from public sector banks to private firms will require these concerns to be addressed. A significant proportion of the so-​called stalled projects are in some way or other the result of ‘decision paralysis’. For example, a delayed project necessitates decisions on what and who was responsible for the delay, permissions for time extension and cost escalation, and often even revisions in the scope of work. Each of these decisions is liable to be criticized as benefiting the concessionaire. Officials therefore prefer to play safe and either penalize the contractor unfairly or delay making the decision itself. This sets off a cascade of events which further delays the project and raises costs. Interestingly, India’s experience with bureaucratic decision paralysis mirrors the ongoing trend in China following the massive anti-​corruption drive initiated by President Xi Jinping after assuming office. An empirical study by Peng Wang and Xia Yan of the impact of corruption and anti-​corruption on the efficiency of China’s bureaucratic system in developing a local economy found4: Government officials . . . are primarily “motivated to avoid blame and ensure their political survival” rather than to seek credit through taking the initiative. As a result, local officials prefer to reduce bureaucratic interactions and avoid providing official hospitality in order to protect themselves. Moreover, the Chinese government has introduced a lifelong accountability mechanism in many policy areas, such as environmental protection and economic development, in order to prevent government officials from making irresponsible decisions. The lifelong responsibility system makes public officials accountable for their administrative errors and the losses resulting from their decisions. An adaptive response from local officials, as China Daily reveals, is “trying

4 Peng Wang and Xia Yan, ‘Bureaucratic Slack in China: The Anti-​Corruption Campaign and the Decline of Patronage Networks in Developing Local Economies’, The China Quarterly, 19 November 2019 (https://​www.cambri​dge.org/​core/​journ​als/​china-​quarte​rly/​arti​cle/​burea​ucra​ tic-​slack-​in-​china-​the-​ant​icor​rupt​ion-​campa​ign-​and-​the-​decl​ine-​of-​patron​age-​netwo​rks-​in-​ dev​elop​ing-​local-​econom​ies/​8CBF5​E3B2​DEF1​9012​ED98​0C2F​62FD​DFE) (Accessed on 27 December 2018).

142  State Capability Challenges and Weaknesses to do as little as possible” because “the more work one does, the more chances there are for one to commit an error, which will possibly result in disciplinary penalties.” . . . dilemma for local officials: if they stick to the rules on official hospitality they risk offending their superiors; if they disregard the rules, they risk being investigated and punished. As a response, local government officials have reduced their interactions with different government departments in order to avoid this dilemma. This has directly led to the decline of the use of personalized bureaucratic ties in developing local economies . . . “Not hoping for merit, just avoiding the risks” (buqiu yougong, danqiu wuguo 不求有功，但求无 过) has become the survival strategy of Chinese bureaucrats in a highly uncertain political environment.

Any reform to address this problem has to be done with great care so as to not upset the delicate institutional balance that is critical to India’s vibrant democracy. It is not enough for the government alone to profess commitment to fairness in such assessments and investigations. It requires introspection, foresight, and leadership of an exceptional nature from all institutional stakeholders. An acknowledgement of the problem is, therefore, an essential first step. The different institutional issues are examined below.

8.2  Judiciary The Indian judiciary has established itself as one of the most independent in the world and has to its credit many path-​breaking innovations that have strengthened India’s democracy and governance over the years. Many transgressions of democratic or constitutional norms by the executive have been stopped or reversed by it over the years. Many distinguished and erudite judges have stood up for those who have otherwise been neglected and ensured that justice is done. For a society used in colonial times and earlier to an all-​powerful and arbitrary executive, the post-​Independence Indian judiciary has been a transformative experience in constitutionalism and democracy. It has decisively changed the culture of administration by reducing arbitrariness. However, this book is on the limited issue of state capability. Viewed from that narrow prism, one of the reasons for decision paralysis is the

Decision Paralysis  143 state of constant flux in administrative law in India because of the following features of the legal system:

(i) Courts often go into the merits of executive action rather than confine themselves to the Wednesbury principles or procedural fairness. Therefore, persons aggrieved by a regulatory or administrative order have a chance of success in challenging it even if it was procedurally correct and fair. (ii) The principles that the Courts expect to underpin administrative action have changed frequently and often without cogency in terms of previous decisions.5 (iii) Stare decisis (settled precedent) often cannot be assumed because the Supreme Court has quite often reversed its own earlier decisions. Thus legal uncertainty is high. (iv) Most courts have heavy backlogs of cases and final orders can take years at each level. (v) Courts often grant interim stays of executive actions that may be in operation for several years; interim/​interlocutory orders are far more common in India than in other common law jurisdictions. (vi) Higher courts tend to over-​ rule lower courts with a high frequency. (vii) Though there are no rigorous statistics, it is generally believed that the proportion of cases appealed is very high. Partly this is because of the high probability of reversal on appeal, the relative ease of obtaining interim stays (see iii above), and relatively low legal costs at the High Court level. (Private litigants, especially well-​heeled ones, find the cost of litigation disproportionately small compared to its potential benefits. Government litigants, who anyway do not bear any personal cost and may be criticized by an auditor or vigilance commission for causing loss to the exchequer by not contesting the decision, prefer to appeal on the pretext of protecting public interest. Appeal therefore becomes the default option.)

5 Arun Shourie in Courts and Their Judgements, New Delhi: Harper Collins, 2002 gives a brilliant, if dated, but still very relevant, account of this problem.

144  State Capability Challenges and Weaknesses Figures 8.1 to 8.3 outline the route to a finality that decisions take in different situations. Figure 8.1 relates to decisions taken by decisions made within the traditional executive and by regulators (like the RBI) who do not have a special tribunal attached to them. Here, there is no statutory right of appeal but the High Court can review the matter under its writ jurisdiction based on the general principles of judicial review of administrative action. Appeal to the Supreme Court is by special leave or if the High Court certifies the case as fit for appeal due to its legal or constitutional importance. The dotted lines reflect the fact that appeal is not a

Administrative Decisions: The Route to Finality (Type 1) Supreme Court

High Court (Bench) [“letters patent appeal”]***

Writ jurisdiction under Art.32 (only for violation of fundamental rights)*

Appeal - by special leave of Supreme Court - if certified by High Court as Involving substantial point of law

High Court (Single Judge)

Executive** (may include an administrative appeal within Executive branch)

Writ jurisdiction under Art. 226 (Judicial review of administrative action under Wednesbury principles etc. as considerably expanded by Indian Courts)

* Not routinely entertained ** Also applies to regulatory bodies not linked to a special tribunal - e.g. Reserve Bank of India *** May not be available in certain High Courts / for certain matters. Appeal / writ by leave of the Court (not automatic) Statutory appeal ( i.e. right to appeal)

Figure 8.1  Administrative decisions: The route to finally (Type 1)

Decision Paralysis  145 Regulatory Decisions: The Route to Finality (Type 2) Supreme Court

Appeal - by special leave of Supreme Court - if certif ied by High Court as involving substantial point of law

Statutory appeal if provided under the Act creating the Tribunal

High Court (Bench) Writ jurisd iction under Art.136 (by special leave of Supreme Court)* Writ jurisdiction (under Art. 226 as expanded in Chandra Kumar’s case)** Special Tribunal [1] (Single Bench) Appeal Regulator [1] Examples: TDSAT, CAT, SAT, Appellate Tribunal for Electricity * Rarely entertained after Chandra Kumar’s case ** Almost always entertained after Chandra Kumar’s case Appeal / writ by leave of the Court (not automatic) Statutory appeal ( i.e. right to appeal)

Figure 8.2  Regulatory decisions: The route to finally (Type 2)

matter of right, though in practice High Courts rarely refuse to hear writs on administrative matters. Figure 8.2 relates to a regulator with an attached tribunal with a single bench. The appeal to the Supreme Court may be statutory (i.e. leave of the court is not needed) or may be subject to special leave. Figure 8.3 relates to regulators (who may be part of the administration) for whom a national regulator with multiple benches exists. This is similar

146  State Capability Challenges and Weaknesses Regulatory Decisions: The Route to Finality (Type 3) Supreme Court

Appeal -by special leave of Supreme Court - if certified by High Court as Involving substantial point of law High Court * (Bench)

Statutory appeal under the Act creating the Tribunal***

Writ jurisdiction (under Art. 226 as expanded in Chandra Kumar’s case)** National Special Tribunal [2] with several benches*

Executive (Central)*

Executive (State level)*

[2] Example: NGT * The State where the cause of action arises and the Tribunal bench may be In different States. If so, more than one High Court may have j urisdiction. The government of one state may have to argue a case In the High Court of another state. ** Currently maintainable, per interim decision of Supreme Court and decisions of Madras & Uttarakhand High Courts, even if statutory appeal to Supreme Court is provided. *** If not provided for in the statute, then appeal by special leave under Article 136 would apply. Appeal / writ by leave of the Court (not automatic) Statutory appeal ( i.e. right to appeal)

Figure 8.3  Regulatory decisions: The route to finally (Type 3)

to Figure 8.3, but with the additional possibility of more than one High Court being involved—​this is because a bench of a national tribunal may cover more than one state. It is not inconceivable that more than one High Court may have jurisdiction in the same matter. Yet another variant is the Company Law tribunals where apart from the executive, there is a National Company Law Tribunal (NCLT) and then a National Company Law Appellate Tribunal (NCLAT), with further recourse to the High Court and Supreme Court.

Decision Paralysis  147 Another important issue is that of judicial activism. Judicial activism, manifesting as judgements on Public Interest Litigation (PIL), has been around for nearly four decades. It was welcomed as a healthy response to address both executive excesses and inaction. Several landmark judgements like those that came down hard on corruption, safeguarded democracy, protected the environment, restored rights of the underprivileged, and so on have become not only causes celebres but also contributed greatly to public welfare. While largely laudable, this activism has had its share of negative consequences. Some legal experts feel that instead of interpreting law as laid out in the statute, judges have often delivered not only their subjective interpretation of justice but also made laws across a wide variety of subjects. The former Judge of the Supreme Court, Justice Markandey Katju, lists out a series of judgements where judges have effectively legislated6: First, in Arun Gopal v. Union of India (2017), the Supreme Court fixed timings for bursting Diwali fireworks and prohibited the use of non-​ green fireworks, although there are no laws to that effect. Second, in M.C. Mehta v. Union of India (2018), the court annulled the statutory Rule 115(21) of the Central Motor Vehicle Rules, 1989, when it directed that no BS-​4 vehicle should be sold after March 30, 2020, and that only BS-​6 vehicles can be sold after that date. Third, in Subhash Kashinath Mahajan v. State of Maharashtra (2018), the court amended the Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act, 1989, by annulling Section 18 which said that no anticipatory bail will be granted to persons accused under the Act; by requiring a preliminary enquiry; and by prohibiting arrest under the Act except with permission in writing by the appropriate authority. Fourth, in Rajesh Sharma v. The State of Uttar Pradesh (2017), the court felt that Section 498A of the Indian Penal Code was being misused. So it amended that Section by requiring complaints under that provision to be sent to a Family Welfare Committee constituted by the District Legal Services Authority, although there is no such requirement in Section 498A. Finally, the National Green Tribunal (NGT) ordered that no 15-​year-​old 6 Markandey Katju, ‘When Judges Legislate’, The Hindu, 16 November 2018 (https://​www.thehi​ ndu.com/​opin​ion/​op-​ed/​when-​jud​ges-​legisl​ate/​arti​cle2​5508​359.ece) (Accessed on 3 January 2019).

148  State Capability Challenges and Weaknesses petrol-​driven or 10-​year-​old diesel-​driven vehicle will ply in Delhi, and the Supreme Court has directed impounding such vehicles, though neither the NGT nor the Supreme Court are legislative bodies.

His conclusion on whether judges can legislate is pertinent here: In Ram Jawaya v. The State of Punjab (1955), the court observed: “Our Constitution does not contemplate assumption, by one organ or part of the state, of functions that essentially belong to another” . . . Making laws is the function of the legislature. As observed in Union of India v. Deoki Nandan Aggarwal (1991), “The power to legislate has not been conferred on the courts.” In Suresh Seth v. Commissioner, Indore Municipal Corporation (2005), the court observed: “Under our Constitutional scheme, Parliament and Legislative Assemblies exercise sovereign power to enact laws” . . . If judges are free to make laws of their choices, not only would that go against the principle of separation of powers, it could also lead to uncertainty in the law and chaos as every judge will start drafting his own laws according to his whims and fancies.

There have been economic costs to such activism.7 Harish Salve, one of the country’s most eminent lawyers, blamed the Supreme Court for the economic slowdown in 20198: I squarely blame the Supreme Court. I can understand holding people responsible for the wrong distribution of licenses in 2G . . . Blanket cancellation of licenses where foreigners are investing . . . see when a foreigner invested it was your rule which said he must have an Indian partner. The foreigner did not know how the Indian partner got a license. Foreigners invested billions of dollars, and with one stroke of the pen, the Supreme Court knocked all of them out. That’s when the decline of the economy began . . .

7 ‘Ease of Doing Business’ Next Frontier: Timely Justice, Economic Survey 2017–​18’, Volume I, Chapter 9 (https://​mof​app.nic.in/​eco​nomi​csur​vey/​eco​nomi​csur​vey/​pdf/​131-​144​_​Cha​pter​_​ 09_​ENGL​ISH_​Vol%2001_​2​017-​18.pdf) (Accessed on 3 October 2019). 8 Interview of Harish Salve by Indira Jaisingh, The Leaflet, 15, September 2019 (https://​the​leaf​ let.in/​squar​ely-​blame-​supr​eme-​court-​for-​econo​mic-​slowd​own-​har​ish-​salve-​in-​conve​rsat​ion-​ with-​ind​ira-​jais​ing/​) ( on 3 October 2019).

Decision Paralysis  149 The Supreme Court, the way they have been completely inconsistent in dealing with commercial cases, causes grave concern in the minds of investors. So, you increase the price of an investment. You cancelled coal mines by one stroke of the pen, without examining the merits of every case. Much genuine foreign investment in the coal industry went flat. Then what happened? Indonesian coal and other world coal prices softened up. It became cheaper to import. Correct? A few million people are without jobs in India. Indian coal mines are lying closed, and we are importing coal. That is putting pressure on the economy. Iron ore mines. I must say, I’ve read that judgement of the Supreme Court and have always fought for the environment. That judgement is a howler. You’ve shut down all iron ore mines—​iron ore cases of Goa. Justice Pattnaik’s bench opened the Goa mines. There were people who invested in a barge. You know how Goa works. People who bought barges, people who bought trucks, you shut down. So then, demand falls.

Thus, interpretations of law without full appreciation of the decision context and influenced by subsequent events have resulted in contracts being abrogated for matters that go far beyond the scope of contractual terms. In one judgement, the NGT imposed an ‘environmental tax’ on commercial vehicles entering Delhi in addition to an existing toll tax, directed investments by toll operators over and above their contractual obligations, and even fixed the toll collection charges for the operators.9 Routine administration too gets affected by judicial interventions. Procurement and human resource matters are frequently affected. It is not uncommon to have promotions and disciplinary actions being stuck for years due to litigation. In such cases, courts’ interim stay orders can freeze matters for years, thereby stifling any room for administrative action in those areas. In fact, it is commonplace for Secretaries and Heads of Departments of Education, Health and Tax Departments across states to spend more time attending to employee litigation than administration or policy formulation. 9 ‘Commercial Vehicles entering Delhi Ordered to Pay “environment tax” ’, Indian Express, 8 October 2015 (http://​indian​expr​ess.com/​arti​cle/​cit​ies/​delhi/​com​merc​ial-​vehic​les-​enter​ing-​ delhi-​to-​pay-​enviro​nmen​tal-​tax-​ngt/​) (Accessed on 3 October 2019).

150  State Capability Challenges and Weaknesses Article 136 of the Constitution empowers the Supreme Court to ‘grant special leave to appeal from any judgment, decree, determination, sentence or order in any cause or matter passed or made by any court or tribunal in the territory of India’. Its sweeping nature, institutionalized in the form of Special Leave Petitions (SLPs), evidently means that it has to be exercised in cases involving a substantial question of law or a gross miscarriage of justice. But this ‘extraordinary’ jurisdiction appears to have been reduced to a regular appellate one—​34,500 SLPs were filed in 2014, of which 43% were admitted, in a Supreme Court with a caseload of just over 60,000.10 The Competition Commission of India (CCI) is a regulator established in 2009. A newspaper article that consolidated information accessed from the government informs that 97% of the penalties imposed by the regulator since inception have been stayed by the courts or appellate authorities.11 In 2018–​2019, of the 68 complaints received about anti-​competitive practices, the CCI completed investigation and passed orders on 51 and imposed penalties of Rs 357.85 crores, of which by 31 March 2019, it collected just Rs 1.41 crores. Most of the orders are being litigated.12 For sure, the CCI cannot absolve itself of its share of the blame for the quality of its orders. But it is unlikely to have been so bad that 97% of its orders are not only contested but continue to remain under litigation. Tribunals were introduced as a progressive step to minimize litigation and reduce the caseload on the higher judiciary through state and central legislation. Unfortunately, instead of helping to clear the backlog of pending cases, tribunals have become another institutional layer of litigation, since the losing party invariably ends up going on appeal before the courts. The courts have tended to entertain even petitions on procedural issues on cases being heard by the tribunal and go beyond points of law in appeals on tribunal decisions. The ongoing experience with the

10 ‘The True Worth of a Senior Advocate’, Livemint, 16 September 2015 (http://​www.livem​ int.com/​Polit​ics/​FFg​FOFn​zN8r​qvRN​WTTg​ugM/​The-​true-​worth-​of-​a-​sen​ior-​advoc​ate.html) (Accessed on 5 October 2019). 11 ‘Most of CCI’s Penalties Are Stuck in Courts’, Livemint, 14 December 2015 (http://​www. livem​int.com/​Polit​ics/​0lD​EKk4​J2v5​Jah9​q2cP​YwJ/​Most-​of-​CCIs-​penalt​ies-​are-​stuck-​in-​court. html). 12 Arjun Srinivas, ‘Why Anti-​Trust Is a Jumbo Indian Problem’, Livemint, 3 August 2020 (https://​www.livem​int.com/​news/​india/​why-​antitr​ust-​is-​a-​jumbo-​ind​ian-​prob​lem-​115​9637​ 5452​543.html).

Decision Paralysis  151 Insolvency and Bankruptcy Code 2016 has only served to reinforce this perception.13 Activism has extended to issues of social and religious concern: the Supreme Court adjudicated a PIL on a bull-​taming sport, Jallikattu, played in parts of Tamil Nadu during Pongal celebrations. Such rule-​making by judges is perceived by some to border on the ancient practice of kritarchy.14 Another practice observed in recent times is that of court directions forcing governments into a fait accompli on purely administrative matters. In a recent instance, the Supreme Court made strong observations on the Andhra Pradesh government’s decision to conduct the board examinations for 12th standard despite the Covid-​19 pandemic. The court found inadequacies with the state government’s assurance that Covid-​19 protocol will be followed and only 15 students will be seated in a hall; asked why the state government could not follow other state boards, CBSE, and ICSE and cancel the examinations; and warned orally during a hearing that ‘if any fatality occurs, the state will be held responsible’.15 This left the state government with little choice but to reverse its original administrative decision and cancel the examinations as desired by the court. Thus a careful balancing of multiple objectives—​the educational interest of students with the health of the population—​which is the kind of decision the executive is expected to take (and is accountable to the people for), was replaced by a judicial diktat to prioritize one objective over the other. Overall these tendencies, and the initial acceptance of frivolous writ petitions, have had significant consequences on executive decision-​making. It has also resulted in increasing the judicial workload and thereby reducing the effectiveness of courts. A new dimension is an emergence of ‘Suo Motu writ petitions’, where no litigant approaches the court but the court on its own motion decides to initiate litigation against the government. In some such cases, the judge uses his or her personal experience and adjudicates on public issues. In a

13 Gulzar Natarajan, ‘The Birth Pangs of India’s IBC Continues’, Urbanomics, 18 July 2019 (https://​gulza​r05.blogs​pot.com/​2019/​07/​the-​birth-​pangs-​of-​ind​ias-​ibc-​contin​ues.html). 14 Kritarchy, Wikipedia (https://​en.wikipe​dia.org/​wiki/​Kritar​chy). 15 ‘Andhra Pradesh Cancels Class 12 Exams After Supreme Court’s Critical Remarks’, Livelaw. in, 24 June 2021 (https://​www.live​law.in/​top-​stor​ies/​and​hra-​prad​esh-​canc​els-​class-​12-​exams-​ after-​supr​eme-​cou​rts-​criti​cal-​rema​rks-​176​224) (Accessed on 21 July 2021).

152  State Capability Challenges and Weaknesses recent case, a judge of the Delhi High Court observed the fellow passengers in a flight not wearing their masks properly. He initiated a Suo Motu writ petition. He commented as follows: It was noticed that, though all the passengers had worn masks, many passengers had worn the masks below their chin and were exhibiting a stubborn reluctance to wear their masks properly. This behaviour was seen not only in the bus transporting the passengers from the airport to the flight but also within the flight itself. It was only on repeated entreaties made (by me) to the offending passengers that they condescended to wear their masks properly. On the cabin crew being questioned in this regard, they stated that they had directed all the passengers to wear masks, but were helpless in case they did not comply.

The judge then adjudicated the petition he had initiated and issued legally binding orders16 on what airlines should do. Such Suo Motu writ petitions are a further expansion of the scope of writ petitions beyond the already-​liberally interpreted scope of PIL. They allow a judge to effectively take up public issues not only beyond the traditional maxim of locus standi but even beyond the newer realm of PIL. Suo Motu writs arguably create the situation of being a judge in one’s own cause, in contrast to the traditional natural justice doctrine of nemo judex in causa sua potest. These are practices that are not observed in other democracies. Many argue that such judicial pronouncements, while often very popular in the short run with the litigants (some have characterized cases filed by social activists as ‘publicity interest litigation’), cannot be a substitute for executive action and that such activism may distort incentives and affect judicial workload and effectiveness. Sudeep Mahajan, a former Additional Advocate General, made the following observations: In the wake of the coronavirus’s second wave hitting its peak, there have been some very interesting observations from the higher judiciary.

16 ‘Covid 19: Take Action Against Passengers Not Wearing Masks Properly, Delhi HC judge Tells Airlines’, Scroll.in, 09 March 2021 (https://​scr​oll.in/​lat​est/​989​068/​covid-​19-​take-​act​ion-​ agai​nst-​pas​seng​ers-​not-​wear​ing-​masks-​prope​rly-​delhi-​hc-​judge-​tells-​airli​nes) (Accessed on 21 July 2021).

Decision Paralysis  153 On April 24, a Division Bench of the Delhi High Court observed that it will ‘hang any person’ who tries to obstruct oxygen supplies to hospitals in Delhi. A bench of the Madras High Court said on April 27 that the Election Commission is singularly responsible for the second wave and should probably be tried on murder charges for its failure to ensure adherence to Covid-​19 protocols. The Supreme Court was later constrained to call these observations ‘harsh and inappropriate’. The SC further emphasised that ‘there is need for judges of superior courts to exercise restraint’ . . . However, the judicial quote that probably takes the cake is, when, the Delhi High Court, on June 1, while directing the government to prioritise younger people over older ones when it came to administering Amphotericin-​B, a drug used to treat mucormycosis, observed that the young are the ‘future of the country’ and need to be saved, while the elderly have ‘lived their life’. The question, however, is: What causes the higher judiciary, at times, to make such sweeping remarks when the role assigned to it in the Constitution makes it almost necessary that the observations from a judge, holding a court, should always be tempered with moderation, restraint, discipline and even wisdom? Is it the extraordinary hard work that the judges need to put in, in order to keep up with ever increasing piles of case files? Or is it because the higher courts have come to firmly believe that in this country, they are the ultimate arbiters and therefore every order or decision or policy of the state must get their stamp of approval, for things to move forward at all? . . . In Census Commissioner & others v. R Krishnamurthi, where a high court had issued directions to achieve social justice, the Supreme Court held, ‘it is not within the domain of the court to Legislate . . . The courts have the jurisdiction to declare the law as unconstitutional . . . But the courts are not to plunge into policy making by adding something to policy by way of issuing a writ of mandamus. There, judicial restraint is called for’. It further added, ‘. . . refrain and restraint are the essential virtues in the arena of adjudication because they guard as sentinel’. One quote mentioned therein is from . . . Francis Bacon: ‘Judges ought to be more learned than witty, more reverend than plausible and more advised than confident . . . Let the judges also remember that Solomon’s throne was supported by lions on both sides: let them be lions but yet lions under the throne’.

154  State Capability Challenges and Weaknesses . . . The question is why do courts in India often forget or neglect such pearls of wisdom? Given the global scale of the pandemic, why haven’t courts anywhere else in the world, distrusting the state, taken it upon themselves to arrange for oxygen, medicines and vaccines? Or taken to directing which section of patients should be attempted to be saved and which should be allowed to let go for having ‘lived their life’? These kinds of directions are beyond the jurisdiction of the courts. These are clearly the arenas that must be left for the government to govern. The assumption that everyone in the government is either corrupt or incompetent cannot always be a valid starting point of all judicial actions. The courts must never disregard that the Preamble of the Constitution unequivocally makes the ‘people of India’ the source of all power that flows from its various provisions. It is these very people who go to polls every five years to elect their representatives to Parliament and the state legislatures. Only these representatives, empowered by an almost-​sacred process of elections, are answerable and accountable to the people of India and therefore they alone have the power to enact laws and to frame policy. The people of India in turn have the right and the power to change their representatives, in the next elections, if they do not like the laws enacted and the policy framed by them . . .17

The effects of the tendencies noted above manifest in the World Bank’s Doing Business Survey rankings for 2019, where India ranked a lowly 163rd in the ease of enforcing contracts,18 judicial delays being the biggest contributor to this. Validating this, the Economic Survey 2017–​18 found19 : 1. Delays and pendency of economic cases are high and mounting in the Supreme Court, High Courts, Economic Tribunals, and Tax Department, which is taking a severe toll on the economy in terms of stalled projects, mounting legal costs, contested tax revenues, and reduced investment more broadly; 17 Sudeep Mahajan, Judges are arbiters, not lawmakers, The Indian Express, July 12, 2021 (https://​indian​expr​ess.com/​arti​cle/​opin​ion/​colu​mns/​coro​navi​rus-​sec​ond-​wave-​delhi-​high-​ court-​oxy​gen-​elect​ion-​com​miss​ion-​7399​907/​) (Accessed on 24 July 2021). 18 Doing Business 2019, World Bank Group (https://​www.doingb​usin​ess.org/​en/​repo​rts/​glo​ bal-​repo​rts/​doing-​busin​ess-​2019)(Acces​sed on 12 October 2020). 19 ‘Ease of Doing Business’ Next Frontier: Timely Justice, Economic Survey 2017–​18, Volume 1, Chapter 9 (https://​mof​app.nic.in/​eco​nomi​csur​vey/​eco​nomi​csur​vey/​pdf/​131-​144​_​Cha​pter​_​ 09_​ENGL​ISH_​Vol%2001_​2​017-​18.pdf)(Acces​sed on 12 October 2020).

Decision Paralysis  155 2. Delays and pendency stem from the increase in the overall workload of the judiciary, in turn, due to expanding jurisdictions and the use of injunctions and stays; in the case of tax litigation, this stems from government persisting with litigation despite high rates of failure at every stage of the appellate process. The following views expressed by Justice S.C. Dharmadhikari, a renowned former Judge of the Bombay High Court provide an apposite conclusion to this section20: “The prerogative writs that can be issued by the Supreme Court and the plenary powers vested are for enforcement of the fundamental rights . . . Although it is said that the power of High Courts... under Article 226 is wider, it is undisputed that those approaching these courts must possess a legal right and must demonstrate a corresponding legal duty of any state authority . . . . . . [I]‌f there is no statute holding the field, courts cannot legislate in the garb of issuing directions and passing orders. Even if the Supreme Court passes orders under Article 142, it can only be to safeguard fundamental rights and to do complete justice while doing so—​not to create an ad hoc legal regime and take over the role of the legislature. Even the executive power of administering and enforcing the law cannot be taken over... .. Judicial power has in-​built limitations. Judicial power is not conferred to run government affairs. Judicial power will not enable the court to ignore the will of the legislature . . . Merely because the nature of the proceedings is a public interest litigation, this essential legal position does not undergo any change . . . ..Court orders ought to have a binding effect and must be capable of implementation. Judges and lawyers should not call upon the State to do what is virtually impossible . . . ...In our enthusiasm to make constitutional courts supreme and powerful, we must not effect a reversal of roles.”

20 S.C. Dharmadhikari, ‘Every healthcare issue is not for courts’, Business Line, May 8th and 10th 2021 (Accessed on 18 November 2021).

156  State Capability Challenges and Weaknesses

8.3  RTI and Its Effect on Civil Service Behaviour The enactment of the RTI Act in 2005 was an important milestone in Indian public administration. Indeed, without exaggeration, it was a milestone in public administration globally as the Indian RTI law has influenced not only many developed countries but even the information policy of the World Bank. Much has been said and written about its salutary purpose and effects. This part of the chapter is not an attempt to assess RTI in its entirety; rather, it is a very narrow look at the limited issue of the effects of RTI on civil service behaviour (and in this section the discussion covers both the higher and lower services). The RTI Act, 2005 works on the presumption of mandatory disclosure, except for a few exemptions, and enshrines the fundamental right of a citizen to demand and obtain information from ‘public authorities’. It helped shine light and unleash a giant wave of transparency on the functioning of insular public bureaucracies. It empowered the poor farmer to access his patta certificate, the harassed employee to know why her promotion was denied, and the complainant to see the contents of the First Information Report (FIR) filed by the Police against them. It also allowed civil society workers to obtain information about the considerations that led to decisions at all levels of the government, including those of the Cabinet. Apart from a handful of exceptions, vast swathes of government became transparent overnight. This was rightfully celebrated as a triumph of India’s democracy and a major victory for its citizens. The effects of the RTI Act on the civil service appear to have gone through a series of phases. Most officers, schooled in the British traditions of official secrecy and the default position that everything is restricted, had difficulty adjusting to the new law. Initially, the reaction was often to find a way—​even if contrary to the Act—​to deny requests. Many learnt painfully, through orders of the Information Commissions and the courts, that times had decisively changed since they were eventually compelled to provide it and sometimes fined. Then came the various definitional debates about whether one or other type of information was included or not; in every case the battle for restriction was lost and information had to be disclosed. The only institution that successfully waged the battle against disclosure was the Supreme Court and the judiciary.

Decision Paralysis  157 This led to the second phase: many ills were exposed because the files had been prepared without anticipation of disclosure. Juicy skeletons fell out of many a closet (if one pardons the unsavoury mixed metaphor). The initial impact was to reveal wrongdoing that previously would have remained unknown. Then came the third phase, where the civil servants realized that they were operating under a different regime. They first began to make sure that the note files (the records of internal deliberation) were now more carefully written on the assumption that they would be published. While the hope would be that this led to different advice and different decisions, in many cases what it led to was that the same advice as before was tendered orally; after decisions were taken, the file was written in a defensible manner. The fourth phase was when civil servants realized that they could harness the power of RTI to further their own career interests. A massive number of RTI applications are now filed by government servants at all levels, usually in the name of another (a spouse, friend, etc.). This creates additional work for the offices (in providing details, dealing with appeals, etc.). The extra work is performed either by diversion from other duties (since RTI has a time limit and clear penalty, it takes priority over normal citizen services) or by creating additional staff. In a sense, the establishment matters of government servants—​when raised through RTI—​take precedence over public service. The latest phase coincides with the trend of activist auditors and courts, and media trials, with attendant debilitating impact on public decision-​ making. The Act’s inclusion of every correspondence made within and by government, including those part of the ‘deliberative processes’, which resulted in a decision, led to an expansion of transparency beyond what was practiced even in the most mature developed democracies. In fact, the US Freedom of Information Act protects recommendations and advice which are made as part of the ‘deliberative process’ leading to decision-​ making. This expansion of scope immediately transformed a far less formal and more freewheeling decision-​making process into an excessively formal one where opinion is expressed with great restraint. For example, why would a Director or Joint Secretary risk going against conventional wisdom on any matter? Similarly, no official would even dare suggest a change in a tender specification or contracting process,

158  State Capability Challenges and Weaknesses howsoever beneficial, if the change could be misinterpreted as being also favourable to a large (and possibly stigmatized) corporate firm likely to participate in the bids. The Supreme Court’s observations in a recent case capture the problem21: There is a pervasive fear of RTI. Officials have become afraid of putting anything on files. We are informed that Mantralaya in Mumbai is getting paralysed because of incessant RTI applications . . . There are innumerable instances of blackmail through RTI . . . unbridled right to anyone and everyone to ask for any information is putting a strain on officials.

It is noteworthy that nearly all developed countries and international organizations restrict the disclosure of ‘deliberative information’ (basically, note files in the Indian context) because disclosure would affect the tendering of frank advice internally and is a communication that deserves ‘executive privilege’.22 The perception of rampant corruption in India and the general mood of suspicion of both civil servants and ministers mean that advocating non-​disclosure of anything risks being seen as supporting corruption. However, this is an area that may benefit from a dispassionate review. In the context of a similar debate on the issue of information disclosure and transparency in the United States, in a 2009 essay, Harvard Law School Professor Lawrence Lessig cautioned against excessive transparency.23 There is no questioning the good that transparency creates in a wide range of contexts . . . But we should also recognize that the collateral consequence of that good need not itself be good . . . Sunlight may well 21 ‘Dhananjaya Mahapatra, RTI Queries by Anyone and Everyone Paralysing Administration: SC’, The Times of India, 17 December 2019 (https://​times​ofin​dia.ind​iati​mes. com/​india/​rti-​quer​ies-​by-​any​one-​and-​every​one-​par​alys​ing-​adm​inis​trat​ion-​sc/​arti​cles​how/​ 72786​062.cms) (Accessed on 3 January 2020). 22 Her Majesty’s (HM) Government, The Civil Service Reform Plan, Institute for Government, London, UK, 2012. 23 Lawrence Lessig, ‘Against transparency’, The New Republic, 9 October 2009 (https://​newr​ epub​lic.com/​arti​cle/​70097/​agai​nst-​trans​pare​ncy) (Accessed on 10 June 2018).

Decision Paralysis  159 be a great disinfectant, but as anyone who has waded through a swamp knows, it has other effects as well24

He argued that given its ‘salience’ in the debate and the ‘assumptions of our political culture’, any accusation that a legislator took campaign donation in return for supporting a legislative proposal immediately sticks. Further, he argues, the ‘limited attention-​span’ of the audience rules out the possibility of an informed debate on the issue. He wrote: At this time the judgment that Washington is all about money is so wide and so deep that among all the possible reasons to explain something puzzling, money is the first, and most likely the last, explanation that will be given. It sets the default against which anything different must fight. And this default, this unexamined assumption of causality, will only be reinforced by the naked transparency movement and its correlations . . . To understand something-​-a​ n essay, an argument, a proof of innocence-​-​ requires a certain amount of attention. But on many issues, the average, or even rational, amount of attention given to understand many of these correlations, and their defamatory implications, is almost always less than the amount of time required. The result is a systemic misunderstanding-​-​ at least if the story is reported in a context, or in a manner, that does not neutralize such misunderstanding . . . you will begin to see the attention-​ span problem everywhere . . . Think of politics, increasingly the art of exploiting attention-​span problems-​-​tagging your opponent with barbs that no one has time to understand, let alone analyse.

Further, apart from the typical poor farmer or harassed complainant, an increasingly important beneficiary of the RTI has been the corporate sector. In cases of litigation with the government, corporate lawyers rely on the voluminous file notings and inter-​governmental correspondence accessed through the RTI Act to strengthen their case by exposing discrepancies and inconsistencies within those records. In complex and long-​drawn matters involving multiple decision makers at the same level

24 Presumably, the allusion is to the fact that sunlight also promotes the growth of weeds and pests.

160  State Capability Challenges and Weaknesses over time, it is not difficult to find loose ends that can be selectively subjected to the test of literal interpretation in their favour. Others have observed that the well organized (usually the relatively privileged or the corporate sector) may benefit the most because they can use it the best. The widespread use of RTI by government servants in service of their career interests is a case in point. An assessment of the effects of RTI on civil service behaviour suggests that: 1. On the positive side, it has made it easier for good civil servants to discourage the taking of egregiously bad decisions by pointing out that they would be indefensible in the public realm even with the best of reasoning. Civil servants are also taking greater care in thinking through and documenting the rationale for decisions. 2. On the other hand, the pressure on honest officers not to dissent on file appears to have increased because whereas earlier the dissent would only come out rarely, now there is a greater threat of disclosure. RTI has also made it easier for civil servants to leak details that they would like to leak by getting someone else to file an RTI application. For this reason, it has made ministers even more suspicious of dissent notes on files because the minister may wonder whether this is purely an internal expression of opinion or is a planned act of sabotage. 3. The disclosure of note files has reduced the recording of internal disagreements on files and made them less interesting and also less reliable for an ex post view of what actually happened. Instead, disagreements are aired verbally and only the final ‘consensus’ view is put in writing. This has often reduced the transparency of the record of decision through the censoring out of some viewpoints expressed internally. It is likely to be more difficult in the future to find out what really underpinned a particular decision, whether for historians or researchers or auditors or courts. 4. Overall, the RTI, in steady state, does not (at least not yet) appear to have had a beneficial effect on honesty and integrity of civil servants. The underlying premise for transparency is simple—​sunlight is the best disinfectant.25 In governance this translates into making public

25 The quote is attributed to Justice Louis Brandeis.

Decision Paralysis  161 information that went into decision-​making within the government. But, as the experience of India suggests, the reality is more complex. And this is also attested by evidence from both the public and private sectors. In one study, researchers analysed the changes in the behaviour of the Federal Reserve Open Market Committee (FOMC), the interest rate–​ setting committee of the US Federal Reserve, after the statutory enforcement of transparency laws in 1993.26 They found, Using various documents released by the Federal Reserve from 1976-​-​ 2007, covering both years in which the FOMC knew its deliberations would eventually be made public, and years in which it believed no records were kept, we find that FOMC deliberations became more similar to the always-​public press releases in the transparency regime. FOMC members shifted their comments towards popular economic subjects, such as “inflation” and “growth,” and away from personal opinions, like “think.”

In other words, with greater transparency, the FOMC members shifted towards a more cautious and politically correct stance in the Committee meetings. Another study of workers in a large mobile phone manufacturing facility in China found that excessive transparency in organization design can adversely impact workers’ productivity and organizational performance.27 In fact, they found the cost to be a 10–​15% reduction in performance. The authors point to a transparency paradox: . . . maintaining observability of workers may counterintuitively reduce their performance by inducing those being observed to conceal their activities through codes and other costly means; conversely, creating zones of privacy may, under certain conditions, increase performance. Empirical evidence from the field shows that even a modest increase in group-​ level privacy sustainably and significantly improves line 26 Michael Egesdal, Michael Gill, and Martin Rotemberg, ‘How Federal Reserve Discussions Respond to Increased Transparency’, SSRN Working Paper, 21 October 2015 (https://​pap​ers. ssrn.com/​sol3/​pap​ers.cfm?abst​ract​_​id=​2676​429) (Accessed on 3 October 2019). 27 Ethan S. Bernstein, ‘The Transparency Paradox: A Role for Privacy in Organisational Learning and Operational Control’, Administrative Science Quarterly 57, no. 2 (2012): 181–​216 (https://​www.hbs.edu/​facu​lty/​Publ​icat​ion%20Fi​les/​Bernstein_​TransparencyPa​rado​x_​AS​Q_​Ju​ ne20​12_​c​daae​e20-​3a45-​4a07-​8867-​9761a​5d4b​5e8.pdf) (Accessedon 3 October 2019).

162  State Capability Challenges and Weaknesses performance, while qualitative evidence suggests that privacy is important in supporting productive deviance, localized experimentation, distraction avoidance, and continuous improvement.

Clearly too much of even the best disinfectant is poisonous.

8.4  Vigilance and Investigative Agencies Once an issue is ‘exposed’ through an RTI application or by a CAG audit, it attracts the attention of a plethora of anti-​corruption agencies—​Lok Ayukta, Anti-​Corruption Bureaus, Vigilance Directorates and Commissions, Central Bureau of Investigation (CBI), etc. Investigations into white-​ collar crimes, especially those involving public policy decisions, are at best of times extremely difficult. Given the attribution problems associated with policymaking actions, much less the availability of direct evidence of gratification, it can be a challenge to establish culpability beyond reasonable doubt in such cases. In matters involving technical specifications of tenders or decisions relating to financial market regulations or policies on public resource allocations, investigations also demand proficiency in these areas. From 1988 and till amended in 2018, the Prevention of Corruption Act had a section, Section 13(1)(d) (iii), under which a public servant who ‘obtains for any person any valuable thing or pecuniary advantage without any public interest’ was guilty of corruption. The key element of this offence is that there was no need to establish mens rea on the part of the officer or to establish recklessness or even negligence. Any order which results in a pecuniary benefit to a private party without public interest is liable to be prosecuted as corruption. Thus, an act of a public servant which—​without any prior knowledge on his or her part—​ends up benefiting someone can potentially be characterized as an act of corruption. Thankfully, this section was repealed in 2018 though it still applies to acts performed earlier. The good intention behind this section was that it is often easy for a corrupt public servant to conceal the act of bribery and it is often difficult for the prosecution to establish mens rea. Removing the requirement of a guilty mind would make it easier to nail the corrupt.

Decision Paralysis  163 However, the practical effect of the section was that innocent and honest officers could be prosecuted for corruption, with hindsight, on the basis of decisions that ended up unwittingly causing a gain to some party (and even if there were no loss to the exchequer). Its impact was deeply debilitating. A decision, by definition, involves an exercise of judgement. In even a ‘rules-​based’ regime, the very decision of finalizing rules would invariably benefit one party at the cost of another. It is fallacious to presume that ‘following the rules’ amounts to an algorithmic application of those rules. Deciding technical and financial eligibility specifications of even an open and competitive bid document involves the exercise of considerable subjectivity, which often benefits one party or category of prospective bidders at the cost of another or others. Taken to its logical conclusion, every decision could be questioned. For example, it is one of the fundamental principles of contracting that risks be allocated to those who can bear it the best. In many infrastructure contracts, this also means that the government bears the burden of all residual risks; but absolving private partners of risks that cannot be mitigated could be seen as favouring them at the cost of the taxpayers. Officials would be unwilling to stick their necks out and caution against such improper risk allocation for fear of being accused, in any subsequent enquiry, of having caused loss to the government if the risks materialized. It is no surprise that many infrastructure contracts in India have an unfair allocation of risks against the private party. Similarly, in the case of delayed projects, decisions involving apportionment of responsibility for the delays are invariably an exercise of judgement. Officials prefer to play safe and lay the blame on the concessionaire by default and let them seek legal redress. Much the same applies to all cases involving a reference made to a Ministry to adjudicate on a difference of opinion between an agency under that Ministry and a private provider engaged by it. The Ministry officials invariably prefer to lean towards the public agency, even when evidence favours the private contractor, so as to avoid taking a decision that would be perceived as having favoured a private party at the cost of the government. The private party is left with no option but to seek arbitration or litigation. All these incentive distortions apply with equal relevance to all spheres of policymaking. Second, the relatively poor investigative training, skills, and resources of the anti-​corruption agencies in dealing with white-​collar

164  State Capability Challenges and Weaknesses crime vis-​à-​vis their counterparts in developed countries mean that they generally take the easy route of basing a prosecution solely on a ‘criminal conspiracy’ supported by notes on a file, without any proof of actual corruption on the part of a civil servant. For instance, where B, a businessman, has done something illegal, and C (a civil servant) has written something on a file favourable to B’s application or request, the favourable noting becomes evidence of C conspiring with B and hence makes C liable for the criminal actions of B—​even though C might have been acting with no improper motive. For instance, when an industry requests something from the government for a project which has positive economic and employment effects, a well-​meaning and honest officer may indeed recommend a favourable decision. If it later turns out that the industry had broken the law in some respect or bribed a minister without any knowledge on the part of the civil servant, the positive recommendation can become a source of suspicion and prosecution. By way of illustration of the twin effect of these tendencies, in 2001, after an election, a newly elected state government launched a prosecution against a former city mayor (of an opposing political party), alleging corruption in the award of high-​value civil work. Two years earlier, the rules had been amended by delegating enhanced authority to elected local bodies, at the initiative of the Secretary of the Municipal Administration Department, an officer of impeccable integrity and exceptional competence, who had been very serious about implementing the spirit of the 73rd and 74th Amendments. She had written a strong note seeking enhanced delegation for all local bodies and had piloted a bill that was passed by the State Assembly. She was initially named by the prosecution as a co-​conspirator since her action of delegating enhanced powers enabled the mayor to proceed with the alleged corrupt transaction, which resulted in a wrongful gain to the contractor. The charges against the officer were eventually dropped but only after she was named as a corruption suspect in the press; she was also never empanelled for posts in the Government of India as a result. The result of this tendency of ‘investigation by file reading’ instead of ‘investigation by tracing the money’, which has gathered speed in the last decade, is that many officers (of all types) feel it is always best to initially oppose on file any request from the private sector or indeed a private

Decision Paralysis  165 citizen even if genuine, as that is the best way to not be accused of conspiring in corruption. Overcoming the negative recommendation would then require a committee which, by its very size and diverse membership, cannot easily be accused of being guilty of conspiracy. This greatly slows down and complicates policymaking and further increases the citizen-​unfriendliness of the government. The approach (while possibly deterring some egregious forms of corruption) has also had a deleterious effect on decision-​taking, with decisions being made in the interest of self-​protection of officers rather than in public interest. Procedural purity has become more important than substantive correctness. There are some ‘rules of thumb’ that some officers (honest as well as dishonest) follow to minimize the risks of prosecution under these provisions: 1. Initially oppose all requests, however genuine, even if you intend to later accept them; your initial opposition will protect you as the investigators and courts will presume you were ultimately coerced into changing your mind either by your superior or by a committee or the Cabinet. 2. Accept the lowest tender meeting the minimum specifications on paper, even if the bidder’s technical capacity or ability is dubious; hidden costs of delay or bad quality, even if known, are ignored as these will not be apparent on file. 3. If there are any unforeseen situations during contract execution that might require re-​negotiation of terms, pursue a legal dispute with the contractor even if it delays a project and increases the project cost; any cost increase through losing an arbitration or re-​tender (even a corrupt arbitration or re-​tender) is unlikely to be viewed as corrupt, whereas even an honest exercise of re-​negotiation will appear corrupt. 4. Buy from the public sector whenever possible, even if quality or timeliness is poor and even if the supplier has a bad track record. 5. Do not assess quality otherwise than on quantifiable criteria—​ignore qualitative aspects that cannot be reduced to numbers on file. 6. Whenever possible, take decisions by committee even if it is slower and/​or subtracts quality by bringing in the inputs of unqualified

166  State Capability Challenges and Weaknesses and inexperienced outsiders—​the presence of outsiders is more important than the quality of the decision. 7. Mechanically apply precedent; do not exercise discretion even when the situation calls for it or when hardship will be caused; for any change from precedent, always either promulgate a new rule or create a large committee. The Indian parliamentarian and member of the Constituent Assembly, H.V. Kamath (1907–​1982), who began his career as an Indian Civil Service officer, is quoted as saying the following to say about how the government functions: If you can, don’t move; If you must, move slowly; If pushed, move in circles; If in doubt, appoint a committee.

The corrupt (at political and official levels) are quite capable of being corrupt without breaking the formal rules and while adhering scrupulously to procedure, since ‘bribery simpliciter’ need not involve any violation of rules and cannot be detected through file reading. Unfortunately, the investigative and prosecutorial approach often does not recognize this. Investigators and disciplinary authorities should acknowledge the very strong likelihood of policy choices not playing out as expected, even engendering completely unintended or unforeseeable rent-​seeking mechanisms. Not even the most far-​sighted and enlightened policymaker can anticipate all the emergent dynamics in a complex policy environment and design a policy that mitigates all possible failure pathways. In fact, in large-​scale projects, re-​negotiations and periodic course corrections are par for the course. The challenge lies in being able to scrutinize the decision-​making process to identify transgressions from the due process and/​or mala fide actions. It is necessary for investigators to distinguish three types of decisions.

Decision Paralysis  167   (i)  A Diligent Decision That Ended Up with a Negative Outcome It is necessary to understand that a good decision does not always mean a positive financial result. An example is a decision to hedge the exchange rate risk in a contract. At a given point of time, there is no fool-​proof way of predicting whether the rupee’s rate of exchange with, say, the yen will depreciate or not. ‘Expert’ opinion is rarely correct in such matters. If a decision is taken to hedge the risk, then the decision will give a favourable result if the rupee actually falls. But if it rises, the decision to hedge may prove to be financially negative. It may have been better to remain unhedged. A decision to hedge should not be criticized merely for this reason—​it could be criticized only if there was no due diligence before taking the decision. Two identical decisions with contrasting outcomes in cricket may help illustrate the point: In the Oval Test of 1979, India in the fourth innings needed 438 runs to win. India had reached 366 when the second wicket fell. Since quick scoring was needed, the captain (S.Venkataraghavan) sent the inexperienced Kapil Dev in at No. 4 instead of the seasoned G.R. Viswanath who was the regular No. 4. The hope was that Kapil Dev, known to be a fast scorer, would accelerate the scoring. The move failed because Kapil Dev got out for a duck in five balls. Vishwanath was more successful scoring 15 in 11 balls but also got out. India ended up drawing the match, by reaching 429 for 8, but Venkataraghavan was criticised for the move because it did not pay off. By contrast, in the fourth innings of the Sydney Test of 2021, with India chasing 407, at the fall of the third wicket, the captain (Ajinkya Rahane) sent in Rishabh Pant at No. 5, ahead of his normal spot. Pant scored 97 quick runs and greatly helped India’s cause. Rahane has been hailed for this decision. The decisions were identical—​to send in a fast-​scoring batsman to try and win the game. But one was applauded (because it paid off) while the other was criticised because it did not. Similarly in the World Cup Final of 2011, captain M. S. Dhoni promoted himself in the order to

168  State Capability Challenges and Weaknesses accelerate scoring; the decision paid off with Dhoni scoring a match-​ winning knock and India won. It was hailed as a great decision. If India had lost, the captain may have been sharply criticised and even accused of putting his personal interests ahead of the team! Civil servants are often in similar situations.

  (ii)  A Negligent Decision A negligent decision would be one taken without proper care but with no corrupt or bad intention. Such decisions need to be dealt with through administrative measures, which may be corrective or even punitive (disciplinary proceedings including dismissal, etc.) But they should not be the subject matter of criminal prosecution, except in those limited instances where negligence itself is a crime.   (iii) A mala fide Decision A decision taken with corrupt intent should of course be met with prosecution for corruption. In the hedging example, if the decision taker had obtained a commission from the financial institution placing the hedge, it would warrant prosecution even if the hedging decision was good.

8.5  Comptroller and Auditor General In an influential work in the nineties, Michael Power drew attention to the problems with excessive reliance on the use of audit mechanisms to exercise oversight and administrative control.28 He cautioned against it being viewed as ‘an all-​purpose solution to the problem of administrative control’. He wrote,

28 Michael Power, The Audit Explosion, Demos, 1994 (https://​www.demos.co.uk/​files/​theaud​ itex​plos​ion.pdf) (Accessed on 17 March 2019).

Decision Paralysis  169 Audits may provide comfort to stakeholders who are remote from day to day practices but, in doing so, they often deter substantive inquiry which would empower stakeholders. Audit arrangements can bring an end to dialogue inside and outside organisations, rather than helping it. In so far as audit is directed at evoking good feelings about organisational practices it may also become a new form of image management rather than a basis for substantive analysis. In this sense, the fact of audit is becoming more important than the how of audit.

Another paper examined the effect of bureaucratic control on the activities of field workers in international development organizations.29 It finds that such control attempts, in order to improve performance, perversely end up sometimes ‘having net effective opposite those intended’. Oversight gone too far leads to ‘a reluctance among officials to exercise judgement and take decisions’. It writes, First, controls may induce agents to focus on meeting formal requirements rather than the service they are meant to deliver. Second, controls may preclude agent initiative and productive use of the asymmetric information to which agents uniquely have access.

These findings on the impact of audits on bureaucratic incentives resonate strongly with developments in India for over a decade. Article 149 of the Constitution empowers the Comptroller and Auditor General (CAG) of India to exercise such powers as entrusted to it ‘in relation to the accounts’ of central or state governments or public entities. This would include financial audit of all public revenues and expenditures and accounts of public entities. But in recent years, like with other institutions, the scope of CAG activities has expanded dramatically to encompass performance audits and even assessment of policies. But

29 Dan Honig, ‘When Reporting Undermines Performance: The Costs of Politically Constrained Organizational Autonomy in Foreign Aid Implementation’, International Organization 73, no. 1 (2019): 171–​201.

170  State Capability Challenges and Weaknesses this functional expansion has not been accompanied by sufficient appreciation of the decision-​making context. In the political history of India in the last decade, the headline number of Rs 1,76,000 crores (nearly USD 30 billion) will figure prominently. That was the presumptive loss, as estimated by the country’s CAG, caused by the allocation of the 2G telecommunications spectrum at non-​market rates. The CAG was hailed for bringing this to light, though many neutral and credible experts felt the estimate of loss was grossly exaggerated. The expanded role of ‘audit’ has also led to negative consequences. In the real world, with its numerous uncertainties, constructing counterfactuals is often impossible. Controlled experiments are possible only in natural sciences. In real life and in policymaking, the context is never a constant, and yet, it is an important variable in decision-​making. It is misleading to attribute the loss caused in one situation to simplistic extrapolation from another unrelated context—​which is what the CAG had done in the 2G case. Even unintentional omissions and errors may be interpreted as having caused presumptive loss to the government. The possibility that their decisions can be subjected to such ex post facto extrapolations and presumptive loss assessments, even several years after the decision, encourages officials to avoid hard decisions involving asset or resource valuations. And when forced to make a choice, the bureaucracy reflexively prefers auctions as their universal choice for discovering price, irrespective of the context and sector. This overlooks the fact that in imperfect markets, with the possibility of future re-​negotiations, auctions can sometimes generate sub-​optimal outcomes. Developers may bid aggressively to bag the resource block or the contract, with the assurance that they can renegotiate later. Another example is banking. Banks are now struggling to address the issue of rising NPAs, which threatens to cripple the banking sector and stifle economic growth itself. The commonest strategy globally to address NPAs is to auction them to private ARCs after writing off a share of their loans. But this runs against the likelihood that at least some of these assets may generate windfall gains for the ARCs, thereby raising the possibility of attribution of presumptive loss to the public exchequer: ‘If the ARC

Decision Paralysis  171 could realise Rs. 200 for this asset, why did you sell it to them for just Rs. 150?’ This, with the attendant disciplinary and other proceedings, incentivizes bankers to postpone the recognition of bad assets, which in turn increases the cost and exacerbates the NPA problem. Much the same logic of presumptive valuation can be applied generally, ex post, to assess the financial consequences of any decision involving an application of judgement to prefer one policy against another or change certain elements of the existing policy. Given the uncertainties involved, policies may often deliver less than expected. In fact, across contexts and places, very rarely do ex post assessments of benefits of a public policy exceed ex ante estimates. Such implementation gaps are a feature of most public policy decisions. While this is the main channel through which auditors contribute to decision paralysis, they are no more immune from the usual corruption that afflicts other agencies and this too engenders decision paralysis. It is well known that, while the top offices of the institution have been of impeccable integrity, the Indian Audit Department under the CAG has its own share of corrupt officers and the CBI has periodically arrested CAG officers. Corruption is believed to be fairly widespread in the audit of areas that are themselves prone to corruption, like revenue audit and audit of works departments. There is a documented case of an audit party openly demanding, in writing, free phones and talk time from a public sector telecom company even for its watchmen, drivers, and cleaners for a full year. When auditors visit offices for their periodic audits, many junior officers feel they should be ‘taken good care of ’ or else they can raise difficult audit objections, which can linger interminably. This typically means at the least exemplary hospitality—​accommodating them in the best hotels in mofussil towns, transporting them in the best vehicles, providing them with sumptuous meals (and often drinks), and taking them to the nearby prominent sight-​seeing locations and temples. Newly recruited IAS officers working as sub-​collectors have vivid experiences of their first CAG auditor visit, with the Head Clerk’s hesitant approach carrying a request to the new officer for approval of such ‘special’ requirements. In recent years, the CAG has taken very commendable steps to discourage these practices.

172  State Capability Challenges and Weaknesses

8.6  Finance Function In the public bureaucracies, at both central and state governments, Ministries formulate schemes and proposals, and then seek the concurrence of the Finance Ministry (central) or Department (state) (loosely referred to as ‘Finance’); depending on the level of funds involved, approval of the Cabinet or chief/​prime minister. As would be expected, given the scarce resources and large competing demands, Finance often cuts down on programme allocations. This is obviously a legitimate function of the Ministry.

Decision Paralysis  173 In many cases, Finance goes beyond fixing an allocation ceiling and expresses views on programme economics (a unit rate compensation of Rs. x instead of Rs. y), procurements (why not through the local government agency or Self Help Groups, SHGs), financials (why not leverage resources from corporate social responsibility funds or Public Private Partnerships, PPP), contracting strategies (one model of PPP over another), or manpower deployment (one administrative structure over another, x number of people to do a task instead of y). Some of these prescriptions are often drawn from generalizations or thumb rules or ‘best practices’.

There are pros and cons to this. In favour of it is the fact that administrative departments may not always come up with fully thought-​through

174  State Capability Challenges and Weaknesses or well-​designed proposals. Finance, with its experience of dealing with multiple departments, may often have a wider perspective and be better able to spot opportunities for economies of scale or to identify models that have failed. On the other hand, officers in Finance with limited experience in the subject matter may go too far and prescribe too much or prescribe something not implementable or not sensible. Such prescriptions often assume the de facto force of a veto. Finance can theoretically be over-​ruled, but it is difficult in practice. The manifestations of sub-​optimal prescriptions of this kind are often felt during implementation and can become the difference between success and failure. But the other side of this is that Departments generally prefer to have the crutch of a Finance approval before going ahead with their expenditure decisions. It provides reassurance against potential audit objections and vigilance investigations that Departments would be loath to forgo. The following is a real-​life example encountered by one of the authors early in his career in the 1990s. The Public Works Department of a State sought concurrence of Finance to buy new office clocks. Finance returned the file saying that ‘purchase of office furniture of less than Rs. 10,000 is within the Department’s powers’. The PWD replied stating that ‘clocks are not furniture, hence Finance’s concurrence is sought’. The Finance Department then replied pointing to the printed list of examples under the definition of ‘office furniture’ which explicitly included clocks!

The Ministries in Government of India follow the Integrated Financial Advisor system, which combines both internal financial management as well as external expenditure oversight responsibilities. Accordingly, all Ministries and Departments have an Integrated Finance Division headed by an Additional Secretary & Financial Advisor (AS&FA), who reports to both the Secretary of the Department concerned as well as the Secretary, Department of Expenditure, Ministry of Finance. The idea behind this approach is to advance the good causes of the Department or Ministry without violating financial propriety. A good financial controller should not only stop bad expenditures but also find a means to allow good expenditures. Doing the latter requires a good understanding of the General Financial Rules (GFR), especially its

Decision Paralysis  175 various exception provisions. Subordinate staff in the Integrated Finance Divisions typically emphasize the ‘compliance’ aspect more than the Ministry’s mission. Without a good knowledge of the rules, Financial Advisers often struggle to use the panoply of opportunities within the GFR to constructively advance the cause of the Department. Many Financial Advisers have limited previous exposure to the financial rules. Secretaries have the right to over-​rule their Financial Advisers, but few actually do so.

9 Challenges in Execution This chapter begins with a broad discussion about the challenge of implementation or execution in India’s context and the paradox of outstanding examples of institutional success in execution amidst pervasive state failures. It then discusses the challenges of balancing strictness in enforcing the law against the associated likelihood of harassment or corruption and concludes by examining the trade-​off between coordination, which demands a strong convening centre, on the one hand, and empowerment of line departments and field functionaries, on the other.

9.1  Execution in Perspective The most critical deficit for India’s long-​term sustainable development is neither its fiscal nor current account deficits. Instead, it is arguably its (less discussed) implementation deficit. It is possible to have the best structured and incentive-​compatible policies and yet not have much real impact in addressing the underlying problem. One can change structures of personnel (e.g. lateral entrants) and management (e.g. results framework) and still achieve only limited change. For example, in terms of the conceptual basis, analytical frameworks, and incentive compatibility, there is little to criticize in the design of the National Health Mission. It is state-​of-​the-​art, based on evidence, and worthy of emulation as a policy. But the delivery of health care outcomes, across its entire spectrum of activities, could have been much better. In the context of environmental regulations, Sunita Narain of the Center for Science and the Environment, a New Delhi-​based think-​tank, puts it aptly1: 1 ‘Why India Is One of the Most Polluted Countries on Earth?’, The Economist, 6 December 2018 (https://​www.econom​ist.com/​asia/​2018/​12/​08/​why-​india-​is-​one-​of-​the-​most-​pollu​ted-​ countr​ies-​on-​earth) (Accessed on 16 December 2018). State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0009

178  State Capability Challenges and Weaknesses We have all the institutions and a lot of the right laws, but where is the actual capability, the personnel, the tools? At the level of outcomes, the implementation deficit manifests in the form of poor student learning outcomes, poor maternal and child health outcomes, malnutrition, low firm and farm productivity, weak enforcement of regulations, poor quality of infrastructure and public services delivery, and so on, not to mention the persistence of poverty and lower value for money from public expenditures. At the operational level, the implementation deficit manifests in the form of widespread absenteeism of public functionaries, rampant harassment and/​or corruption in accessing basic public services, incomplete or inordinately delayed infrastructure projects, weak implementation of regulations, poor supervision, deficient accountability, and so on. Discussions on the diagnosis of the implementation deficit have largely been confined to the conventional stereotypes—​politicization, corruption, and the general ‘inefficient’ nature of government officials and bureaucracies. Needless to say, the solutions have accordingly been along expected lines—​outsourcing, privatization, consultants, digitization, and so on. The reality is far more nuanced. For sure, the conventional wisdom on the diagnosis is relevant. All those are real problems and need to be addressed. But the problems go much deeper. So, even where these stereotypes are addressed, state capability may remain weak. In fact, prima facie attractive solutions like those using technology can sometimes come in the way of thinking hard about the real challenges of implementing programs where the technical limitations of technology get exposed and where there are deep human behavioural change dimensions. For example, as we shall see in Chapter 13, the use of digital dashboards for supervision can offer the illusion of control without addressing the problem. James C. Scott, in his classic book, refers to the urge to simplify and undertake administrative ordering of the problem context and the high-​ modernist ideology where the neatness of science and technology trumps attention on all other factors.2

2 James C. Scott, Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed, Yale University Press, 1998.

Challenges in Execution  179 Take the example of irrigation. Successive governments have spent trillions of rupees on irrigation projects, with disproportionately lower impact compared to the actual public expenditure.3 Irrigation projects are among the simplest technically: construct a dam and reservoir to hold water, and then build a network of canals to distribute it to farmers. But irrigation structures like large and small dams and water storage units, by themselves, contribute little to irrigation, unless complemented by canals and field channels. In fact, even among canals, the outcome is realized only with field channels that reach the farm. Contractors, though, are more interested in the dams and storage and, at the most, the larger upstream canals. They find the single-​location, regular construction work associated with dams easier and more attractive than the right-​of-​way acquisition problems and other transaction costs that characterize widely spread canals and channels. In the case of larger projects, the components are given as separate contracts, and it is commonplace to find the dams and other large engineering structures in place without the canals. Similar incentives drive government stakeholders. Irrigation departments, most often interested only in contracting out works, push projects, making unrealistic irrigation coverage estimations. In cases of projects done with assistance from the Government of India, state governments routinely make over-​optimistic irrigation coverage claims to get project approvals. The process of acquiring right-​of-​way for downstream canals and field channels, which involves engaging and negotiating with local landowners and resultant transactional challenges, can be very challenging. Then there are also the issues of excess water drawals and wastage by those in the higher reaches. It is therefore no surprise that dams have been built with vast differences between the promised and actually delivered irrigation coverage.4 In fact, an audit of the original estimate of the irrigation potential and the realized coverage from all major and 3 Robert Chambers and Robert Wade, ‘Managing the Main System: Canal Irrigation’s Blindspot’, Economic and Political Weekly, 15, no. 39 (September 27, 1980) (https://​www.jstor. org/​sta​ble/​4369​122?seq=​1) (Accessedon 3 October 2019). 4 Robert Wade, ‘Access to the Irrigation Department: The Tail-​ End Problem Under South Indian Irrigation Canals’, Indian Journal of Public Administration, 26, no. 2 (April 1, 1980) (https://​journ​als.sage​pub.com/​doi/​abs/​10.1177/​00195​5611​9800​207) (Accessed on 3 October 2019).

180  State Capability Challenges and Weaknesses minor irrigation contracts across the state and central governments is most likely to reveal massive gaps. While there are no easy answers to the problem, one possible strategy would be to package construction contracts as irrigation contracts rather than dam construction contracts. Instead of constructing a dam, the terms of reference in the bid should be clearly redefined as ‘creation (or stabilization) of 5000 Acres’, with levels of water access. This would require much closer engagement among the contractor, irrigation officials, and the beneficiary farmers and far more rigorous project preparation work. The project reports for each structure would have to evolve bottom-​up, capturing local requirements and practical considerations so as to ensure that outcomes remain at the centre of the project. However, it is also likely that, given the uncertainties associated, contractors may refrain from bidding for such projects. The risks associated with not getting the coverage estimations right can be mitigated by appropriate safeguards with respect to upstream water availability. But would contractors be willing to assume the associated risks? Even routine activities of the government like running schools and hospitals face similar challenges. In all these cases, there are multiple functional activities involving both horizontal and vertical coordination and covering large numbers of officials. It is impossible to tightly script all these interactions and tightly prescribe the rules of the interactions. Even when scripted, it is no guarantee of actual execution to desired standards. Even apparently simple monitoring of activities poses practical problems of credible measurement, data capture, and reporting of actionable information. In activities where quality is of essence (a teacher’s engagement in the classroom, doctor’s treatment of patients, engineer’s assessment of quality of work, sanitary inspector’s monitoring of cleanliness, etc.), capturing performance in a satisfactory manner is among the biggest challenges.

9.2  The Systems Approach to Public Service Delivery Systems change has become a buzzword in development today. Its recent origins have been in the philanthropic sector and motivated by external

Challenges in Execution  181 The State (Executive) Politicians

Policy Makers

P (D oliti , F, cs I, M )

C (D ompa , F, c I, M t )

Citizens/Parents/Students Client Power (D, F, I, M)

Coalitions/Inclusion Rich/ Privileged

Poor/ Marginalized

Management (D, F, I, M) Frontline providers Organizations (teachers, (ministries, principals, schools) headmasters)

Instructional services (from teachers to students)

Figure 9.1  Systems approach to public services delivery Source: Lant Pritchett, Presentation on System approaches in education, 14 May 2018 (https://​ www.bsg.ox.ac.uk/​sites/​defa​ult/​files/​2018-​06/​Pritch​ett_​syst​ems%20app​roac​hes%20in%20Ed​ ucat​ion.pdf) (Accessed on 28 December 2018)

agents’ desire to change the ‘system’ so as to address particular development problems. Systems change involves understanding the root causes of the problem and its context, finding solutions, and engaging in their iterative implementation. All this is done as part of a conscious and multi-​dimensional process. The presumption is that ‘wicked problems’ in development cannot be solved with piecemeal interventions, however innovative, and demand simultaneous system-​wide interventions to change both institutional processes and behaviours and attitudes of personnel. The World Development Report 2004 drew attention to the importance of a systems approach to examining public services delivery (Figure 9.1).5 It identifies the different individual and collective stakeholders involved. It then locates the multiple principal–​agent accountability relationships 5 World Bank, World Development Report 2004, Making Services Work for Poor People, 2003 (https://​openkn​owle​dge.worldb​ank.org/​bitstr​eam/​han​dle/​10986/​5986/​WDR%202​004%20-​ %20Engl​ish.pdf) (Accessed on 28 December 2018).

182  State Capability Challenges and Weaknesses and their design elements. Finally, it defines how the relationships and design elements interact. Public service delivery to citizens involves four important principal–​agent relationships—​citizens to the state (political); state to organizational providers (compact); organizations to frontline workers (management); and citizens to frontline workers or organizational providers, as relevant (client power). Each of the four accountability relationships has four design elements—​delegation (define what principal wants from the agent), finance (payment to the agent), information (on the performance of the task), and motivation (compensation for performance). Importantly, the performance of the systems is critically dependent on how the design elements of each interact to ensure tight accountability in the four relationships and how the four relationships themselves interact with each other. Accordingly, a systems assumes the following theoretical framework—​ the principal entrusts a task to an agent, who is then paid to perform the task; and the agent does the task, which is assessed for how well it has been done; and the agent is incentivized accordingly with a reward for good performance and punishment for bad. In public systems, all the four relationships and their interaction with their respective design elements struggle to establish accountability. The strength of electoral accountability, the ‘long route’ accountability, is very loose, even in mature democracies. The challenge of contracting quality and other poorly quantifiable indicators of service delivery, lack of clarity with objective, and also the weakness of state capability makes compact accountability tenuous. The state capability deficiencies weaken the management accountability. Finally, the low level of decentralization and attendant poor empowerment of frontline officials coupled with poorly defined objectives diminish client power considerably. Effective service delivery demands systemic coherence at three levels—​ within a relationship of accountability; between the same element across relationships; and between relationships. Take the management relationship in most departments—​goals without the requisite autonomy over the use of resources to perform the task; no credible measurement and monitoring system; and disconnect between performance assessment and structure of compensation. An example of the second level of incoherence is the difference in the messages communicated in each relationship.

Challenges in Execution  183 So different stakeholders carry back different, contradictory, or unrelated messages about the goals and performance of the system. Finally, agents intersect with multiple relationships, and the design elements in them can be vastly different, thereby creating problems of aligning the agent’s incentives and realizing systemic coherence. The result is that the theory stumbles. The agent does the task badly, its assessment is practically difficult, and incentivization is misaligned. All of development is an effort to generate system coherence to drive change. But the complex requirements make an effort very difficult. However, instead of ‘boiling the ocean’, the conceptual framework points to a few headline enablers that are essential for the success of any attempt at systems transformation. First and foremost, there has to be clarity of goals across all four relationships and those goals have to be aligned with each other. Second, agents at each level have to be equipped with the resources, powers, and capability to be able to do the task entrusted to them. Third, the principals at each level should be able to credibly measure and consistently assess how well the agent has done the task. Finally, principals should have the power to reward or penalize agents based on their performance. Assuming the requisite political will, the biggest challenge with getting the execution capability right will be at the level of frontline functionaries. They are central to the strength of both the management and client power relationships.

9.2.1  Problems at the Frontline In the vast majority of cases, policies are ultimately implemented by frontline or street-​level bureaucrats. The citizen or corporate experiences the policy or program in terms of how it is implemented by the frontline official. This makes the accountability and empowerment of frontline officials critical to any systemic transformation efforts. Unfortunately, functionaries at the cutting edge, across sectors, are overburdened and under-​resourced and generally feel disempowered. Consider the plight of important frontline managers like a Block Education Officer or Sub-​Divisional Medical and Health Officer. While both are responsible for the execution of education and health activities

184  State Capability Challenges and Weaknesses in their respective jurisdictions, they are heavily overburdened and have hardly any influence on program design. They inherit an implementation plan, most often without clearly defined objectives, which is transmitted from the state government or district administration, and which offers limited flexibility for local adaptation. They have sorely limited resources, physical or human, to even execute their mandate effectively, much less embrace any innovation. This results in the following: • Dysfunctional management: Even where delegation and finance are addressed, itself debatable, organizational providers struggle to credibly assess and incentivize performance. So they resort to top-​ down indicators and limit discretion. The result is disempowered and demotivated frontline functionaries. • Dysfunctional client power relationships: None of the four design elements hold—​people do not entrust the task, do not pay for it, cannot assess outcomes, and cannot manage performance. The result is unaccountable frontline functionaries. This creates a paradox. The Indian state has proactive disclosure standards under the Right to Information Act, social audits, decentralized planning, door-​step public service delivery, use of technology for e-​governance applications, legally enshrined rights-​based entitlements, and so on: all very progressive signs of accountability. But does this form of accountability translate into the substance of accountability? To answer this, it is worth examining the two dimensions of accountability within an administrative system—​the top-​down and bottom-​up accountabilities. The top-​down accountability typically involves lack of trust. The policymakers distrust the implementers, and the higher-​level implementers distrust those at the frontline. Limiting discretion in implementation is therefore commonly regarded as a touchstone for a good policy. The corollary to this is disempowerment, which flows in the reverse direction from the bottom to the top. The frontline officials feel disempowered at not being able to improvise and adapt the program or the intervention to the requirements of the local context. Interestingly, this sense of disempowerment is increasingly being felt by even those at the top levels of the bureaucracy. There is a case to be made the while the

Challenges in Execution  185 disempowerment of those at the bottom is dictated by real fetters placed on their operational discretion, that of their bosses is more a rationalization driven by abdication. This creates a very convenient and self-​reinforcing equilibrium for the overwhelming majority of participants. Those at the top levels of government, either as policymakers or program implementers, are comforted by having placed sufficient guardrails and restrictions to limit discretion, which ostensibly ensures things do not go wrong too badly. For the well-​ intentioned among them, there is an illusion of control. For the not-​so-​ well-​intentioned, there is the assurance that they can easily subvert the process to realize their objective, all the while retaining the form. The lazier and laidback ones are happy to let things play themselves out once the guardrails are in place. While this frustrates the minority of committed frontline officials, for the majority, the status quo provides a convenient excuse for inaction and apathy.6 In all the interplay of deficient trust and feeling of disempowerment, the substance of accountability is lost. The system gets stuck in a sub-​optimal but bureaucratically and politically acceptable equilibrium. Fundamentally, sustainable transformations cannot be realized just through rigorous monitoring and require the full ownership of programs by frontline managers and implementers. There are at least three reasons for this requirement: • Activities that are human engagement intensive (called ‘thick’ activities by Lant Pritchett—​see Chapter 12) demand personal ownership by the agents. • Enforcement works at the margins; enforcement is effective when compliance is the norm and only the minority need to be forced into conformity. • High-​powered incentives work at the margins; as with enforcement, incentives work only when the majority are willing to play

6 Yamini Aiyar argues that India’s frontline bureaucracy is shaped more by ‘the attraction of government jobs as a source of mobility and power rather than the achievement of professional goals’; see: ‘How to Strengthen India’s Frontline Bureaucracy’, Hindustan Times, 11 October 2018 (https://​www.hin​dust​anti​mes.com/​colu​mns/​how-​to-​str​engt​hen-​india-​s-​frontl​ine-​bure​aucr​ acy/​story-​MP7​H1z7​otbY​362b​pLja​YzH.html) (Accessed on 30 December 2018).

186  State Capability Challenges and Weaknesses by the rules of the game and only a small minority are prone to manipulation.7 Such program ownership, in turn, requires empowerment and motivation. Empowerment requires supporting delegation and financial powers. Motivation requires a combination of both accountability to organizational providers and citizens, as well as engendering some form of intrinsic motivation. A fundamental transformation is required in the bureaucratic culture. This in turn has two key elements: greater emphasis on the qualitative dimension of accountability and increased discretion for field functionaries. These are discussed in Chapter 13.

9.3  The Paradox of Elections and Census In a recent paper, Devesh Kapur examined the paradox of why the Indian state both failed and succeeded at the same time.8 Contrasting the widespread stories of development failures, he points to the successes of elections, census, Aadhaar, Goods and Services Tax, and so on. He writes, The Indian state has delivered better in certain situations and settings: specifically, on macroeconomic rather than microeconomic outcomes; where delivery is episodic with inbuilt exit, rather than where delivery and accountability are quotidian and more reliant on state capacity at local levels; and on those goods and services where societal norms and values concerning hierarchy and status matter less, rather than in settings where these norms and values—​such as caste and patriarchy—​are resilient . . . The reasons, we argue, lie in the understaffing of local government, the precocious democracy of India and its anomalous sequencing of universal franchise, and India’s “societal failures” manifest in caste and gender discrimination.

7 Paul Milgrom, ‘Employment Contracts, Influence Activities, and Efficient Organisation Design’, Journal of Political Economy, 96, no. 1 (1988): 42–​60 (https://​eco​npap​ers.repec.org/​arti​ cle/​ucpjpo​lec/​v_​3a96_​3ay_​3a19​88_​3​ai_​3​a1_​3​ap_​3​a42-​60.htm)(Acc​esse​don 3 October 2019). 8 Devesh Kapur, ‘Why Does the Indian State Both Fail and Succeed?’, Journal of Economic Perspectives, 34, no. 1 (Winter 2020) (https://​pubs.aea​web.org/​doi/​pdfp​lus/​10.1257/​jep.34.1.31).

Challenges in Execution  187 So, amidst the several government failures and state capability weaknesses, there are some standout exceptions. Consider the Election Commission of India (ECI), which can rightfully stake a claim to be arguably the most efficient bureaucracy, public or private, anywhere in the world. What explains the success of the ECI? As of January 2014, the ECI had a sanctioned strength of just 641, of which 276 posts were vacant.9 In other words, the ECI conducts the world’s largest democratic elections, perhaps one of the largest logistical exercises in the world, with a mere 365 full-​time people! And this number has hardly changed over time—​327 and 319, respectively, in January 2006 and 2010. Yes, it relies on the entire central and state government machinery to conduct the electoral process. But given the highest fidelity in execution required, zero-​defect as they say, this is truly a staggering achievement. And the electoral process is not just about holding the elections itself; it is about the always-​on preparatory process like the updation of electoral rolls. This is a massive exercise in managing and coordinating among numerous stakeholders. This is perhaps the most efficient leveraging of external resources in any organizational setting anywhere in the world. It is an example of the paradox of India’s state capability that it can execute well-​defined tasks like elections, census, disaster relief, etc., with unparalleled proficiency and yet do simple and regular things like running mid-​day meal kitchens in the most appalling manner. How is it that Indian bureaucracy delivers spectacularly with programs like Pulse Polio, elections, and Kumbh Mela? What explains the handful of outstanding examples of successes with improving maternal and child health outcomes, or women’s empowerment, or increasing access to sanitation that occasionally come from various corners of India? There are two possible answers. One, and relevant to the first question, is that the state is exceptional at doing things that have a short duration and a clearly defined destination. A common thread that goes through all those activities is meticulous 9 Report of the Seventh Central Pay Commission Report, November 2015 (https://​www. fin​min.nic.in/​sites/​defa​ult/​files/​7cpc​_​rep​ort_​eng.pdf?downl​oad=​1) (Accessed on 21 December 2018).

188  State Capability Challenges and Weaknesses planning which maps functionaries to task and time, simple and clearly defined monitoring goals, rigorous supervision, and insulation from political influence. This is generally achieved through a massive mobilization of the entire local administration, often augmented with manpower from other areas. It helps that the short duration of these activities allows for such cross-​departmental mobilization. All this is underpinned by a strong bureaucratic (and in some cases, political) commitment at the highest levels to achieve the objective. The success with these activities stands in stark contrast to the egregious failure with implementation of regular government programs in the same area by the same personnel. It shines a light on the sources of their implementation failure. An apolitical, adequately staffed and funded, and appropriately trained bureaucracy, with clearly defined outputs and timeframe, when monitored rigorously, gets things done. All these ingredients—​the absence of political interference, adequacy of resources and appropriate training, clearly defined and monitorable tasks—​are either deficient or intrinsically not feasible in many government programs. Mundane activities like running mid-​day meal kitchens or distributing old age pensions to beneficiaries or delivering ante-​natal services to pregnant women appear insurmountable systemic challenges, mainly because these characteristics are absent. The examples of success in such activities are, most often, a result of individual initiative. A few fiercely committed officers bring great passion and effort into planning and implementing programs. Their energy and leadership mask the chronic deficiencies in state capability and results in positive deviance. There are a few features that inform the actions of all such successful officials. They make detailed plans, with clearly defined intermediate milestones and final outcomes with timelines, all of which can be monitored. They rely on some improvised information reporting system to rigorously monitor compliance. Some of the more ‘tech-​savvy’ among them use computer applications for internet-​based reporting. Others rely on brute force reviews, done with unfailing regularity. These reviews are generally done personally at a painstaking level of granularity to identify weaknesses and poor performers, which are then followed up and addressed. It is no accident that their reporting systems are generally at variance from the ineffectual routine state or nation-​wide reporting systems for the same program or activity.

Challenges in Execution  189 Another prong of their monitoring is intensive field inspections, most often done randomly and without intimation so as to validate the information from the aforementioned reports. Such personal inspections, which are often followed up with strong and certain disciplinary action on errant officials, are a strong deterrent against slacking. But given the large geographical jurisdictions, such personal inspections take up a disproportionately large amount of time and effort and involve unfavourable work trade-​offs. A handful of motivated and sincere subordinates are co-​opted as internal champions in these efforts. In a few cases, committed local non-​ profits become important partners. The smarter officers keep open active feedback channels—​ through inspections and interactions with field functionaries, citizens, and political representatives—​that reliably convey information on the actual implementation. Though all these are the physical ingredients of routine bureaucratic implementation, its actual realization in a typical bureaucratic environment is infrequent. It is commonplace that officials are severely handicapped by an enfeebled administrative machinery, whose weaknesses are exacerbated by inadequate, poorly trained, and disillusioned staff. Most often, they are left to fend for themselves, in completely unknown terrain, with just a handful of people for support. The level of direct personal involvement needed for success requires huge mental bandwidth for information processing and monitoring and a level of commitment that is understandably possessed by only a handful of very sincere and energetic officials. Their success is despite grave deficiencies in state capability and a triumph over extreme adversity. In short, such successes come from more-​than-​call-​of-​duty efforts, which are institutionally non-​replicable. In some cases, the systemic changes do endure, but reflecting the deeply personal nature of such interventions, many such successes are short-​lived, failing to outlive the officials themselves. Though the main focus here is on the role of All India Service (AIS) officers, much the same applies to other non-​AIS officials of the state government. The successful among them overcome much greater adversity than a similar AIS officer. The point is as much to highlight why things sometimes appear to work in certain places as to why its systemic replication may be difficult given our current state capability.

190  State Capability Challenges and Weaknesses A bureaucracy is an impersonal rules-​based organization of a group of people who work together to achieve certain defined outcomes. Bureaucracies therefore have to work on their systemic strengths rather than the personal initiative of individual bureaucrats. A capable bureaucracy is one that delivers outcomes when administered by the average bureaucrat rather than being reliant on the serendipitous presence of an exceptional bureaucrat. The typical Indian bureaucracy, at all levels, is most likely to end up short on this test. In the context of discussing exceptions, it is relevant to acknowledge the success of the Indian state with various large scale infrastructure and other projects like the construction of the large national highways network or the enrollment of citizens under Aadhaar project or opening of bank accounts under the Prime Minister’s Jan Dhan Yojana (PMJDY) or the construction of toilets under the Swachh Bharat Mission. There are specific institutional causes for these successes. The discussion in this book therefore does not dwell on these successes, but rather draws relevant lessons from them for addressing these more complex development challenges.

9.4  Coordination vs. Empowerment One of the defining features of large bureaucracies is the urge at leadership levels to be in control. This reflects the fundamental tension between control and discretion. Control requires being prescriptive and minimizing the exercise of discretion. It spawns one-​size-​fits-​all bureaucratic processes and programme implementation guidelines. Note that the desire for control is not necessarily perverse—​in fact, the public often expects the government to be in control (especially when things go wrong). So if a local police force fails to detect a terrorist plot, criticisms of the Union Home Ministry often follow swiftly. If there is a disease outbreak in a town, the public expects the state government—​ not the municipality—​to take charge. Parliament and legislatures hold Ministers responsible. Citizens love local empowerment—​except when it fails, in which case they want centralized control. The complex nature of the problems being sought to be addressed and the sheer diversity of contextual challenges demand the exercise of

Challenges in Execution  191 discretion to adapt implementation to suit the context. Further, delegation empowers frontline functionaries and thereby creates the conditions for state capability to develop. Public choice theory captures this dilemma in the principal-​agent framework. The principal is concerned about the transmission of information to agents in a large bureaucracy10: As a piece of information is passed from individual to individual, the content of the message becomes increasingly distorted. The magnitude of this distortion, or “noise,” is a function of the complexity of the information and the number of people in the transmission chain. As the complexity of the information and the length of the chain increases, so too does the magnitude of the noise introduced into the initial message as it is passed from person to person. The solution to this problem . . . is the decentralization of decision-​making to reduce the length of the transmission chain and the complexity of the information.

But such decentralization, in turn, generates perverse incentives among agents that come in the way of the principal’s ability to control outcomes, While decentralizing decision-​making reduces the length of the transmission chain, it also poses the problem of ensuring that the goals of the decentralized decision nodes are aligned with the broader aims and goals of the organization. As such, a complete solution to the problems associated with information deficiencies and incentive compatibility includes de-​centralizing decision-​making while simultaneously structuring the bureaucracy so that the actions of the decentralized decision centres align with the broader goals of the organization. For obvious reasons, this is an extremely difficult task in practice.

This dilemma has profound implications for the bureaucracy and political system.

10 Christopher J. Coyne, ‘The Politics of Bureaucracy and the Failure of Post-​ War Reconstruction’, Public Choice, 135 (2008): 11–​ 22 (https://​www.resea​rchg​ate.net/​publ​icat​ ion/​5154773_​The_​Politics_​of_​Bureaucr​acy_​and_​the_​fail​ure_​of_​p​ost-​war​_​rec​onst​ruct​ion) (Accessed 18 December 2018).

192  State Capability Challenges and Weaknesses In the case of the bureaucracy, the delegation of powers and procedural requirements are calibrated to ensure that the leadership is able to limit discretion at lower levels and retain adequate high-​level control over the implementation of various programs. It is feared that excessive delegation of authority and procedural discretion would incentivize officials to compromise on implementation and/​or engender corruption. Never mind that control rarely addresses these concerns and perhaps even increases its likelihood. But it provides a fig-​leaf for officers to not only cover their backs but also retain control over the administration for exactly those considerations. In India’s policy and program implementation context, there are structural factors that contribute to centralization and limits on the discretion of frontline functionaries. The most important is the role of IAS officers at leadership positions in the districts and in departments. The fact that they are (by the very nature of the respective recruitment processes) more likely to be higher capability individuals compared to the line department heads perhaps contributes to their dominance. Besides being the head of the district administration, politicians have found it administratively more convenient as well as practically more effective to rely on the District Collectors. The weak state capability at the frontline of implementation only exacerbates this reliance. A local legislator realizes that his or her work is more likely to be done with the Collector’s direct intervention. Finally, this reflects the power dynamics at the state level, with the dominant role of the Chief Ministers. The cliché that the only three people who matter in India’s government are the PM, CM, and DM is not totally inaccurate. Accordingly, young IAS officers posted in various field positions—​Sub Collectors, Project Officials of various agencies, Municipal Commissioners, CEOs of Zilla Panchayats, District Collectors, etc.—​ possess unparalleled authority. Globally, there are few comparable examples of such across-​the-​board power vested in unelected bureaucrats, especially at such very early stages of their careers. For sure, given the extremely weak state capability, on average, things work reasonably well with this approach due to the enormous energy and commitment of young officers working in these posts. However, it is a patchwork solution.

Challenges in Execution  193 Given their youth and the attendant energy and passion, coupled with the weakness of the system they administer, these officers centralize authority and exercise it aggressively. Most often they function virtually as heads of departments and directly exercise line departmental authority. It is a very rare officer who steps back and empowers the line department heads or gives them full autonomy. Rarer still are those who allow departmental heads to take critical decisions and stand by them in good times and bad. All this is not to gloss over the self-​inflicted apathy, ineptitude, and corruption among departmental officials themselves. For even if the departmental officials are empowered, most probably they would be reluctant to assume responsibilities, defer decisions, and continue to await instructions from the top, and, given the circumstances, they are unlikely to be any more effective. This, at least initially, would be the most likely typical response. It does not help that there is some crisis or the other all too often within each department and the Collector’s active engagement becomes critical to defusing the situation. The campaign mode of program delivery of central and state governments centralizes authority even more in the hands of the District Collector. Realizing the enormous convening power of the District Collector, departments at the state level prefer to engage with them rather than their own line functionaries on important issues. Consider a District Education Officer or District Health Officer. Each has the responsibility of managing a jurisdiction with a population of 1–​ 2 million and thousands of teachers and health personnel. They are arguably the equivalent of the Permanent Secretary of Education (or Health) of a small country. How many of them are empowered to act independently? After a career of subservience and taking orders, how many of them still retain the initiative to act independently even if empowered? And one is not even talking about the Block Education and Health Officers, not to mention the field inspectors or supervisors in either department. These power dynamics are mirrored at the various Line Departments and Ministries of state and central governments. The Directors or Commissioners and Secretaries, generally being IAS officers, exercise the same level of centralized authority over their subordinates. Ministers too find it convenient to rely on their Directors and Secretaries to get work

194  State Capability Challenges and Weaknesses done. Only in cases of extraneous considerations do they go over their heads and engage with the departmental officials. In all these cases, Municipal Commissioners, District Collectors, Heads of Departments, and Secretaries, the limited delegation of powers means that the subordinate officials have limited powers to take a decision on any important operational issue. The culture of looking up to the bosses invariably follows. The immanent signals of authority get quickly entrenched and shape the actions of all stakeholders. Emergent trends have only amplified these dynamics, in particular, the practice of campaign mode implementation, especially for those which require considerable public outreach for social mobilization. Accordingly, the state or central government announces a mission-​mode program with tight implementation timelines. The entire administrative machinery of the district is mobilized to support the implementation of the program. Supervision of administrative units—​blocks or Gram Panchayats—​are entrusted to a Nodal officer, a senior officer drawn from across departments. At the cutting-​edge, under the leadership of the relevant line Department (for the particular programme) all the other departmental officials are mobilized to execute the campaign. As a practical approach, it has been relatively successful in achieving the desired immediate objectives. The Departments acting alone could not be expected to realize such targets in such short times. But it also comes with certain costs. For a start, it creates the conditions for the erosion of line departmental accountability and lets afoot the moral hazard that others will be there to bail out with campaigns. It also weakens the Departmental leadership and makes the District Collector directly responsible for implementation. Further, the inherent nature of such campaigns, with their immediate objectives, not only papers over departmental weaknesses but also exacerbates them. This power dynamic has its equivalent in political terms in the reluctance of state-​level politicians to decentralize powers. Accordingly, state governments, either directly or through their agents at the district or city levels, retain control over important administrative and financial matters of all departments. Despite the passage of the 73rd and 74th Constitutional amendments on decentralization of powers to rural and urban local governments, many state governments have not ceded

Challenges in Execution  195 meaningful control over funds, functions, and functionaries to local governments. Like the officers, politicians may cite the same excuse: corruption and chaos if power is decentralized to ill-​equipped local governments. But there is a perception that the real reason may be the fear of losing control over important political constituencies and (in some cases) the patronage and rent-​seeking opportunities that come with retaining power at the state government level. The circularity of this argument in both cases cannot be missed—​weak state capability prevents delegation and decentralization, and insufficient decentralization prevents the acquisition of capability and accountability.

PART IV

POS SIBL E SOLU T IONS

10 Addressing Policymaking Challenges—​I In Chapter 7 we attempted to identify the shortcomings in India’s policymaking processes. This leads to the question: What can we do to improve policymaking?

The issues highlighted in the analysis on problems with policymaking can broadly be divided into three types. The first of these is structural—​ too much fragmentation, too much implementation workload on policymakers, and poor structure and process for involving outside experts and stakeholders. The second kind of problem lies with the competence and management of the people who man the structure—​inadequate professionalism of the policymaking staff, inadequate competence of the specialists, cadre management, promotions, etc. The third kind of problem concerns the issues with the policymaking processes—​decision paralysis, the uncertainties associated with complex policies, etc. Accordingly, this chapter makes some suggestions on each of these three problems. It also discusses two of the commonly suggested reforms, on new public management (NPM) approaches and performance-​based pay. The final section outlines certain insights into policymaking on complex issues that can be a useful guide for officials when they engage with policymaking.

10.1  Structural Reforms 10.1.1  Reduction in Fragmentation It was observed that, a priori, there are both benefits and disadvantages from fragmentation. Broadly, the benefit is specialized knowledge while the demerit is weaker coordination and integration. State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0010

200  Possible Solutions Having concluded that the present level of fragmentation is excessive, the question that arises is, how to go about reducing it. Applying the principles described in Chapter 8, the first reform would be to achieve a progressive decrease in fragmentation as one goes up the hierarchy. This would mean fewer Secretaries, each of whom would handle more than one of the existing sectors. The result would be that coordination and integration will be achieved far more smoothly. When deciding which portfolios to ‘broadband’, the second principle—​the degree of interconnection and overlaps between sectors—​would be the guide. Figure 7.1 below is a schematic diagram of the proposed structure. It stands in comparison to Figure 7.1 in Chapter 7. In Figure 10.1, at the higher levels of government, there is a progressive reduction in the number of ‘compartments’—​denoted by fewer radial lines. The Government of India has already taken steps in this regard, merging the Ministry of Overseas Indian Affairs with the Ministry of External Affairs and the Ministry of Drinking Water and Sanitation with the Ministry of Water Resources. Besides it is also worth considering merging the separate Departments within Ministries into a single Ministry without departments. Many Departments of the Government of India can just as well be managed by a Joint Secretary level officer reporting to the Secretary of a larger Ministry.

Figure 10.1  Proposed structure for policymaking process (Source: T.V. Somanathan and O.P. Agarwal, Public Policy Making in India: Issues and Remedies, ResearchGate, February 2005)

Addressing policymaking challenges—I  201 In fact, Ministries with a largely coordinating role should learn from the model of the ECI. As discussed in Chapter 9, the ECI manages to coordinate effectively with states to manage the massive electoral process with just 365 people. Ministries like Agriculture and Tourism, whose direct role is limited, could do well to focus their resources on more effective coordination and support for state governments. A similar rationalization is desirable at the level of state governments too. Indeed, it is perhaps even more significant at the state level, given the greater scarcity of talent to lead important positions at Secretariat and Directorates.

10.1.2  Separating Policymaking From Implementation The proposal to reduce fragmentation invites the question: How will the Secretaries cope with such enhanced responsibilities when they are already overworked? It is true that senior-​level civil servants in the Government of India appear to be constantly overworked. But this is mainly because of: • the heavy burden of day-​to-​day administrative (implementation) work, which occupies far more time than thinking on policy issues • the high degree of centralization of administrative powers. The proposed reform is that implementation responsibilities should be entrusted to Boards or Agencies, headed by a Director-​General, in the rank of Joint Secretary or Additional Secretary. The Director-​General would have full powers of decision-​making on non-​policy issues. There are existing examples—​the National Highways Authority of India (NHAI) in the Ministry of Road Transport and Highways, the Director-​General of Hydrocarbons in the Petroleum Ministry. While the Director-​General’s primary responsibility would be implementation, he would also provide essential inputs for policymaking. He would, thus, be a bridge between policy and implementation. The Secretary will be responsible for policymaking and have no implementation responsibilities. He would only get feedback on the progress of implementation, largely to aid future policies or to correct existing policies. Files on individual implementation

202  Possible Solutions decisions should not go to the Secretary and (where Minister’s orders are necessary) go directly to the Minister. Nor should the Secretary attend meetings on implementation issues. This will not only release Secretaries from their excessive routine workload but also give policymaking the focus it deserves. This change is depicted in Figure 10.1: the thin lines separating policy and execution are replaced by thicker lines, denoting a much stronger separation of execution and policymaking. However, there are pitfalls in completely isolating the Secretary from implementation. The flow of information and policy-​relevant ideas can be weakened if the policymaker is not also the implementer. Secondly, lack of authority over current implementation can, in the real world, lead to a perception of diminished ‘power’ with an attendant downgrading of the importance of the policymaking function. There is a way around this: the Director-​General should report to the Secretary and her/​his annual performance appraisal should be carried out by the Secretary. This should ensure that the Secretary continues to have access to information and that the policymaking role is not seen as a secondary or unimportant one. The restructuring proposed above—​fewer departments—​could admittedly pose a problem in accommodating a large council of ministers. One possibility is to have a Deputy Minister or Minister of State overseeing the Boards/​Agencies either independently or reporting to the Cabinet Minister. In fact, such separation of the policy advice function from the implementation or service delivery function has been a key ingredient of governance reforms in the United Kingdom, Australia, New Zealand,1 Malaysia,2 and other countries.

10.1.3  Decentralization/​Enhanced Delegation of Authority A very widespread tendency in Indian public administration, across departments and tiers of government, is a high degree of centralization of 1 Commonwealth Secretariat, Current Good Practices and New Developments in Public Service Management: A Profile of the Public Service of New Zealand, 1995a. 2 Commonwealth Secretariat, Current Good Practices and New Developments in Public Service Management: A Profile of the Public Service of Malaysia, 1995b.

Addressing policymaking challenges—I  203 authority on financial and other matters at higher levels. This is found at the district level (where power is centralized with the Collector or other district functionary) and in the state and central Ministries. In this section, solutions to this problem are discussed separately in the Secretariat and District contexts. (While delegation is discussed here in the context of policymaking, it is equally potent in improving execution.) 10.1.3.1 Delegation in Central/​State Secretariats and Departments While centralization is an old tradition, a recent change has been that more and more decisions are taken by large inter-​departmental committees chaired by very senior officers. The logic is usually as follows: 1. the lower levels are less competent; 2. the lower levels are more corrupt; 3. committees will bring collective wisdom (more heads are better than one); 4. central authorities can better resist improper local pressures; 5. since even senior officers may be corrupt, committees will reduce the risk of corruption; and 6. therefore in the interest of quality of administration and to reduce fraud and corruption, decisions must be taken at higher levels and by committees. There is some truth in these considerations, although with the prevailing levels of corruption among senior officials in the states the assumption that lower levels are more corrupt is probably no longer valid. However, centralization also has disadvantages. First, it removes the locus of the decision further from the locus of the problem; by making the decision-​taker more remote it reduces empathy, knowledge of field reality, and the ability of the citizen to be heard by the actual decision-​taker. Second, it makes decision taking slower. Third, it facilitates centralized corruption: if, say, transfers of teachers or award of contracts were handled at the block level, a corrupt minister or Secretary (either acting on his own or in response to political interference) would not be able to fully ‘control’ the transfers or contracts and hence may not be as successful in extracting bribes. Some local politicians may be corrupt, but the amounts involved are likely to be much lower and decisions would be taken much faster

204  Possible Solutions and with a greater voice for the affected parties. Committees, instead of preventing corruption, may provide the corrupt with ‘safety in numbers’ and embolden the taking of a corrupt decision which (if it was left to an individual officer) one may not have been willing to take. Given that corruption and fraud are (unlike in the 1950s) not merely lower-​level problems but may also exist at the top, it is time to do away with the automatic preference for centralization of financial, personnel, and procurement powers. The benefits of greater delegation of authority will outweigh the drawbacks in most circumstances. Therefore, a systematic ‘zero-​base’ review of delegation of powers across a wide range of functions is recommended, with a bias towards increasing delegation.

Addressing policymaking challenges—I  205 But such delegation will have to be accompanied by associated accountability mechanisms. One important requirement could be the introduction of performance appraisals for the lower civil services and their integration into promotion and other important service decisions. Indiscipline and poor work culture at lower levels is difficult to remedy and a major obstacle to governance reform. There are several levels, especially in state governments, where there is no formal performance evaluation. Some form of performance evaluation needs to be present at all levels, even for ‘Multi-​Tasking Staff ’ (office assistants) and drivers. This will, if nothing else, help discipline and productivity. At clerical levels, it could be input into promotion decisions. For example, if promotions from one clerical level to the next were based on, say, 80 marks for seniority and 20 marks for appraisals, there might be a greater degree of responsiveness; at the same time, the scope for arbitrariness and favouritism would be constrained since the weight of performance would only be 20%.3 10.1.3.2 Delegation in District and Local Administration Any meaningful attempt to address chronic state capability problems requires strengthening frontline functionaries within various departments; it is critical that frontline functionaries be nurtured to find their feet, become empowered, and lead and not merely follow. The tension between the need for coordination and the benefits of empowering field functionaries, referred to in Chapter 7, needs to be addressed. A broad approach to this is to centralize decisions on the process by which decisions will be taken but delegate substantial authority for decisions within that process. For example, the Collector would decide the steps involved in sanctioning benefits under a scheme or programme and make this uniform across the district; but individual decisions would be substantially delegated to lower levels with only a few individual cases coming to the Collector for approval. This is broadly reflected in Table 10.1. Given that Indian Administrative Service (IAS) officers are in departmental leadership positions at sub-​state, state, and national government levels, it is of paramount importance that they are sensitized 3 Strengthening performance appraisals of senior civil servants has been touched upon earlier as part of the discussion on changing promotion policy.

206  Possible Solutions Table 10.1  Delegation of Powers and Control Over Process Trade-​Off Matrix Control vs. levels of delegation

Levels of delegation of authority

Process control Low

High

Low

Worst of all worlds—​ demotivated staff/​poor implementation and limited control over outcome.

Prevailing arrangement—​ high degree of control. Low level of initiative by staff.

High

Bureaucrat’s nightmare—​loss of control. May promote innovation and be motivating for lower levels, but higher risk of embarrassing failures

Best outcome—​keeps people motivated and creates conditions for building state capability while controlling quality of execution.

about the importance of strong line departments. The young IAS officers posted in the field should be made to realize the importance of this. It is for training institutions like the Lal Bahadur Shastri National Academy of Administration (LBSNAA) and the State Administrative Training Institutes (ATIs) to help officers internalize this insight. The tacit emphasis on the ‘heroic individual officer achieving spectacular results’ as the role model in the LBSNAA needs to be replaced by a recognition of cases where officers have built systems and processes and developed the line managers to deliver. Collectors ought to resist the temptation of becoming de facto heads of line departments. They have to realize their role should be to support heads of departments to perform their functions. They should be leaders of a team, stepping in as required to guide, capacitate, coordinate, and monitor. This requires a very different mindset from the current direct execution approach, internalizing the importance of building systems and the value of team spirit. Contributing to team strengthening should become a norm for IAS officers, acknowledged both in peer groups and, if possible, in performance assessments. The 73rd and 74th amendments should be implemented in their entirety in letter and spirit and within a finite time. The pace of decentralization can be varied across contexts and departments based on practical considerations that trade off with the government’s policy priorities.

Addressing policymaking challenges—I  207 Adequate safeguards, without defeating the purpose of the decentralization, can be considered during the transition period to mitigate egregious failures. Much has been written about this.4

10.2  Improving Integration and the Flow of Knowledge From Outside Government The third broad reform would be to create structures that ensure the availability of policymakers of non-​governmental inputs and subject matter expertise. To this end, each Ministry or Department should have a ‘Policy Advisory Group’. This would consist of: 1. Selected top civil servants, covering related sectors. To ensure that the groups do not become one more bureaucratic mechanism without clout, only Secretary-​level officers should be on these groups. 2. Stakeholder/​Industry representatives. 3. Academics with expertise in the field. These Policy Advisory Groups should cut across departmental viewpoints and offer integrated policy suggestions. Consultation of the Policy Advisory Group and a consideration of the Group’s views would be required on major policy shifts.

10.3  New Public Management Approaches No discussion of state capability improvement and administrative reforms is complete without examining the role of NPM approaches. The ‘NPM’ movement was a set of policies for civil service and governance reform adopted initially by several developed Commonwealth countries (Australia, New Zealand, Canada, and the United Kingdom) and later by other countries. Some of its key features were: 4 T.R. Raghunandan, Decentralisation and Local Government, Orient Blackswan Private Ltd., 2012.

208  Possible Solutions 1. Adoption of private sector managerial techniques; 2. Introducing competition in the provision of public services; 3. Separation of policy and execution, with the latter under specialized autonomous agencies; 4. Separation of purchasers of services from providers of services; 5. Use of performance contracts or memoranda of understanding (MoUs) between Ministries and the autonomous agencies; 6. Performance-​linked pay and promotion with bonuses and other such measures for achieving targets set in the contracts/​MoUs. Several of these (performance-​linked pay especially) are popular with economists and private-​sector commentators in India. The logic of most such commentators usually runs approximately as follows: 1. This is how the private sector operates; 2. The private sector is more efficient and competent than the government; 3. Therefore, logically, those techniques, if applied in government, must necessarily improve efficiency. The factual basis for (1) and (2) is questionable, but even if one assumes them to be true, both economics and public administration theory indicate that these methods are unlikely to be successful in core public-​ sector activities (as opposed to commercial activities which happen to be run by the government). From the perspective of economic theory, the economist Oliver Williamson, not exactly a proponent of government intervention in the economy, stated in the context of core sovereign functions: ‘Replication of a public bureau by a private firm with or without regulation is impossible’. He goes on to state that ‘practices that are widely condemned (low-​powered incentives, convoluted bureaucratic procedures, excesses of employment security) actually serve legitimate economizing purposes in this context.’5 Lant Pritchett, another economist, not known for his love of government intervention, stated that ‘the provision of key, discretionary, transaction-​intensive services through the public 5 O.E. Williamson, ‘Public and Private Bureaucracies: A Transaction Cost Economics Perspective’, Journal of Law, Economics and Organization 15, no. 1 (March 1999): 306–​342.

Addressing policymaking challenges—I  209 sector is the mother of all institutional and organizational design problems’ (i.e. there are no simple solutions and certainly not simple market-​ based ones).6 Empirical evidence shows decisively that NPM is not a panacea and indeed can make things worse.7 The ability of purchasers and providers to be separated, for purchasers to articulate their preferences in a competitive environment, and the efficacy of decoupling policy from delivery are all questionable, as O’Flynn documents.8 In respect of introducing competition, she points out that ‘evidence shows that [competitive regimes] are usually costly to implement and rarely deliver genuine competition . . . such approaches have resulted in increased transaction costs due to the high costs of contract preparation, monitoring and enforcement’. The Organisation for Economic Cooperation and Development’s (OECD’s) own review of NPM stated that the ‘reforms produced some unexpected negative results’. The OECD review felt the competitive model ‘failed to understand that public management arrangements not only deliver public services but also enshrine deeper governance values.’9 The early adopters of NPM are now rolling it back. In developing countries, NPM is even more risky because NPM only succeeds if what Manning calls ‘OPD’ (Old Public Discipline) is present.10 Heywood points to diminished integrity even in a civil service renowned for probity and political neutrality (United Kingdom) through the excessive adoption of NPM.11 Writing from a continental European perspective, Drechsler identifies the serious flaws inherent in the NPM ideology.12 If

6 L. Pritchett and M. Woolcock, ‘Solutions When the Solution Is the Problem: Arraying the Disarray in Development’, World Development 32, no. 2 (February 2004): 191–​212. 7 N. Manning, ‘The Legacy of the New Public Management in Developing Countries’, International Review of Administrative Sciences 67, no. 2 (June 2001): 297–​312. 8 J. O’Flynn, ‘From New Public Management to Public Value: Paradigmatic Change and Managerial Implications’, The Australian Journal of Public Administration 66, no. 3 (September 2007): 353–​366. 9 Organisation for Economic Cooperation and Development (OECD), Public Sector Modernisation, Policy Brief, Paris: OECD, 2003. 10 Manning, ‘New Public Management’. 11 P.M. Heywood, ‘Integrity Management and the Public Service Ethos in the UK: Patchwork Quilt or Threadbare Blanket?’. Paper presented at Conference on Integrity Management and Collaborative Governance, ICAC Centre of Anti-​ Corruption Studies, Hong Kong, September 2010. 12 W. Drechsler, ‘The Rise and Demise of the New Public Management’, Post-​Autistic Economics Review 33, no. 2 (September 2005): 17–​28.

210  Possible Solutions a country suffers from a corrupt and politicized civil service, NPM does not stand much of a chance. Overall, while some NPM practices may indeed play a positive role if applied in the right context, widespread application of NPM will not be the solution to the problems of the Indian Civil Service.

10.4  Performance-​Based Incentives in Government Arguably no issue gets as much attention in administrative reform as performance management. Conventional wisdom would have it that public systems, in contrast to private organizations, do not focus on performance and align incentives accordingly. Performance-​based incentives, it is often argued, are the most effective strategy to improve outcomes in any organization. Corporate sector remuneration accordingly has two parts: a fixed salary and a variable performance-​linked bonus. Can the same strategy succeed in public sector bureaucracies like education, health care, and other regular government departments? On the issue of linking pay to performance, the evidence is very unclear, even in developed countries. From an OECD perspective, Bourgon found no evidence that performance-​related pay resulted in improvement. He found that ‘performance pay may promote behaviours that emphasize success in the short term at the expense of achieving long term results’ and ‘does not allow enough recognition of the collective and collaborative efforts’. He adds that ‘performance rewards would be susceptible to political influence and make public servants too responsive’ and points to the Australian move to abolish performance pay for Secretaries.13 If the case is weak in developed countries with stronger institutions, the case for performance-​based pay in the civil service of developing countries appears to be non-​existent.14 Besides, the incentive effects of performance-​based 13 J. Bourgon, ‘The Public Service of 2025—​ Themes, Challenges and Trends: Human Resources Management Trends in OECD Countries’, 23 August 2008, Quebec City, (https://​ silo.tips/​downl​oad/​human-​resour​ces-​man​agem​ent-​tre​nds-​in-​oecd-​countr​ies) (Accessed 12 November 2021). 14 Z. Hasnain, N. Manning, and J.H. Pierskalla, ‘Performance-​Related Pay in the Public Sector: A Review of Theory and Evidence’, Policy Research Working Paper No. 6043, Washington, DC: World Bank, 2012.

Addressing policymaking challenges—I  211 remuneration are too complex and vary sharply across contexts, thereby making credible performance payment systems difficult to design.15 (The Government of India has a system of performance-​related pay for public sector enterprises, but its results have not been particularly good.) The starting point for meaningful performance-​based assessments is figuring out the right things to measure. The Office of Management and Budget (OMB) in the Office of the President of the United States defines the characteristics of good performance goals16 • Quality over quantity. Performance goals should be relevant to the core mission of the program and to the result the program is intended to achieve. This generally argues for quality over quantity, with a focus on a few good measures. However, programs should not feel compelled to collapse complex activities to a single measure, particularly if that measure is a proxy for the true objective. • Importance to budget decisions. Performance goals included in the Performance Assessment Rating Tool (PART) should provide information that helps make budget decisions. Agencies can maintain additional performance goals to improve the management of the program, but they do not need to be included in the PART. • Public clarity. Performance goals should be understandable to the users of what is being measured. Publicize (internally and externally) what you are measuring. This also helps program partners understand what is expected from the program. • Feasibility. Performance goals should be feasible but not the path of least resistance. Choose performance goals based on the relevancy of the outcomes and not for other reasons—​not because you have good data on a less relevant measure, for example. If necessary, terminate less useful data collections to help fund more useful ones. • Collaboration. Agencies and their partners (e.g. states, contractors) need to work together and not worry about “turf ”—​the outcome is what is important. 15 Axel Engellandt and Regina T Riphahn, ‘Incentive Effects of Bonus Payments’, IZA Discussion Paper No 1229, July 2004 (https://​pap​ers.ssrn.com/​sol3/​pap​ers.cfm?abst​ract​_​id=​ 571​725). 16 OMB, ‘Performance Measurement and Challenges and Strategies’, 18 June 2003 (https://​ geor​gewb​ush-​whi​teho​use.archi​ves.gov/​omb/​perf​orma​nce/​challe​nges​_​str​ateg​ies.html).

212  Possible Solutions The authors are sceptical of its success when implemented broadly in a country like India. Using some of the considerations mentioned in the OMB framework, here are some reasons why performance-​based pay can run into problems in real-​world implementation in the government.

10.4.1  Quantification and Measurement Problems The fundamental prerequisite for any credible performance-​based pay system is its ability to quantify and measure outcomes. To start with, there is a strong case that not all dimensions of performance in social sectors can be quantified. Complicating any measure of quantifiable outcome is the role of widely varying exogenous factors like the social and economic background of student/​patient/​beneficiary, family environment, and historical and legacy factors, all of which exert considerable influence on the final outcome. Furthermore, it is important to have some level of broad consensus about the variables used to measure performance outcomes and the actual measurement process itself. Such consensus is most often not forthcoming. For example, how do we reliably measure learning outcomes among primary school students? Even assuming some level of consensus among all stakeholders about what should be measured, what are the instruments available for its measurement? How do we ensure that instruments like examinations are not subverted? These are questions for which there may not be answers which have some reasonable degree of consensus. In general, measuring the quality of service delivery is hard. It is harder still when the service delivery involves direct engagement with individual citizens. Once there is broad acceptability of the performance measure and the process of capturing it, its credibility also depends on the rigour of monitoring. Too much rigour in the measurement and data collection process means high administrative costs; too little rigour, and the purpose is defeated. It is natural that any such measurement system, especially if deployed for not-​so-​low stakes decisions, is bound to invite attempts at subversion. Even assuming the availability of a credible-​enough performance measurement framework, like in the case of maternal and child health

Addressing policymaking challenges—I  213 interventions, the challenge of reliable data collection remains. How do we ensure that the Auxiliary Nurse Midwife (ANM) is not reporting inaccurate figures? Super-​checks and sample validations by third-​party agencies, while useful, may not be credible enough when done on a state-​wide or national scale. Besides, they take up their bit of administrative effort and cost. An incentive is effective only when its magnitude is beyond a particular threshold. Too low an incentive fails to evoke the desired or even any performance response. This is more likely given the already high salary levels among lower and middle government employees. However, too high an incentive, apart from being not cost-​effective, also distorts incentives by raising stakes and encouraging false reporting and data manipulation. The optimal incentive level varies both across sectors and with each sector from place to place. For example, what works in education may not be the same in health care. In education itself, performance incentives may differ from primary to secondary school, from one geographical area to another, and from rural to urban. Given all these factors, calculating the incentive with any reasonable level of accuracy becomes a very complicated task. Further, if there are too many incentive slabs, then that creates another set of dynamics. Finally, the practical administration of such incentives is difficult.

10.4.2  The Administrative Uncertainty Principle There is a kind of Heisenberg’s uncertainty principle in public administration. (Heisenberg’s uncertainty principle states that that the position and the velocity of a sub-​atomic particle cannot both be measured simultaneously. If one is determined, the other is indeterminate and vice versa.) In the authors’ experience, data gathered in public administration can either be accurate or can be useful for decisions—​but not both at the same time. Take the case of a census. A census that merely tries to anonymously count the population and its characteristics almost always elicits truthful responses from citizens. The citizen responding to questions about his caste, income, etc., has no incentive to lie. But such anonymous data is of no use in the targeting of subsidies because while the census might

214  Possible Solutions show that there are (say) five poor Scheduled Caste women in a village, it does not identify them by name. So the numbers are accurate but not administratively useful. On the other hand, if the respondent’s identities were recorded, the data would be very useful in the administration of social welfare programmes. But if the respondents to the census knew that it would be so used, they would be tempted to falsify responses so as to obtain higher benefits—​e.g. people who are not Scheduled Castes may claim to be so. Now the data would be usable—​but inaccurate. This phenomenon is found repeatedly in administrative data-​gathering and is a major obstacle to measuring outcomes and performance. (Indeed, for this reason, future rounds of Socio-​economic and Caste Census data are likely to be less reliable than the first round when the use to which data would be put was not known.)

10.4.3  The Moral Issues With Public Service Delivery Besides quantification and measurement, a deeper problem with performance management involves the moral choices faced while implementing programs. Performance management measures are inherently skewed towards capturing outputs and outcomes, reflecting the efficiency of service delivery. But public services are also about fairness and distributive justice to enable access to certain services across widely varying environments. Besides, public servants, especially those dealing with individual citizens demanding statutory services, are also supposed to be compassionate, responsive, and respectful. It is very difficult to measure any of these in a satisfactory enough manner, certainly at scale. A focus on efficiency can, therefore, force trade-​offs that lead to the crowding-​out of these moral values. Reflecting on this, Bernardo Zacka, who wrote a book chronicling the lives of street-​level bureaucrats, writes17:

17 Bernardo Zacka, ‘Bureaucrats to the Rescue’, Boston Review, 21 September 2017 (http://​ bosto​nrev​iew.net/​polit​ics/​berna​rdo-​zacka-​bure​aucr​ats-​res​cue).

Addressing policymaking challenges—I  215 Focusing on objectives rather than procedures was meant to liberate bureaucrats from the tangle of red tape, boosting their efficiency while providing the public with a more objective standard for accountability. Both aspirations are of course laudable, provided we do not fetishize them. While greater accountability is welcome, setting clear metrics as goals risks displacing or distorting an agency’s overall mission. In such cases, we achieve a pyrrhic victory: securing greater control over the actions of bureaucrats at the cost of impoverishing their mandate. Lest we succumb to such facile temptations, the directives we provide bureaucrats must remain underdetermined. We cannot content ourselves with assessing performance in terms of efficiency or compliance to standards. Bureaucrats, after all, do not merely execute. They must also grapple independently with complex normative questions. To govern them as if they were mere instruments is the surest way to hollow out their moral promise. If bureaucracy is to stand a chance of being a public good, we must evaluate and manage bureaucrats with that in mind.

This has resonance across most development sectors. How do we capture the personal interest shown by certain public-​school teachers that encourages more parents to send their children to a school? How do we evaluate the empathy and connect that certain frontline officials establish with their stakeholders? How do we measure the regard that patients develop for certain doctors? How do we measure the intense commitment and sincerity displayed by certain officials during the Covid-​19 pandemic response? In other words, performance management approaches struggle when faced with assessing the work of bureaucrats. Then there are the operational challenges with their implementation, which we now turn to.

10.4.4  Institutional Culture, Incentive Distortions, and Unintended Effects Any performance-​based payment initiative has to be institutionalized both administratively and culturally. In fact, its widespread cultural or social acceptance as an incentive to reward good performance underpins the success of the administrative implementation. This plays a critical role in creating the stakeholder pressures and institutional vigilance

216  Possible Solutions that are vital to ensuring that forces to subvert the system from within are foiled. But cultural internalization is difficult to achieve when implemented in countries with wide social and cultural diversity. In the absence of socialization about its benefits among the stakeholders, willy-​nilly subversive tendencies creep in. Once a significant share of the employees have partaken of bonuses at some time or the other, it is only a matter of time before they come to view these bonuses as entitlements and unions and vested interests enter to defend the new status quo. The incentives turn into an entitlement. For example, there is a system of ‘bonus’ for railway employees, originally dependent on the financial performance of the railways. But over the years, it has become a fixed sum equal to precisely 78 days of wages, regardless of actual performance. Public servants, especially field-​level functionaries (teachers, medical supervisors, bill collectors, tax inspectors, building inspectors, and so on), are among the highest-​paid workers (among comparable levels) in many developing countries.18 This also partially explains the scramble for these jobs. Yet, their actual performance is generally significantly lower than what their formal responsibilities demand. So it is commonplace to have tax collectors who assess and realize just 30–​50% of the estimated demand, electrical engineers managing substations with distribution losses in the range of 30–​40%, or teachers who attend school only a third of the working days, and whose student learning outcomes are abysmal, or primary care hospital doctors achieving just 30% universal immunization and 40% institutional deliveries. This is business as usual in many low-​income countries. Thus, one has the (relatively) high-​paid officials achieving such low outcomes. Performance incentives in such circumstances face two challenges. The first is incentive distortion. The issue shifts from improving outcomes using existing levers to one getting the most optimal design of incentives. Implicit in the latter is an acknowledgement that the poor status-​quo is inevitable and ‘alright’ and we are now therefore looking at the performance incentive lever to improve outcomes. The moral hazard 18 T.H. Gindling et al., ‘Are Public Sector Workers in Developing Countries Overpaid?’, World Bank Policy Research Working Paper 8754, February 2019(http://​docume​nts.worldb​ank.org/​ cura​ted/​en/​398​3615​5111​7216​050/​pdf/​WPS8​754.pdf).

Addressing policymaking challenges—I  217 from this can create unexpected outcomes. Poor performance soon gets to becoming psychologically rationalized by truant officials. Further, not only the base salary but the structure of incentives too become factors in wage negotiations. Second, the implementation of financial performance incentives can have unexpected effects. The most likely impact is that over time, performance incentives come to be viewed as an entitlement. Or for example, over time, it is reasonable to have expectations and attitudes that re-​balance baseline performance downwards and revision upwards of incentives. When an incentive gets scaled up, employees tend to forget the baseline salary (which is then taken for granted) and there is the danger of the incentive becoming seen almost as an entitlement—​if you want us to improve the performance, give us more incentives! The following is a real example: During the budget season in the Finance Department of a state, staff staying beyond normal working hours were paid overtime provided they stayed an extra 3 hours. With the efflux of time, staff got used to getting overtime throughout the season and many stopped staying the full extra 3 hours. In the early 1990s, a new Deputy Secretary started checking the presence of staff at the end of the 3 hours and denied overtime to those not present. Far from feeling guilty, a lot of staff felt aggrieved at having to stay on to get the overtime that they used to get without staying! The incentive became an entitlement and the enforcement of the intended linkage to performance became a problem. (The Deputy Secretary stuck to his guns and eventually the staff complied.)

There are several examples of distortions along these lines which have been engendered by financial performance incentives. The practice of bonuses is commonplace in public sector power (generation, transmission, and distribution) companies in India and their impact in terms of actually lowering transmission and distribution losses is perhaps nil. A ‘productivity-​linked bonus’ is paid to railway staff but in practice is not linked to productivity; it has been kept at 78 days’ wages for many years irrespective of railways’ actual performance.

218  Possible Solutions

10.4.5  Maintaining Credibility The success of performance management initiatives is, to a large measure, dependent on its acceptability among the large portion of its stakeholders. This credibility rests on tenuous foundations. It is easily shaken by a few jolts, which most often ends up giving a convenient excuse for opponents to question its reliability. Even where outcomes can be measured, there are problems with comparability and standardization of performance assessments. For example, how do we compare the realization of maternal and child health outputs of a medical officer or Collector in a remote versus an urban district? The variations in conditions even across districts in a state, much less across states themselves, are so large as to defy any practical comparative assessment system that can be applied to something as high-​stakes as a personnel performance assessment. Such inconsistencies in comparability are difficult to avoid and can quickly raise questions about fairness, which is central to the credibility of the process. Another factor complicating matters is the collective action nature of outputs in many government activities. A teacher requires the support of the headmaster as much as the field nurse requires that of the medical office of the primary health centre. A district agriculture officer is a manager of block agriculture officers. If many of those posts are vacant or filled by in-​charge officials, then the fairness of assessing the district officer becomes questionable. Once a few lapses get highlighted, especially high-​profile ones, a downward spiral is never far away. Loss of credibility inevitably follows. It suits the vested interests to publicize such shortcomings to add credence to their opposition. The political and related challenges associated with performance measurement systems are formidable.19 Any government service delivery channel, especially in democracies, is embedded in a political system. It is therefore natural that the processes and administering stakeholders are exposed to political dynamics. The most salient manifestation of this is

19 Jenny M. Lewis, ‘The Politics and Consequences of Performance Measurement’, Policy and Society, 34 (2015) (https://​www.tand​fonl​ine.com/​doi/​full/​10.1016/​j.pol​soc.2015.03.001).

Addressing policymaking challenges—I  219 the employee unions. In these circumstances, there are many areas where political considerations can take precedence in the conceptualization, implementation, and sustainability of performance-​based pay initiatives. Once such considerations creep in, it dilutes the program’s objectivity and raises hackles among rival political and social groups, thereby denting the credibility of the process.

10.4.6  Scale Dynamics Much, if not all, of the evidence of success with performance-​based pay in social sectors comes from small-​scale experiments. Such studies fail to account for the dynamics that emerge once a performance-​based pay program is expanded. The administrative challenges are the most formidable. For example, how does one address the problem of measuring the performance of thousands of teachers and administrators scattered across each taluk and then comparing them across dozens of taluks?

10.4.7 Conclusion In conclusion, we are inclined to believe that any performance-​based pay system for government officials, while unobjectionable at a theoretical level, may be very difficult to implement, most certainly for political and administrative reasons, in the prevailing environment in countries like India. While it may succeed in a limited area, scope and time, it may not yield the desired results with a more ambitious scope and pan-​Indian area of implementation.

10.5  Other Reform Ideas This closing section examines some of the other proposals commonly suggested as part of administrative reforms. The ones with question marks are ones the authors do not regard as desirable or effective.

220  Possible Solutions

10.5.1  Restrict Civil Service Entry to Those Who Have Completed a Course in Public Administration? The 2nd Administrative Reforms Commission (ARC) recommended that the Government of India establish new National Institutes of Public Administration to run bachelor’s degree courses (broadly on the pattern of the successful National Law Schools). Only graduates of these institutions or those who have completed a bridge course conducted by them would be eligible to write the civil services examination. It is not clear which problem currently faced by the civil service this ARC recommendation is intended to tackle. In one sense, this approach is similar to the French Ecole Nationale d’Administration (ENA) system, where young graduates enter the ENA and then become civil servants. A benefit of this approach would be that candidates would have spent their college years preparing themselves for a public service career and the successful among them might be better skilled in the early years of their career. However, this approach would have the serious drawback of removing from the pool many candidates who have an interest in the civil service but feel they should have an alternative career if they are unsuccessful; very promising candidates who are unwilling to spend time on the bridge course which has no value to them would not attempt the examination. It would also produce a ‘monoculture’ within the civil service and take away one of its big strengths—​the broad and multi-​disciplinary character. Also, graduates of the new public administration institutes who fail to pass the civil service examination would reach a dead end; since their qualification would be largely worthless in the private sector, they may eventually ‘demand’, through political or judicial means, absorption in government. On the whole, this proposal is not a good one.20

20 S.K. Das had proposed recruitment at the end of higher secondary school as an alternative to examinations for entry to medical colleges, IITs, and the likes. He argued that this would widen the pool by bringing in bright students who might otherwise enter the private sector and not take the examination later, remove issues of age relaxation, and increase the degree of idealism and dynamism among the recruits. This would be a fundamental change whose consideration is beyond the scope of this paper. Das, Building a World Class Civil Service for Twenty First Century India, Oxford University Press, 27 September 2010, 172–​174.

Addressing policymaking challenges—I  221

10.5.2  Creation of Civil Services Board/​Enacting a Civil Services Act? The 2nd ARC and others have recommended the creation of a Civil Services Board (CSB) under a Civil Services Act, which would be in charge of personnel management at central and state levels. It is unlikely that this will be an effective solution; there is already a statutory framework and the record of such boards and committees (especially at the state level) does not create confidence that they would function without political interference.

10.5.3  Curb Post-​Retirement Opportunities by Aligning Retirement Ages A simple but highly effective solution to the problem of top civil servants seeking post-​retirement opportunities is to amend all the statutes relating to tribunals, regulatory bodies, and training institutions to bring their retirement age to exactly the same level as the civil service retirement age. Thus, these opportunities should be taken up before, not after, retirement. This should also apply to judicial members whose retirement age should be identical to the retirement age of High Court/​Supreme Court judges (as the case may be).

10.5.4  Training in Specific Skills Using Modern Online Learning Technology The primary focus of trainings continues to remain the physical residential courses at, for instance, Duke University or Harvard’s Kennedy School of Government—​which are currently part of in-​service training programmes. In addition to them, and perhaps more importantly, the DoPT and the National Academy of Administration should, in collaboration with leading Western universities, commission, develop, and use very carefully designed and updated online courses. These would be of three types.

222  Possible Solutions The first would be targeted and subject-​specific modules to provide each Joint Secretary/​Director with an up-​to-​date curriculum giving the latest trends (in the sector the officer is assigned to) in India and abroad, with an overview of the laws and regulations applicable, the key pending policy issues, and so forth. Such courses should be made compulsory, with officers posted in a particular ministry or sector (at the centre or even in the states) required to complete the online course with a pass mark within, say, two months of joining—​failing which they will not be qualified for their next increment. About 200–​400 such modules may need to be developed, with at least one for each Department and not more than one per division. Courses could be updated systematically once a year. This represents a quick and feasible way of improving competence and should neither face political resistance nor require additional funding. The second area where targeted training can improve capacity across a variety of sectors is training in procurement and financial procedures. Many officers (from generalist or specialist services) have a poor understanding of the rules and principles of good public procurement and of the financial code. Quite apart from deliberate malfeasance, many officers make mistakes purely out of ignorance. Furthermore, their ignorance creates undue fear of these matters, which leads to slower decisions, the appointment of unnecessary committees, the prompting of repeated queries, and so on—​all of which slow down administration and reduce quality. The UK Civil Service has recently started a capacity-​building effort aimed at procurement and certain other targeted common skills.21 A set of modules on procurement and financial rules should be built as an online training tool using the best modern content and delivery techniques, updated regularly, and tailored to Indian rules and regulations. A third area would focus on good decision-​taking processes. The traditional civil service rules have strong procedural safeguards against bad decisions, but these are mostly through wide consultation. They do not adequately address the typical cognitive mistakes made in decision-​ taking. A targeted training module for senior officers at the level of Joint Secretary and above should be developed on techniques like those advocated by Nobel Laureate Daniel Kahnemann in his recent practical note 21 H.M. Government, Meeting the Challenge of Change: A Capabilities Plan for the Civil Service, UK Government London, UK: April 2013.

Addressing policymaking challenges—I  223 on ‘quality control for decisions’ (which is content-​neutral and applicable to all sectors).22 Overall, a re-​deployment of training funds to develop tailored and very specific online training content from the best providers is likely to produce good results. In 2020, the Government of India had launched Mission Karmayogi and created the Capacity Building Commission, which seeks to do exactly this.

10.6  Personnel Management 10.6.1  Career Progression—​the ‘Informed Generalist’ The need for an ‘informed generalist’ was discussed in Chapter 7. There may be a strong case now for promoting some degree of specialization for the IAS.23 The IAS was modelled after the colonial era Indian Civil Service as a generalist service to deliver the core functions of the state—​collect taxes and maintain law and order. The challenge of development in a large, populous, and impoverished country was probably not on the radar screen when the IAS was designed. But it soon became apparent that this development task would become central to public administration, especially at the state level. The IAS adapted to these changing dynamics by re-​ tooling itself as a ‘development agent’ and, on the whole, acquitted itself quite creditably. As economic reforms deepened and the state started yielding to the market, the nature of administration changed, demanding domain knowledge, especially at the policy level. This raised questions about the role and relevance of a generalist civil service like the IAS. Two views emerged. The first is the argument that the best leadership is provided by generalists who have a breadth of understanding and experience. Specialists, 22 D. Kahneman, D. Lovallo, and O. Sibony, ‘The Big Idea: Before You Make That Big Decision . . .’, Harvard Business Review 89, no. 6 (2011): 51–​60. 23 The following discussion draws from the op-​ed by D Subbarao and Gulzar Natarajan, ‘Generalist vs Specialist’, Indian Express, 25 September 2017 (https://​indian​expr​ess.com/​arti​cle/​ opin​ion/​colu​mns/​upsc-​gen​eral​ist-​vs-​spe​cial​ist-​ias-​recr​uitm​ent-​4859​491/​).

224  Possible Solutions no matter how competent, tend to have tunnel vision and are often not equipped to take a broader view. Domain knowledge has to feed into policymaking, but that can be accomplished by domain experts advising the generalist leader in decision making. In this world view, a good IAS officer can head the Department of Agriculture as competently as she would the Department of Shipping. The opposing view is that the IAS, as generalists, tend to overweigh their experience of the process and form an over-​understanding of policy content. Only someone who has learnt the subject from the trenches, as it were, can provide competent leadership in a functional area. Having the IAS head specialized areas is an inefficient arrangement.

Addressing policymaking challenges—I  225 This tension between breadth and depth is a debate that has frowned on moderation. But there is no need to look for binary solutions. The complex and interconnected nature of policymaking demands that specialist expertise has to go with generalist experience. Notably, the Constitution Review Commission 2002 suggested the ‘need to specialize some of the generalists and generalize some of the specialists’. That seems to be a wise dictum for the way forward. That raises the challenge of managing specialization. When does an IAS officer start to specialize? How will the system be operationalized? The example of the private sector is instructive. There, young professionals are typically recruited into specialized areas and they rise to generalist leadership positions negotiating their way up the hierarchy. What we have, or should have, in the public sector is in fact its reverse. Young recruits should join the IAS as generalists, acquire breadth, and then go on to acquiring depth. The first decade of an IAS career is typically spent in field postings with responsibility for policy execution, which hones their administrative and people management skills, apart from giving them an invaluable understanding of the ground realities. From there they graduate into policy formulation positions both at the centre and state levels. This transition provides the ideal marker for starting to specialize—​combining the soft skills they have learnt with the hard skills of a specialized domain. Managing specialization can be a complex challenge. How many specializations should there be? How should officers be allocated among the specializations? What should be the respective weights for expressed preferences and revealed competencies? Once allocated to a specialization, how should an officer’s career be managed? A starting point can be to categorize ministries broadly into three groups—​welfare ministries, regulatory ministries, and economic ministries since experience suggests that each of the three categories demands broadly similar behavioural attributes and aptitudes. A couple of principles should inform the process. First, allocating officers across specializations cannot, and should not, be reduced to a formula. It is best to work the system flexibly, allowing specialization to emerge gradually through a process of deliberate iteration at the mid-​career level. This will facilitate officers specializing as they move up the hierarchy based on revealed aptitude and record of performance.

226  Possible Solutions Needless to say, because the system needs to be flexible, it places the onus on the government to make it predictable and transparent. Second, specialization need not be mandatory. Some IAS officers may prefer to remain generalists. Indeed, the system too is always in need of some generalists. One of the tasks of cadre management will indeed be to match the supply and demand across specializations and generalists. Once allocated to specializations, officers should be afforded opportunities to deepen their domain knowledge through study and training. Also, since IAS officers are recruited at a young age, they hardly ever experience the government from the outside. They should therefore be allowed, even encouraged, to work outside the government, preferably in a non-​governmental organization for a few years irrespective of their area of specialization. This is bound to make them more useful and relevant civil servants. This effort to optimize generalist experience with specialist domain expertise should apply to lateral entrants as well. Just as regular recruits are required to specialize, lateral entrants should be required to ‘generalize’ through field postings so that they have an opportunity to dirty their hands. Giving the IAS an optimal blend of breadth and depth is a complex challenge. The way forward lies in eschewing binary solutions and embracing a more nuanced, iterative process of active but careful cadre management. No doubt, this is not easy to do—​but the effort is worthwhile.

10.6.2  Reforming Appointments/​Postings It is settled wisdom that senior-​level appointments are a process of match-​making. Employers select those they find most appropriate from among interested eligible applicants. The current process of postings under the Central Staffing Scheme (CSS) can be improved to meet these objectives. Postings at state and central government levels are generally done without the formal involvement of the officer posted. Officers of different services, from the levels of Deputy Secretary to Secretary to Government of India, apply to the Department of Personnel and Training (DoPT) for deputation to the various CSS and non-​CSS

Addressing policymaking challenges—I  227 posts under the Government of India. They are scrutinized for eligibility by the DoPT and their names are placed on an offer list. The CSB, headed by the Cabinet Secretary, then meets and recommends them to various posts on certain administrative and other considerations. The CSB’s recommendation is then placed before the Appointments Committee of the Cabinet (ACC), headed by the Prime Minister of India, which approves the appointments. While the officers can express their preferences in their applications, this is not a primary consideration in the selection process. Further, while the Secretary concerned is a member of the CSB, there is no other Departmental consultation in terms of the choice of officers. Admittedly, the process allows for a degree of flexibility to accommodate the views of the officer and Department concerned. It is for the government of the day to set the informal boundaries of this accommodation. In the past, there have been times when the informal practice was for the Ministry to suggest the name which the CSB used to recommend and ACC approve. However, this created its set of problems in terms of allegations of favouritism and even corruption. In the existing process, the views of the principal, the Government of India represented by the ACC, prevail over that of the employing agency, the Department concerned, and the preference of the employee. The current process could be modified to allow for internal advertisement and applications among those within the zone of consideration without affecting government’s unfettered right to appoint any officer to any post. It would help minimize unintentional mismatches and encourage a greater proportion of win-​win outcomes, better matching of competence with need, and increased motivation. The challenge is to design a process for posting officers which lays down certain objective considerations, promotes specialization among officials, and allows for reasonable accommodation of the preferences of officials and Ministries, without compromising on the government’s need for reasonable control over the process and postings. It is with this in mind that the following is a suggested process for CSS postings: a. The majority (see item ‘e’ below) of CSS postings at Director, Joint Secretary, and Additional Secretary levels should be done through

228  Possible Solutions





a process where the Ministry/​Department concerned notifies the vacancy, evaluates and interviews applicants based on DoPT approved parameters and recruitment rules, and recommends a panel of two (or three) names for ACC approval. b. This would dispense with the offer list mechanism and those not considered will have to apply elsewhere. c. In order to avoid frivolous applications, there can be limits on the number of times an officer can apply to one Ministry. d. Each deputation should be for a tenure of five years. In case the officer gets empanelled during the course of the deputation, he/​she must apply again within three months and be posted as a Joint Secretary or revert back to the cadre. e. Some posts, considered sensitive, could be kept outside the scope of this process.

This process could address many of the concerns with the prevailing and previous processes while meeting the reasonable requirements of the officer, Department, and the Government of India. It retains the overall control exercised by the government through the ACC while also accommodating the interests of the officer and the Department. In terms of the importance of these national-​level postings, especially at Joint Secretary and above, it is only appropriate that the selection be done through a rigorous enough process of choosing from among a set of professionally competent and interested officials. This will eliminate the possibility of officers being posted against their choice and therefore make them more accountable for their performance. Further, it is not as though this is a completely new process. Already a similar process is followed for the non-​CSS postings of the Government of India. The post is notified, officers apply, the Ministry recommends a panel of two or three officers selected through approved Recruitment Rules, and ACC approves the same. This would be extending the process to the CSS postings too. The main criticism of this approach would be that it can result in a skewed distribution of applications, with interest for certain Ministries (Commerce, Finance, Urban, Infrastructure, etc.) at the cost of others (e.g. Defence, Personnel, Home, Rural, Sanitation, Education, Agriculture, Tribal, Minority Affairs, etc.). This is not as serious as it seems.

Addressing policymaking challenges—I  229 Firstly, it is an exaggeration to say that officers prefer only the former category of Ministries. There is no dearth of officers with a passion for social sectors, including Agriculture or Sanitation or Rural Development. Secondly, since not everyone can go to the first category of Ministries, it is inevitable in any case that officers wanting deputation to the Government of India will apply for other Ministries. Assuming that a significant share of officers would like to work in the central government, it is possible with appropriate conditions (like on the number of posts that an officer can apply during a period) to achieve a more balanced distribution of applications among departments. Finally, positions can still be filled through the old system, which would remain available wherever the application system does not work. Another important reform is that postings should be for a reasonable minimum tenure, as suggested in the 2nd ARC. The issue of minimum tenure for AIS officers in a post received the attention of the Supreme Court in its judgement in T.S.R. Subramanian v. Union of India.24 A similar verdict in respect of the IPS has had only limited effect.

10.6.3  Promotions into All India Services In keeping with the principles of meritocracy, neutrality, and an all-​India character, selection procedures need to be changed to ensure that those promoted into the service are the genuinely meritorious among the state civil servants rather than those who are politically favoured.

10.6.4  Reforming Empanelment The guidelines for evaluating the suitability of officers for inclusion in the batch-​wise panels for empanelment as Joint Secretary,25 Additional Secretary, and Secretary26 state that: 24 AIR 2014, SC 263. 25 ‘Consolidated Empanelment Guidelines for Joint Secretary’, Department of Personnel and Training, Government of India, 24 July 2014 (http://​docume​nts.doptci​rcul​ars.nic.in/​D2/​D02​ eod/​empane​lmen​tgui​deli​nes.pdf). 26 ‘Guidelines for Evaluating Suitability of IAS Officers’, Department of Personnel and Training, Government of India, 1 January 2008 (http://​docume​nts.doptci​rcul​ars.nic.in/​D2/​D02​ eod/​Guidelin​esOn​Empa​nelm​entO​fIAS​Offi​cer.pdf).

230  Possible Solutions Empanelment should be considered not as a reflection of the intrinsic merit or otherwise of an officer but the suitability of an officer to occupy senior levels in the central government. Given the background and experience of an officer, she or he may be highly suited to occupy senior positions in state governments. Likewise, another officer, in view of the background and experience, may be considered more suitable for central government posts.

In other words, empanelment is not a merit-​based promotion but a suitability screen for senior positions in the central government. It is applicable for All India Service, Indian Foreign Service, and Group A officers from all services who choose to work with the central government. As indicated, it happens at three levels of seniority—​Joint Secretary, Additional Secretary, and Secretary. In terms of eligibility, for each of the three levels, officers should have a minimum number of years of service in Group A as well as be drawing a certain pay scale. In terms of consideration norms, the officer should have a minimum number of Annual Confidential Reports (ACRs) as they used to be called or Performance Appraisal Reports (PARs) as they are now called and be clear from the Vigilance side. A problem with the PARs (as distinct from the earlier ACRs) as a measure of performance is that there is a distinct ‘grade inflation’ inasmuch as most officers are being rated above 8 on a scale of 1 to 10. The grade inflation has occurred after the practice of transparency was introduced along with the new PARs. Under the ACR system, only an adverse rating was communicated to the officer rated. This meant that superiors felt more free to make detailed subjective descriptions, which made it possible to distinguish different levels of performance. They also felt able to give gradings of ‘Good’ which were not adverse but below the grades of ‘Very Good’ and ‘Outstanding’. With the introduction of PARs and transparency, subordinates began to appeal non-​adverse grades, which earlier they would not have been able to contest. Some officers have even contested grades of 9, alleging that they had previously got 9.5 for the same performance. The process of appeal puts a strain on the superior who gives the grade. As a result many now mechanically give high grades to everyone. Consequently, the near-​universal high absolute grades have become of limited use in relative ratings of officers.

Addressing policymaking challenges—I  231

As a result, in addition to the formal performance reports, in recent years a system of Multi-​Stakeholder Feedback has been introduced. An Expert Panel scrutinizes the ACR and makes informal enquiries to grade each officer. The enquiries help with a 360-​degree assessment of the performance of the officer. This grading is then placed before the CSB, whose mandate is described below27:

27 ‘Consolidated Empanelment Guidelines for Joint Secretary’, Department of Personnel and Training, Government of India, 24 July 2014 (http://​docume​nts.doptci​rcul​ars.nic.in/​D2/​D02​ eod/​empane​lmen​tgui​deli​nes.pdf).

232  Possible Solutions The Civil Services Board (CSB) would take into account the experience profile of officers, carefully scrutinize the ACR dossiers and evaluate such qualities as general reputation, merit, competence, leadership and a flair for participating in the policy making process to recommend the list of officers to be included in the panel.

The recommendation of the CSB is placed before the ACC to select the officers for empanelment from each batch. In an ideal world, the process of empanelment should strike a reasonable balance between the often-​conflicting imperatives of fairness to the civil servant and public service efficiency. There are at least two problems with the concept of empanelment. One, it can be argued that it prioritizes fairness to civil servants over public efficiency.28 Batch seniority is sacrosanct in empanelments—​an officer from a junior batch cannot be considered for empanelment, and therefore posting, without having considered all senior batches. This gives the impression of subordinating the idea of identifying the best candidate for senior positions to seniority considerations. Such adherence to seniority for selecting candidates to occupy senior positions is very rare within bureaucracies elsewhere. This stands out at the level of Secretary to the Government of India. The pool of empanelled officers available to choose for postings is very narrow. They become empanelled very late in their careers, thereby limiting the tenures available for them as Secretaries to mostly less than three years. At the state government level, the Chief Minister can choose from a very large pool of officers to post as Secretaries,29 since inter se batch seniority among Secretary level officers is not a consideration for postings. But at the central government, given the fastidiousness on the seniority of batches, the Prime Minister has a limited pool to draw upon for appointments as Secretaries to government.

28 In recent years, the introduction of the ‘Multi-​Stakeholder Feedback’ has reduced the weight for ‘fairness to the individual civil servant’. It is widely believed to have improved the quality of officers empanelled, though some officers perceive it as opaque and therefore unfair. But it still goes batch by batch, and in that sense respects seniority. 29 The Secretary to the state government is an officer who has been promoted to the Senior Administrative Grade, or the level for empanelment as Joint Secretary in the Government of India.

Addressing policymaking challenges—I  233 The second problem arises from the lack of transparency with performance evaluations. Even in the best human resource settings, performance assessments are challenging. But in a system where a sense of fairness about the civil servant is pervasive, it becomes even more complicated. The multiple considerations used for empanelment and the stakes associated with it30 create a very high level of dissatisfaction and perception of opacity about the process itself from every person not empanelled. All officers empanelled think they got it because they deserve it; all those not empanelled feel they did not get it because the process is not fair. There are no solutions that satisfy the requirements of everyone. The solution suggested below conforms to standard principles followed in bureaucracies globally. In all of them, postings to senior levels of the organization elevate merit over seniority considerations, subject to certain eligibility norms; supersession of seniority for postings to senior positions is an accepted norm. Further, within any cohort of recruits, it is only natural for a small share to occupy the highest positions. The proposed system is: • Dispense with the process of empanelment altogether; this will eliminate the need to have a separate process that empanels people for certain levels of postings. In other words, no one is excluded from consideration • Fix a minimum seniority for all internal candidates based on which service they belong to—​e.g. must have reached a particular grade in the parent cadre to be eligible and/​or have served a number of years at that level • All persons in that level of seniority are eligible to apply for posts • For actual selection and posting from among those eligible, use the internal advertising or direct posting systems discussed above along with a process like the 360-​degree multi-​stakeholder feedback.

30 Even though the government explicitly makes clear that empanelments are not a promotion and officers in their state cadres have already received promotions, and many officers never leave the state government to work with the Government of India, a failure to be empanelled creates a stigmatizing perception.

234  Possible Solutions

10.6.5  Lateral Entry Few issues in civil service reform arouse more passion and acrimony than does lateral entry into the IAS. The conventional wisdom on lateral entry is that it infuses fresh energy and thinking into an insular and complacent, often archaic, bureaucracy. It enables the entry of right-​minded professionals and the adoption of best practices into improving governance. However, this belief has to be seen in the light of the country’s socio-​political context as well as the complex nature of the public policy challenges being sought to be addressed. 10.6.5.1 The Case Against Lateral Entry Many arguments in favour of lateral entry are based on the belief that its success in mature Presidential democracies like the United States can be readily replicated here. They under-​estimate the recruitment, functional, and operational difficulties associated with lateral entry. For a start, a generalized system of lateral entry poses formidable recruitment challenges. Given the erosion in state capability and institutional credibility at all levels, it runs the risk of degenerating into an uncontrollable ‘spoils’ system. This, coupled with the inherent problems of revolving-​door personnel management, raises concerns about accountability. If not properly done, far from infusing fresh energy, lateral entry could further enfeeble the bureaucracy. Functionally, the cutting edge of implementation for the policies formulated in the vast majority of such postings is at the sub-​national levels, most often at the levels of local governments. These policies are implemented in a complex and dynamic ecosystem, involving negotiations among multiple interest groups, several bureaucratic and political layers, and numerous resource and state-​capability constraints. In the absence of adequate field experience, lateral entrants entrusted with such policy formulation are most likely to have only a limited appreciation of these challenges. A few sanitized visits to Primary Health Centres or cursory interactions with field nurses do not equip one with the skills to formulate health care policies for a vast and diverse country like India. Then there are the operational challenges associated with lateral entry. Any cherry-​picked infusion of external talent into only high-​profile posts, apart from adversely affecting morale among incumbents, is also

Addressing policymaking challenges—I  235 likely to distort the incentives of entrants. How do we mitigate the incentive distortions that are likely with a revolving door approach, whereby market talent moves back and forth between the government and corporate world? Restrictions to address these distortions are only likely to turn away precisely those best and brightest sought to be attracted. Further, will the best market talent be attracted by a lateral entry process that offers a mix of high-​profile posts and unglamorous, even drudgery-​filled, ones? Would they be willing to rough it out in a complex and uncertain, sometimes frustrating, work environment for a 10–​ 15 year tenure, at a fraction of their market remunerations? If it fails to attract the best and brightest and only attracts the also-​rans from the market, as is likely, then one may end up with the worst of all worlds—​ a demoralized bureaucracy and lateral entrants of less than desired quality. Finally, there is the issue of its impact on the civil service. If the limited numbers of high-​profile and critical posts get earmarked for market talent, who are more likely to effectively bargain in a ‘spoils’ system, it would adversely affect career progression opportunities. The consequent deterrent effect on civil service aspirants, leave aside the morale of the incumbents, can be significant. These are only a few illustrations of the operational challenges associated with lateral entry. Given all this, far from rectifying the current failings of Indian bureaucracy and improving governance, broad-​based lateral entry as a replacement for the mandarin services could potentially weaken governance and public service delivery. 10.6.5.2 The Case for Lateral Entry But the case for lateral entry is strong too. First, the IAS was designed for the pre-​reform India of a dominant state. The logic of economic reforms that began in 1991 and accelerated with the 2021 Budget is for the state to yield space to the market; it becomes even more imperative for the government to understand the impact of its policies on stakeholders—​ the private sector, the non-​government sector, and the larger public. IAS officers, on the other hand, see the government only from within. Sure, there are efforts to reach out to the stakeholders, but is that an adequate substitute to having within the government itself people who have ‘experienced’ the government from the outside?

236  Possible Solutions Second, IAS officers get recruited at a very young age when it is difficult to test potential administrative and judgement capabilities. Like any real-​ world system, some who are potentially good administrators fail to make it, and some who do make it fall short of requirements. Mid-​career lateral entrants with proven capabilities will help bridge this deficiency. Third, career progression in the IAS is almost automatic. Notwithstanding sporadic efforts to introduce meritocracy, very few get weeded out for poor performance. The only penalty, if at all, for failing to make the grade, is fringe postings. Lateral entry may help push IAS officers out of their comfort zone and challenge them. It is not as if lateral entry has not been tried out; it has, but only at the margin. Both the central government and many state governments have appointed advisers and consultants on an ad hoc basis, for fixed tenures, or even on an open-​ended basis. Indeed some of them, recruited into the Finance Ministry, the Reserve Bank, the erstwhile Planning Commission, and more recently public enterprises, have distinguished themselves. C. Rangarajan and Montek Singh Ahluwalia, just to name two, have been stellar successes. However, the larger experience from such lateral entry has not been happy. Lateral entrants have struggled to fit into ‘the system’ and understand the processes and dynamics of government decision making. They have complained of hostility from the IAS network, which, they believe, sets them up for failure. ‘The system’—​a metaphor for the IAS—​in turn, sees lateral entrants as adversaries who have made their way in, not through an open competitive examination like they have, but because of privilege and connections. 10.6.5.3 A Practical Solution The preceding sections can be over-​simplified as follows: Lateral entry as a substitute for the mandarin civil service is almost certain to fail in Indian conditions. On the other hand, lateral induction of professionals from outside the civil service to a minority of senior posts offers a valuable and necessary addition to the government’s talent pool. However, especially in the Indian context, the risk is that appointments would be made on wrongful or political considerations and those recruited might not work in public interest but in the interest of particular lobbies or vested interests. The allegiance of the civil service to the law

Addressing policymaking challenges—I  237 and Constitution rather than to the political executive (independence and political neutrality) is an important principle in a Parliamentary system of government and this has also been repeatedly upheld by the courts. Any system of lateral induction must preserve this. It is also necessary that the system of induction be fair and in compliance with the norms of equality before law. Therefore, the specific methods used for lateral entry must be such that they • protect the independence and political neutrality of the civil service as envisaged by Sardar Vallabhbhai Patel and enshrined in the Constitution; and • ensure fairness and merit in the manner of induction. This section reviews methods used thus far and suggests a practical way to implement lateral entry in conformity with these principles on a larger and more systematic scale. The Sixth Pay Commission and the 2nd ARC made recommendations for the induction of outsiders such recruitment. In 2008, the Department of Expenditure approved in principle the hiring of outsiders on a contract basis against regular central government posts. The DoPT made a beginning in 2013 and prepared draft guidelines for making these contract appointments.31 Subsequently, in June 2018, the DoPT issued a notification inviting contractual recruitment at the level of Joint Secretary, Government of India, for ten Departments/​Ministries.32 After a process involving screening from over 6000 applicants, in April 2019, the Union Public Service Commission (UPSC) notified the appointment of nine candidates as Joint Secretaries on lateral entry for a contract period of three years extendable to five years.33 These candidates have since joined

31 Department of Personnel and Training, Government of India, Office Memorandum dated 17 June 2013 (http://​docume​nts.doptci​rcul​ars.nic.in/​D2/​D02​est/​AB14​017_​34_​2​ 008-​Estt-​RR.pdf). 32 Press Information Bureau, Lateral recruitment to senior positions in Government of India, 11 December 2018 (http://​pib.nic.in/​news​ite/​Print​Rele​ase.aspx?relid=​186​256). 33 Utpal Bhaskar, ‘Nine from Outside the IAS Set to Become Joint Secretaries in Government’, Livemint, 13 April 2019 (https://​www.livem​int.com/​polit​ics/​pol​icy/​nine-​from-​outs​ide-​the-​ias-​ set-​to-​bec​ome-​joint-​secr​etar​ies-​in-​gov​ernm​ent-​155513​3560​315.html).

238  Possible Solutions their respective postings. Further batches of lateral recruitment have been advertised in 2021. 10.6.5.4 Lateral Entry into the IAS Lateral entry may be done in two ways: firstly, by recruiting laterally to specific or specialized posts within Ministries or Departments and, secondly, by recruiting a pool of persons who will be able to provide leadership and policy inputs in more than one area and may be transferred from one CSS post to another. The recommendations of the ARC and the 2019 recruitment of nine Joint Secretaries cover the former but not the latter. An institutionalized system of annual recruitment into the IAS of mid-​career professionals from diverse career paths would be a better alternative.34 Candidates should be in the 43-​to 46-​year age group and the qualifications should not be too prescriptive in order to privilege proven talent over paper qualifications. The UPSC should design the examination to test analytical skills, maturity of judgement, and personality traits. Lateral entrants, too, like the regular recruits, should be allotted to state cadres, and their inter se seniority should be determined in such a way that the interests of regular entrants and lateral entrants are balanced. A distinguishing feature of the IAS, indeed one that is their claim to pre-​eminence even among other civil services, is their field experience. This cutting-​edge level exposure comes in very handy as IAS officers move up the ladder to policymaking at the highest levels. Some of the Delhi policy mandarins from outside the IAS have never seen a village school or a Gram Panchayat office. That is not just embarrassing; it is dysfunctional. The lateral entrants should therefore have mandatory ‘district immersion’, serving at least three years of their first ten years in field postings. The hard grind of such field postings will make lateral entry self-​ selecting, drawing in only those with commitment and aptitude. Since the lateral entrants will come through an open competitive examination, they will be as much a part of ‘the system’ as the regular entrants, giving no scope for the adversarial relationship seen with ad hoc 34 Duvvuri Subbarao and Gulzar Natarajan, ‘The Case for Lateral Entry’, The Indian Express, 9 August 2017 (https://​indian​expr​ess.com/​arti​cle/​opin​ion/​colu​mns/​the-​case-​for-​late​ral-​entry-​ ind​ian-​adm​inis​trat​ive-​serv​ice-​ias-​upsc-​gov​ernm​ent-​4788​115/​).

Addressing policymaking challenges—I  239 lateral entrants. Besides, the criticism that lateral entrants cherry-​pick their jobs, enjoy a system of revolving doors, come to burnish their CVs, or in some cases even promote private interests from within the government will have no ground. This will have to be complemented with liberalized norms that allow civil servants to work outside government—​with multi-​lateral agencies, non-​profits, and corporates—​for short periods. By enabling exposure to market practices and fresh ideas, this, as much as lateral entry, is likely to help achieve the objectives of lateral entry itself. Besides, this will also enable them to compete on an equal footing with lateral entrants. This lateral entry into the IAS would be a supplement—​not a replacement. There is a lot to be said for continuing to recruit young people into the IAS. They bring in youthful spirit, raw enthusiasm, and unspoilt enterprise into administration. Many of the major programmes that have been mainstreamed across the country such as the mid-​day meal, the employment guarantee scheme, or even right to information are the brainchild of sporadic experiments by these young IAS officers somewhere in the vast hinterland of the country. Yet, an institutionalized lateral entry into civil service will help the government have the best of both youth and experience and take the system closer to the goal of ‘minimum government, maximum governance’. A detailed approach to lateral entry which attempts to meet these objectives is given in the box.

1. Rather than creating a separate new ‘caste’ of lateral entrants, lateral entrants would be integrated into the IAS. This will provide for a. smooth absorption into the system where they would be seen as part and parcel of the civil service; b. better recognition and status for the outside recruits, which would enable them to be effective in dealing with staff and outsiders; c. use of existing training infrastructure and institutions; d. clear norms of seniority; e. ensuring full and automatic coverage of Conduct Rules and other ethical guidelines (this has been a grey area on

240  Possible Solutions











contract appointments, which remains unresolved in the draft guidelines); f. avoidance of ethical and fiduciary risks of entrusting sensitive tasks to short term contract staff; and g. pay scale and other benefits on par with other streams of recruitment. 2. The prestige and perquisites still attached to the top civil services would attract good talent even though the pay would be lower than in the private sector. Such talent may not be attracted by contract appointments alone. 3. This will increase the quality and calibre of the civil services as a whole and help to restore public faith in the system. 4. The IAS (Recruitment) Rules 1954 should be amended to create a new method of recruitment under Rule 4 known as ‘Recruitment by Special Examination of senior professionals’. 5. The IAS Cadre Rules could be amended to create a new cadre known as the Senior Professional Cadre, which will not be attached to any state. However, states may take such officers on deputation with the consent of the Centre, thus preserving the all-​India character of the service; the normal process of releasing officers from state to centre would be reversed with suitable modifications. Since states are not compelled to take them, there would be no objection from them to this new scheme. 6. A suitable cadre strength may be fixed. Initially it may be fixed at 100, with annual recruitment of 25 persons. 7. Since the recruitment is to an All India Service, the selection should be by the UPSC, as per methodology prescribed by the central government. 8. The methodology should allow a wide and open selection of people of excellence from various walks of life but at the same should allow for an objective and fair evaluation of the candidates without undue influence, political interference, or other interference. The eligibility should be such that only persons of some proven merit will be in the zone of selection. The following selection method is suggested.

Addressing policymaking challenges—I  241

















9. Age: Persons aged 40–​50 who are not in civilian employment of the Government of India or any state government at the time of application may be allowed to apply, provided they meet the basic qualifications prescribed. 10. Basic Qualifications: To be eligible for consideration, the candidate must possess a bachelor’s degree or equivalent plus any one of the following qualifications: a. Must have held a post equivalent to Brigadier or higher in the Armed Forces b. Must have held a post of Professor in a University in India or a University of high repute abroad. (The government shall determine ‘Universities of repute’ periodically.) c. Must have been the managing director or whole-​time director for at least two years of a company fulfilling any of the following criteria: i. The company was listed on the National Stock Exchange or Bombay Stock Exchange, or a reputed stock exchange in an OECD country, during any of the last 10 years; ii. The company, at any time in the last 10 years, was required under Section 135 of the Companies Act 2013 to implement the Corporate Social Responsibility policy.* d. Must have been the chief executive (highest executive authority) of a reputed non-​governmental organization (NGO) having at least 1000 full-​time employees and annual expenditure of not less than Rs. 10 crores, both of which shall be evidenced by an audit certificate from the auditor of the NGO. e. Must have worked in the United Nations or its specialized agencies drawing a salary of not less than the minimum of the D-​1 scale for at least two years; f. Must have held a post of the rank of Executive Director/​ Deputy General Manager or of equivalent pay scale in a nationalized bank or any Central Public Sector Undertaking for at least two years (in case of doubt, the Department of Public Enterprises shall decide the equivalence of scale and title);

242  Possible Solutions













g. Must have held a post in any organization drawing salaried remuneration of more than Rs. 10 lakhs per mensem for at least three years. h. Must have been awarded a Padma Shri or any other Padma award. i. Must at any time have been appointed to a Group I All India/​ Central Service through competitive examination conducted by the UPSC (but must not be in service at the time of application). j. Must have appeared in the Common Admission Test for admission to the Indian Institutes of Management and scored a percentile score of not less than 98. 11. Method of Selection: a. Candidates shall be required to submit an application form along with evidence of meeting the basic eligibility criteria; b. Suitable documentary proof may be prescribed as appropriate (appointment order, copy of annual accounts, auditor’s certificate, income tax return, Common Admission Test mark sheet, etc.) but with some discretion to accept alternative documents c. After scrutiny of documents, those eligible shall be required to attend an examination (Civil Services Senior Professionals Examination). 12. Examination: The examination should have a written component consisting of two 3-​hour papers carrying 300 marks each (total 600—​60% weight), which shall be conducted on a single day (as the candidates may not be prepared to spare much time). a. Paper 1 shall consist of an Essay to answer the following question: ‘Describe your work experience in detail covering: i. Major issues, challenges, and problems you have faced in your work; ii. Reasons for various decisions and approaches you took; iii. Important lessons learned from your experience.’ b. Paper 2 shall consist of three questions requiring essay answers: i. ‘What are your practical ideas for making India a better country?’ ii. ‘What are your practical ideas for improving the functioning of the central government?’

Addressing policymaking challenges—I  243









*

iii. ‘What are your practical ideas for improving the functioning of the state and local governments in India?’ 13. Paper 1, by its nature, is a hybrid between interview and examination. It may give some clue to the examiner about the identity of the candidate. This is inevitable, as is the case in an interview. Hence the paper should be valued jointly by a panel of eminent experts who collectively agree on the mark to be given to the candidates. Or, two panels could separately value them and the average may be taken. 14. Paper 2 can be valued in a conventional manner as for any UPSC examination but the examiners should be senior scholars of wide knowledge and repute able to take a very wide perspective. 15. There shall be an interview carrying 400 marks (40% weight). 16. The highest-​ranked candidates based on total marks shall be selected subject to police verification, etc., and then assigned to the Super Time Scale of IAS and placed at the bottom of the scale. 17. They may be notionally assigned the batch of the year whose direct recruit officers are already officiating as Joint Secretaries and placed at the bottom of the batch. They shall form part of that batch for the purpose of promotion to the level of Additional Secretary and Secretary. 18. All those selected under this category shall be deemed to be suitable (empanelled) for appointment as Joint Secretary and Additional Secretary. Empanelment as Secretary shall follow the same process as for others. 19. They may join the New Pension Scheme on the same terms as the direct recruits and shall have the same superannuation age (60).

These are companies with a net worth of >Rs. 500 crores, turnover >Rs. 1000 cr., or profit >Rs. 5 cr.

11 Addressing Policymaking Challenges—​II Overcoming Decision Paralysis

In Chapter 8, we discussed the issue of decision paralysis among officials arising from the activism of oversight agencies. A decision can fail to achieve its objectives due to fraud or incompetence or bad luck. Fraud or corruption can be prosecuted; incompetence or negligence is liable for disciplinary proceedings. But with bad luck, the best one can do is learn and adapt. In this context, any ex-​post examination and adjudication of such decisions should consider three factors. The first concerns making the distinction between intent and outcomes. Policy decisions generally involve weighing the pros and cons of different alternatives and then exercising judgement. It often benefits one group or individual over others. From 1988 till 2018, any pecuniary advantage caused to a private party without public interest by a decision involving them, regardless of whether there was a specific intent to cause such gains, was treated as criminal misconduct. For example, in the coal scam, the mere fact of preferential allocation of coal blocks benefiting certain parties became the basis for a corruption conviction of an officer (known to be personally honest) who was merely following a (bad) policy. The second concerns the application of hindsight to assess policies. Decision contexts can change dramatically. The use of coal and telecom spectrum prices subsequently arrived at through auctions to quantify ‘financial losses’ (presumptive valuation) incurred by earlier allocations through other means is a good example. Among other things, it overlooks the possibility that the initial allocations created the favourable market conditions that generated subsequent higher prices.

State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0011

246  Possible Solutions The third relates to the application of principles that go beyond mere financial considerations in policy decisions. The Comptroller and Auditor General (CAG) guidelines define performance audit as an assessment of ‘the extent to which an organisation, programme or scheme operates economically, efficiently and effectively.’ But assessments on such narrow considerations overlook important political judgements (equity, fairness, relevance, etc.) exercised by democratically elected governments. As can be seen, the so-​ called 4 Cs—​ CAG, Central Vigilance Commission, Central Bureau of Investigations, and Courts—​are central to these problems. Their actions in the aftermath of the resource allocation scams that rocked the country sometime back have convulsed the bureaucracy. For any meaningful attempt to address decision paralysis, all the concerned have to keep those factors in mind while scrutinizing such decisions. The 2018 amendment to the Prevention of Corruption Act 1988, which introduces intent as a requirement for criminal misconduct, is a big forward step. Ex-​post policy assessments by the CAG as part of performance audit should be confined to examining whether public money has been spent in accordance with prevailing rules. It should refrain from exercising judgement on the merits of departmental decisions or policies which have been arrived at by following due process and approved by competent authorities. It is pertinent that the British National Audit Office explicitly ‘does not comment on the merits of policy but aims to conclude whether value for money has been secured’. But it is simplistic to attribute all blame for decision paralysis on the 4 Cs. There are two other important contributors. The first is the tendency for some incoming state governments to systematically re-​open decisions of the previous government and accuse civil servants of wrongdoing. There is no easy solution. The second is the apathy and complacency within the bureaucracy itself. Over time, officers have come to realize that it does not hurt to abdicate and not take decisions. A significant share of officers have settled on an equilibrium where the job has become incidental. It is often rationalized that deciding not to take a decision is itself a decision.

Addressing policymaking challenges—II  247

11.1  Judicial Action In Chapter 8, it was noted that while the Indian judiciary has an excellent track record in many respects, certain kinds of judicial action have contributed to decision paralysis within the government. In this context, the Supreme Court’s reluctance to prescribe explicit and tighter norms for accepting Special Leave Petitions (SLPs), on the grounds that any effort to limit the powers under Article 136 would be tantamount to disturbing the ‘basic features’ of the Constitution, is a matter for concern. Lawyers feel that the acceptance rates of SLPs vary widely across judges and their individual discretion is arguably the critical factor in many cases. This naturally erodes credibility and generates perverse incentives. Therefore, far from diluting its importance, rigorous due diligence, as outlined in a clearer set of internal guidelines which limit the range of individual discretion, would go a long way towards strengthening the Supreme Court’s power under Article 136. The biggest challenge in addressing this problem will be in pushing across relevant reforms. In particular, the most critical thing would be to restrain over-​reach by individual judges. This restraint, to be effective and to preserve judicial independence, should ideally come from within. The issue of judicial activism may therefore be best addressed if it is examined by a Committee of the country’s most credible judges and jurists appointed by the Supreme Court itself. The Committee should, in consultation with all stakeholders, lay down certain principles and rules that limit the range of judge’s individual discretion and draw the line between judicial activism and judicial excess, especially on the issue of entertaining PILs and Suo Motu writ petitions as well as forays into the executive or law-​making domain. The Supreme Court could then mandate all courts across the country to follow those guidelines. A similar process could be followed, and guidelines issued, on the exercise of appellate authority. Another approach is for the higher courts to set strong conventions. The higher courts have a unique role in so far as they not only perform the role of adjudicating on legal issues but also set precedents and create law with their judgements. Judgements of the Supreme Court, more than even legislation, can have a powerful positive effect in alleviating concerns among officials related to the processes surrounding

248  Possible Solutions investigation and prosecution that contribute to decision-​paralysis. It can have a restraining effect on some of the excesses of other institutions too. The decisions of lower courts play an important role in decision paralysis. It is not uncommon to find instances of lower courts not adhering to the precedents and law established by the higher courts and even in egregious violations of judicial propriety. The High Courts and Supreme Court alone are well placed to address these instances and hold the lower courts to account when they come to notice. On an administrative note, given their unitary nature, the High Courts and Supreme Court should respectively hold the Lower Courts and the High Courts accountable in case of complaints about egregious violations of judicial propriety. A commonly observed trend is the practice of inordinate and unexplainable delays in taking up important cases for hearing or passing of long stays that often defeat the purpose of judicial remedy. Further, the conduct of some judges in their courts in passing unrecorded provocative comments on cases as well as on summoned officials of the government has been a matter of much consternation. A very clear and firmly enforced set of guidelines in this regard could go a long way towards enhancing the dignity of the Courts. It is a sign of mature institutions that they acknowledge their failings and put in place mechanisms to rectify them when they come to notice. In a country where such institutional maturity is not always shown, it would be transformative if one of its most respected and exalted institutions, namely the Supreme Court, shows the way with enlightened self-​regulation.

11.2  The Oversight Agencies The amendments to Section 13(1)(d)(iii) of the Prevention of Corruption Act through the Prevention of Corruption (Amendment) Act 2018 were a very welcome step. Since many cases are still being taken up relating to the pre-​Amendment days, the impact of it is yet to sink in among many civil servants. As the Right to Information (RTI) Act nears its twentieth year, it may be the right opportunity to review its implementation. In particular, a

Addressing policymaking challenges—II  249 non-​partisan Committee drawn from centre and states can be constituted to examine, among other things, the possibility of exempting communications related to ‘deliberative processes’ from the Act. Another high-​powered Committee should examine the methodologies and processes of auditors and investigative agencies in light of recent trends and lay down guidelines in this regard. Auditors should preferably confine themselves to examining whether public money has been spent in accordance with prevailing rules and the due process has been followed in decision-​making. It is pertinent that the UK’s National Audit Office ‘do not comment on the merits of the policy but aim to conclude whether value for money has been secured’.1 In addition to financial value for money, policies have considerations like equity and fairness, and most often these criteria matter too. Accordingly, the CAG should refrain from exercising judgement on the merits of departmental decisions or policies that have been arrived at by following the due process and approved by the competent authority. Not only do they not have the competence to make such judgements, but they are also very likely to be swayed by simplistic and first-​ order assessments, which gloss over the deeper considerations that led to such decisions. In particular, they should exercise great caution and have in place adequate administrative controls before constructing counterfactuals and making presumptive valuations. Performance and policy audits should be part of post facto evaluations of decisions and be done by independent and professionally competent third-​party agencies and not auditors. The findings of such assessments should be subject to disciplinary proceedings only if there is prima facie evidence of mala fide in arriving at the original decision. The CAG audits are largely focussed on the loss incurred by the public exchequer due to the actions of government agencies. However, this loss presents only one part of the balance sheet. There is a cost associated with the counterfactual of not taking those actions. It is only appropriate that any cost–​benefit assessment takes into consideration the overall costs, the ones due to acts of commission and non-​commission.

1 An Introduction to National Audit Office, National Audit Office (https://​www.nao.org.uk/​ wp-​cont​ent/​uplo​ads/​2017/​04/​Intro​duct​ion-​Natio​nal-​Audit-​Off​i ce-​04-​2017.pdf).

250  Possible Solutions For example, a performance evaluation may come out in favour of the cancellation of a delayed contract instead of allowing a time extension. The audit may attribute the additional cost due to delay as being due to negligence by the Department. But it does not account for the cost associated with the act of cancellation. So it may be appropriate to also consider the likely cost of premature cancellation. The action may have led to an increase in the overall project cost in a new bid. For example, such cancellation would most likely increase the procurement cost for the government in the next round of bidding, thereby increasing the overall construction cost. Or consider the example of going on appeal on a judicial verdict or an arbitration. It may be a safe thing for officials to exhaust all appeals on a case. However, such appeals come with an associated economic cost. Such costs can be in terms of administrative expenses, the opportunity cost from keeping an economic activity pending, the tax revenues foregone, and so on. An analysis of the tax appeals filed by the central direct and indirect tax departments revealed that the Tax Department was responsible for 85–​90% of all the appeals pending before the Income Tax and Customs and Excise Appellate Tribunals, High Courts, and Supreme Court. However, the Department’s success rate was in the range of just 13–​27%. The Department unambiguously lost in 65% of its cases.2 In both these cases, the prevailing performance audit paradigm incentivizes officials to play safe and keep delaying time extensions or cancel contracts or use appeal as a default provision. In other words, currently audits are skewed towards incentivizing procedural caution. Another thing the CAG could try to do is to expand the scope of analysis in its performance audits. Currently such audits are limited to exploring the losses or costs associated with actions of public agencies. However, in most cases, there are also positive lessons to be learnt. Just as the conventional audits create cautionary precedents for the future, positive learnings can be adopted for emulation as best practices.

2 ‘Ease of Doing Business’ Next Frontier: Timely Justice’, Economic Survey 2017–​18, Volume 1, Chapter 9 (https://​mof​app.nic.in/​eco​nomi​csur​vey/​eco​nomi​csur​vey/​pdf/​131-​144​_​Cha​pter​_​09_​ ENGL​ISH_​Vol%2001_​2​017-​18.pdf).

Addressing policymaking challenges—II  251 Accordingly, it may be useful to have a norm whereby at least some part of the audit report documents positive lessons. Perhaps a fifth or a quarter of the report can be mandated to list out positive learnings. The purpose is to re-​frame audits as a constructive exercise in refining policymaking and administration, one which contributes with both positive and negative lessons, instead of being perceived as a purely ex-​post fault-​ finding exercise. This would also increase the seriousness with which audit reports are taken by the executive—​currently many officers feel that ‘Audit has to find fault and will find fault anyway—​don’t take it too seriously; just follow the procedural routine of responding to it’. If audit occasionally praises, then the criticisms will be taken more seriously. As regards reform of the process of investigating corruption, it has to start with measures to equip the investigating agencies with professional expertise in scrutinizing financial and contract matters. Their composition should become more broad based, including officials with professional expertise in these areas and lateral entrants. It is also essential that officials, especially those involved with high-​stakes policymaking, both at the central and state government levels, be offered adequate legal protections against frivolous investigations and prosecutions thereon. The former demands adherence to a set of permission protocols before the investigative processes can kick in. Most importantly, both investigators and auditors should make the clear distinction between genuine errors and mala fide actions. Finally, there is unlikely to be any dispute about measures to make investigators accountable for their investigations. This assumes great significance in view of the disappointing record of the premier investigating agency, the Central Bureau of Investigation (CBI), in securing convictions on its cases.3 It is important to figure out a way to hold officers accountable for their actions by incorporating conviction into their performance reports, despite the well-​known associated challenges given the long duration of an investigation and the intersecting role of several officials.

3 A.M. Jigeesh, ‘Why Does CBI Have a Conviction Rate of 3%?’, The Hindu Businessline, 9 January 2018 (https://​www.thehi​ndub​usin​essl​ine.com/​news/​natio​nal/​why-​does-​cbi-​have-​a-​ con​vict​ion-​rate-​of-​just-​3/​art​icle​9935​407.ece) (Accessed on 28 December 2018).

252  Possible Solutions As regards the finance veto, there is a need for much better training of officers in Finance so that they focus on overall budget allocations, conformity with prudential rules, reasonable costing and other financial principles, and egregious deficiencies. They should refrain from micro-​ managing issues that are best decided by the competent authorities within the Administrative Department. As regards the issue of the Additional Secretary and Financial Advisor of the Integrated Finance Division of the Department of Expenditure, there are again no simple answers. Some ideas would include a short induction training for all newly appointed AS&FA, which focusses on their dual responsibilities, especially in terms of being a constructive fiduciary agent. They should be made aware of the flexibility and exceptions within the financial rules so as to ensure that they have access to all the opportunities available while scrutinizing expenditure decisions. All these changes have to overcome strongly held conventional wisdom and political correctness. They need to be done with great care and tact such that the delicate institutional balance that is critical to India’s vibrant democracy is not upset. No extent of lateral entry is going to resolve this. Not even change in governments and a strong commitment to good governance can easily correct the incentive distortions created by these trends. It requires foresight and leadership from all institutional stakeholders over an extended period of time.

11.3  Regulatory Institutions: Future Evolution Rubin asserted that the ‘essential administrative character of the modern state is irreversible’4 and that the ‘three branches of government exist only in our minds’.5 He asserted that the three-​branch structure is a problematic and outdated metaphor that inadequately captures the reality of the modern state and which has been retained because of ‘social nostalgia’.6 Rubin argued for alternative metaphors, like that of a network of inter-​ connected institutions. The evolution of Indian constitutional law lends 4 Edward L. Rubin, Beyond Camelot, Rethinking Politics and Law for the Modern State, Princeton University Press, USA, 2005, p. 19. 5 Ibid., p. 15. 6 Ibid., p. 35.

Addressing policymaking challenges—II  253 strong support to Rubin’s contention and reflects an attempt to deal with the emerging contours of the modern regulatory state. A detailed discussion of improvements to the regulatory system is beyond the scope of this book. More than a legislative form or organizational detail, the character and competence of the persons appointed are key. What the American Senator Christopher Dodd said while justifying the expanded discretionary power given to regulators to constrain excesses in the financial services industry during the debate on the passage of the Dodd-​ Frank financial market regulation Bill in the US Congress7 is instructive: We can’t legislate wisdom or passion. We can’t legislate competency. All we can do is create the structures and hope that good people will be appointed who will attract other good people.

11.4  The Minimum Viable Product Approach One of the enduring challenges in development has been that of getting the design of policies and programs right. Governments are always criticized for poor design when a program fails to realize its objectives. The problem is most often with the operational guidelines like identification of beneficiaries, monitoring focus, the release of payments, and addressing delays. The introspection invariably identifies some failings that then get partially or fully addressed in the next iteration of the program. And so on it goes. But this process of program refinement can often be an endless process. Every time you think that the flaws have been identified, something else emerges, and you address them next time round. Clearly, there is no one absolute best program design. In the circumstances, what should a real-​life policymaker or administrator do? Insights from business management literature offer an answer. Businesses have long used the practice of releasing a product with a basic set of features, believed to be enough to appeal to their initial customers. They then collect customer responses through tight and short feedback 7 Binyamin Applebaum and David M Herszenhorn, ‘Financial Overhaul Signals Shift on Regulation’, The New York Times, 15 July 2010 (https://​www.nyti​mes.com/​2010/​07/​16/​busin​ess/​ 16r​egul​ate.html) (Accessed on 28 December 2018).

254  Possible Solutions loops and use the information to refine the product. This initial basic product is the minimum viable product (MVP). This iterative process of product refinement bridges information asymmetry and lowers the risks of failure. While its actual use-​case origins are debatable, it is perhaps in the software industry that this approach gained popularity. It has become even more relevant in the present times with its focus on start-​ups and their development. Tim Harford has this to say about Amazon’s turbulent initial years,8 Bezos combined a grandiose vision with the sketchiest understanding of how the vision could be achieved . . . Bezos was making big claims to his customers... but he didn’t know how those promises were going to be kept. He trusted that they would figure something out. One might have thought that these early weeks were a good time to pause and regroup, to concentrated on making sure Amazon was able to deliver on its early promises. But, like Rommel, Bezos believed in seizing opportunities rather than pausing for breath. In Amazon’s second week of business, he received an email from David Filo and Jerry Yang, the founders of Yahoo. Filo and Yang wanted to list Amazon on the Yahoo home page -​would that be okay? Bezos’s computer guru warned him that it would be like trying to sip from a fire hose. Bezos ignored him and accepted the offer from Yahoo... The workload was inhuman . . . Accepting the Yahoo offer within a fortnight of launch was characteristic of Bezos . . . In 1999, Bezos decided to start stocking kitchen equipment. In Amazon warehouses that had been designed to store, sort and dispatch books, naked carving knives were suddenly scything down the chutes and into the sorting machines. Meanwhile the company’s database would be asking whether the knife was a hardback or a paperback. 1999 was also the year in which Amazon started stocking toys . . . When Christmas came . . . Amazon employees across the US bought Costco and Toys’R’Us inventory in bulk and drove it to the

8 Tim Harford, Messy, Riverhead Books, October 2016.

Addressing policymaking challenges—II  255 Amazon warehouses . . . The scramble was too much for Amazon systems... Products lost somewhere in the vast distribution centres would send Amazon’s databases haywire. Unshipped orders clogged the chutes in the sorting facility, each blockage spawning half a dozen further delays. Internally, the company was on its knees: it launched a ‘Save Santa’ campaign and Amazon staff members were booked into hotels near warehouses (two to a room) and didn’t go home for a fortnight. As Christmas passed, 40% of the toy inventory was unsold and probably unsellable . . . In the summer of 2000 came the dot-​com bust... There was a real possibility that Amazon wouldn’t see Christmas 2001 . . . As Brad Stone writes, “Amazon survived through a combination of conviction, improvisation, and luck”.

Harford points to the concept of OODA, coined by a US Airforce colonel named John Boyd: OODA stands for ‘Observe-​Orient-​Decide-​Act’ -​or, in plain English, working out what’s happening, then responding . . . If you could make quick decisions, that was good. If you had a strong sense of what was going on around you, that was good too.

This approach minimizes the product development cost and commercial risks and allows the business to experiment and elicit the product feature requirements directly from their customers. And given the varying requirements of customer segments in the case of large markets, it helps businesses tailor their products to different market segments. Finally, it also allows businesses to adapt their products to the changes in preferences over time. It can be argued that this is as much relevant in several areas of public policy implementation. The conventional wisdom on the implementation of new programmes and projects is that the leadership plans every implementation step to the last detail, allocating responsibilities and putting in place monitoring mechanisms, and then goes ahead and executes to perfection. Unfortunately, this approach is unlikely to work in many development situations where uncertainties abound. A nation-​wide implementation of any program is a case in point.

256  Possible Solutions

11.4.1  MVP in Government So, taking a leaf out of what businesses do, many public policies and programmes too can be deployed and improved using the MVP approach. All such implementations could be mandated to follow an OODA-​type approach: Create a reasonably good initial implementation plan, an MVP. • Put in place a very good project management unit with some competent officials. • Respond very quickly to emergent problems and refine the implementation plan. • Iterate extensively with short feedback loops and improve the implementation plan as much as possible over as short a period of time as possible. Fundamentally, and of relevance to public policy, it eschews a linear, pre-​planned, and deterministic approach to design and development. Many decades ago, the American political scientist Charles Lindblom wrote an influential paper with the pithy title of ‘The Science of Muddling Through’. He contrasted the ‘root’ method of decision-​making with the ‘branch’ approach. The former relied on a comprehensive evaluation of options in the light of pre-​defined objectives. The latter involved going step-​by-​step and by small degrees from the current situation. He argued that the practical man must follow the latter and that ‘the root method is not usable for complex policy questions.’9 John Kay wrote in 2009 that history had vindicated Lindblom, who was initially ridiculed by those in favour of a comprehensive planning approach.10 The examples of the roll-​out problems faced by three big urban transport projects only highlight the fact that such projects, especially on their implementation side, can be fiendishly complex. The first example is the major overhaul of the public transport system in Santiago, Chile, which was launched in February 2007. The Transantiago scheme involved some 9 Charles E. Lindblom, ‘The Science of “Muddling Through” ’, Public Administration Review 19, no. 2 (Spring, 1959), pp. 79–​88. 10 John Kay, ‘History Vindicates the Science of Muddling Through’, Financial Times, London, 15 April 2009.

Addressing policymaking challenges—II  257 200 km (125 miles) of dedicated bus lanes and the wholesale reorganization of bus routes to integrate them with the city’s metro. But immediately on commissioning all hell broke loose and there were even riots.11 The Transantiago came to be known as a model of how not to reform public transport. It brought misery for commuters: more changes to complete typical journeys, long queues for full buses, and gross overcrowding of the metro. There were many reasons for the failure—​imposition of arbitrary routes that took little account of passengers’ habits, rigid bus routes, inadequate number of buses, inadequate use of pre-​paid smart cards, problems with the satellite technology to control bus movements, etc. In January 2008, the 28 km Gurgaon–​Delhi expressway was opened amidst much fanfare, being projected as a showpiece of Indian urban transportation. The project soon became a commuter’s nightmare, as commute times actually increased.12 The independent consultant to the Project, RITES, got its traffic projection numbers badly off the mark, and the actual traffic on the very first day turned out to be what was projected for 2013. While the Detailed Project Report (DPR), prepared in 1998, estimated traffic in the first year at 80,000 vehicles a day, the actual average daily traffic is now being estimated at 140,000 vehicles. Contrary to the 5% annual traffic growth projected, the actual growth has been more than 9%. With the majority of users not moving into pre-​paid cards, the toll gates could not cope up with the huge traffic. The pilot bus rapid transit (BRT) corridor in New Delhi, which degenerated into chaos immediately after being launched,13 is another example. The 5.8-​km-​long corridor—​constructed on a pilot basis between Ambedkar Nagar and Moolchand in South Delhi—​resulted in massive traffic congestions and an increase in commute times, thereby inviting public wrath. A whole debate got initiated on whether public transport 11 Dario Hidalgo and Pierre Graftieaux, ‘Planning and Implementation Issues of a Large-​Scale Transit Modernisation Plan: The Case of Transsantiago, Chile’ (http://​citese​erx.ist.psu.edu/​view​ doc/​downl​oad?doi=​10.1.1.215.1666&rep=​rep1&type=​pdf) (Accessed 20 December 2018). 12 K.P. Narayana Kumar and Rahul Chandran, ‘Delhi-​Gurgaon Expressway Traffic Stuck in a Numbers Jam’, Livemint, 4 February 2008 (https://​www.livem​int.com/​Home-​Page/​4dG​ Ozne​RczB​2w70​L9Z4​JtN/​Delhi​Gurg​aon-​exp​ress​way-​traf​fi c-​stuck-​in-​a-​numb​ers-​jam.html) (Accessed on 20 December 2018). 13 Rohit James, ‘Far from Global Standards, Here’s What Went Wrong With Delhi BRT’, ITDP India (https://​www.itdp.in/​far-​from-​glo​bal-​standa​rds-​heres-​what-​went-​wrong-​with-​delhi-​brt/​ ) (Accessed 20 December 2018).

258  Possible Solutions systems like BRT were suitable for Indian conditions. The interesting thing about all the three examples, at least the first two, is that these massive engineering projects appear to have failed not in their engineering but on their transport planning side. Transport planning is an extremely complex task, complicated not so much by its technical dimension but by its social dimension. It requires an in-​depth understanding of the specific economic and commercial context of the city, the local transport mix, and passenger habits. Apart from factoring for these major and most obvious stumbling blocks, it is impossible to account for the numerous contingencies that can arise when such massive projects get rolled out. There is limited capacity in India (arguably, even in developed countries) in planning and executing such massive projects. Having been closely involved with the Chennai Metro Railway Project and the BRT project in Vijayawada, from the preparation of DPRs to procurements and initiation of the project execution, the authors think there is a good case for promise of the MVP approach in enhancing the effectiveness of project implementation. In fact, major infrastructure projects invariably face starting troubles. The problems faced at the opening of the Hyderabad and Delhi international airports,14 and the Terminal 5 at London Heathrow,15 are only major recent examples of this reality. People marvel at the engineering dimension of the infrastructure project without realizing that the operational dimension is even more difficult and often impossible to successfully plan and execute. But the initial implementation period is an excellent opportunity to closely study the operational dynamics and get emergent problems remedied. It is critical that all such project management teams have adequate and professionally competent support staff in place during the initial period of implementation of the project. In fact, the success of such grand projects depends on how quickly and effectively we can respond and solve emergent problems.

14 Sandeep Unnithan, ‘Chaos at Airports’, India Today, 25 April 2008 (https://​www.ind​iato​day. in/​magaz​ine/​cover-​story/​story/​20080​505-​chaos-​at-​airpo​rts-​736​136-​2008-​04-​25) (Accessed on 20 December 2018). 15 Peter Woodman, ‘Disastrous Opening Day for Terminal 5’, Independent, 27 March 2008 (https://​www.inde​pend​ent.co.uk/​news/​uk/​home-​news/​dis​astr​ous-​open​ing-​day-​for-​termi​nal-​ 5-​801​376.html) (Accessed on 20 December 2018).

Addressing policymaking challenges—II  259 This approach carries relevance not just to infrastructure projects but to a wide variety of programs. The roll-​out of the Goods and Services Tax (GST) and the implementation of the Insolvency and Bankruptcy Code (IBC) are very good examples of a situation where the MVP approach is relevant. Given the complexity of responses of various stakeholders and resultant uncertainty, the best that governments at centre and states can do is to roll out the platform and keep the powder dry so as to respond to emergent problems. The success or otherwise of the roll-​out is critically dependent on the swiftness and competence of the response. Interestingly, in both cases, acknowledging the complexity of implementation, something close to an iterative approach was adopted. In the case of the GST, its inter-​state Council provided an institutional forum for exactly such revisions. It is true that operational guidelines are often revised in all programs. But these are one-​off and random efforts. Instead, the point being made here is an explicit and time-​limited focus on iterative refinement communicated to all implementing agencies and a dedicated team to respond to the emergent suggestions to quickly incorporate them and refine the program design. Indeed, the first months of implementation should therefore be a safe space where implementing agencies are allowed the flexibility to iterate and figure out the problems and alternatives. This way, some of the critical operational details can emerge bottom-​up from the perspective of effective implementation.

11.4.2  Why the MVP Approach? Like with business products, policies and programs too are formulated under considerable uncertainty and there is very little that can be done to address them ex ante. In fact, there are at least three very dynamic dimensions for most programs (Figure 11.1). One, even more than businesses, in large nation-​wide or state-​wide implementation, the contextual variations of different geographical regions are very large. Second, most programs have different target groups, whose preferences and responses vary based on the program design. Finally, there are general equilibrium dynamics that demand periodic tweaking of program design to address emergent concerns over time.

260  Possible Solutions

Program Design

Geography Population Category Time

Figure 11.1  Three dimensions of program design

But this approach goes against the entrenched belief within bureaucracies that policies and programs, at least for their duration, are cast in stone. The norms and components of the programs are held sacrosanct and program guidelines are issued to ensure fidelity of implementation. While guidelines are often revised, they rarely involve major changes to the original program design. It would require a paradigm shift for government bureaucracies to adopt the MVP approach. There are at least two entrenched features of bureaucracies that would need to be overcome. One, bureaucracies prefer to freeze a program design and focus on the implementation. Iterative adaptation of program design while also monitoring implementation could stretch bureaucratic bandwidth. Two, the ability of bureaucracies to respond quickly to emergent challenges and make revisions to the program design is much less than that of businesses. For example, some of the program design changes may require a Cabinet approval or even a legislative amendment, as was the case with the iteration on the IBC. Even changes to guidelines may require approval by the competent authority, generally the Minister or the Secretary. However, neither of these are insurmountable. The first is a matter of overcoming inertia. After all, it is not as though the entire program design is up for revision. It is, perhaps, just a handful of elements, and that too with most of the work being in the initial months of the program roll-​out. The second is a matter of agility, and here too practical wisdom informs that most of the changes required are likely with operational guidelines, which are less difficult to address, given that it only requires departmental approvals.

Addressing policymaking challenges—II  261 An acknowledgement of the need for such iterative adaptation and its internalization into the bureaucratic culture would itself be a big step with several positive externalities. It would create the conditions for the emergence of a genuinely bottom-​up approach to program design. Apart from being able to address the three program dimensions in a more satisfactory manner than the current top-​down cast-​in-​stone approach, the MVP approach is likely to create positive incentives. Foremost, it is likely to mitigate the risk-​aversion that characterizes bureaucracies. By allowing for course corrections, it frees the policymaker from running the risk of being locked into a particular program design element which, while likely to be very effective, would also have been highly risky and hold the officer accountable for any likely failure. Features like delegation of decisions, deemed approvals, default positions, etc., can significantly enhance the quality of implementation but also run the risk of weakening implementation quality and increasing corruption. An MVP approach allows the freedom to start with the most flexible and effective approach and gradually tweak them to minimize distortions. Further, it also focusses the collective consciousness of the system on outcomes. In the earlier case, since the program design was largely frozen, the stakeholders had only one option—​make the best with what they had. It also routinized and confined monitoring to the process and output indicators which too were more or less fixed along with the program. A culture of iteration would free at least some minds along the bureaucratic chain to think more critically about the program in real-​time and surface emergent operational and outcome concerns. Also, even if someone tries to consciously game the program design, there is always the freedom available for someone else to highlight the problem and demand a course correction. In the current dispensation, this would have been taken to represent insubordination or cynicism about the program. Finally, policy design, in many cases, is less about binary choices and more about choosing along a spectrum—​nature of nudge, optimal tariff, effective integration, robust design, and so on. They generally involve figuring out operational elements on multiple dimensions, not a problem that is amenable to resolution through field experiments. It is of course possible that, in a few cases, there are certain important operational

262  Possible Solutions elements whose identification could do with field experiments. Further, these choices are intimately intertwined with issues of state capability. In fact, they are invariably context dependent and have most likely limited generalizable learnings. Therefore, the most practical approach to addressing them may be to start with an MVP and let the best choice for each context emerge through an iterative process as the implementation proceeds. This approach would also resolve the challenge of integrating evidence into practice and ensure truly evidence-​based policymaking. One of the biggest challenges with the proposed MVP is public expectations. The public, and even more so the press, in India expects new programmes to be perfect from day one and views course corrections as evidence of poor planning or implementation. Therefore, clearer expectation-​setting is imperative if this approach is to be used successfully.

The ‘Hiding Hand’ and Public Policy

There are two problems with addressing fundamental economic or social challenges such as pervasive informality. First, its resolution requires the coordination of several complementary policy levers. The sheer complexity of the challenge generates enormous inertia against any action. Even the action of identifying various reform interventions, sequencing them, and formulating an action plan can be a massive challenge. Second, at least some of the policy changes are likely to be unpopular or would have to overcome entrenched and powerful vested interests. It is in this context that demonetization assumes significance. The underlying reasons and intentions, the implementation issues, and its costs and benefits have been widely debated in India. Reasonable people can and do have differences about all these and many have been very critical. However, it would be far less contentious to claim that demonetization has performed better than conventional approaches in raising structural issues like rampant tax evasion, pervasive informality, low tax base, and predominance of cash transactions to the top of the policy agenda in India’s rambunctious democracy. It is unlikely that measures to discourage the use of shell companies for money laundering, promote information sharing between tax departments and artificial intelligence to analyse suspicious bank transactions,

Addressing policymaking challenges—II  263 encourage the digitization of transactions, and make technological changes to enable the poor to receive and make digital payments through mobile phones would have been undertaken with the same speed and urgency without demonetization. This approach evokes the political economist Albert O. Hirschman’s principle of the hiding hand.* Hirschman said that governments are often either too risk averse or too intimidated to undertake large, complex endeavours. He argued that in such cases, it would help to have a ‘hiding hand’ that encourages policymakers and politicians to underestimate challenges and nudges them to pull the trigger. Although Hirschman’s examples may have been about massive infrastructure projects with long construction times, site acquisition and rehabilitation challenges, and massive expenditures, his reasoning applies just as well to structural policy reforms. * Albert O. Hirschman, ‘The Principle of the Hiding Hand’, The Public Interest (www. nationalaffairs.com/​public_​interest/​detail/​the-​principle-​of-​the-​hiding-​hand) (Accessed 20 December 2018)

11.5  Designing for Scaling of Complex Public Policies While the MVP approach may be useful in cases with implementation uncertainties, it still remains a challenge to figure out the MVP itself. One useful method to develop the MVP and then improve it is described below. Implementing policy in the context of a country like India can be a massive challenge. India’s continental size, large population, and vast diversity coupled with the inherent complexity of many public policy challenges mean that there are very few universal strategies to address them. It is impossible to have one all-​encompassing policy design and implementation strategy for any problem that satisfies the requirements of all categories of states and people. Context matters and policy and implementation designs have to be tailored accordingly. Public policy initiatives should, therefore, avoid one-​size-​fits-​all approaches, allow flexibility in design, and accommodate greater discretion during implementation. In this context, Matt Andrews, Lant Pritchett, and Michael Woolcock had formulated a Problem Driven Iterative Adaptation (PDIA) strategy

264  Possible Solutions to describe a flexible approach towards solving social problems.16 It rests on four core principles—​promote local solutions for local problems; create environments that allow experimentation and emergence of positive deviances; encourage experiential learning, iteration, and adaptation; and scale through subsequent diffusion. The underlying principle has great relevance in the case of policymaking on certain public issues. It is useful to embrace elements of this model of development across sectors, in small pilots, to create the conditions for both encouraging innovation as well as its scale-​up. For certain large initiatives, the Government of India (GoI) should disseminate, in an accessible manner, various models of intervention and all actionable templates and supporting documents that states can readily take off the shelf and implement. This would involve preparing very brief multimedia kits and actionable policy memos explaining the implementation details and benefits of each intervention, illustrative implementation plans, model procurement and contracting documents, hardware and software technology specification standards, and some form of hand-​holding support in accessing certain very specialized services. A web portal can host all these services and facilitate interaction between the GoI and the sub-​national entities. Simultaneously, it should put in place mechanisms that foster competition among states/​districts/​cities. The sub-​national entities should be encouraged to innovate with policy design and implementation, using technology and external expertise, through public-​private partnerships and collaborations with non-​profits, and so on. This becomes all the more essential since public systems are acutely enfeebled. In many cases, the prevailing service delivery models and systems are irreparably damaged and we may need to junk them completely and embrace new models of engagement. In a country with 36 provincial governments, over 730 districts, and 4,041 towns, it is almost certain that at least a few would embrace any initiative. Further, in each such sample, at least one or two would iterate and effectively implement the intervention over a five-​year duration. The 16 Matt Andrews, Lant Pritchett, and Michael Woolcock, ‘Escaping Capability Traps Through Problem Driven Iterative Adaptation (PDIA)’, Building State Capability, Center for International Development, Harvard Kennedy School (https://​bsc.cid.harv​ard.edu/​publi​cati​ons/​escap​ ing-​cap​abil​ity-​traps-​thro​ugh-​prob​lem-​dri​ven-​iterat​ive-​ada​ptat​ion-​pdia-​0) (Accessed on 28 December 2018).

Addressing policymaking challenges—II  265 forces of competition would amplify this dynamic. They could in turn become potential champions for emulation by others and gradual nation-​ wide scaling up of the intervention. A confluence of several such positive deviances across a wide spectrum of interventions, over a period of time, rather than a prescriptive top-​down mission mode strategy for each innovation, stands the best chance of nation-​wide scale-​up of innovative public service delivery interventions. Fortunately, the kernel for such an approach is available with the ease of doing business (EoDB) and Swacchh Bharat Mission rankings. Similar rankings of states on student learning outcomes and logistics performance have also been initiated by the NITI Aayog. It is required to take these rankings down to the district levels and support the states and sub-​ state units with implementation. In the case of the Swachh Bharat Mission (SBM), the then Ministry of Urban Development created a portal that became a technical-​assistance and best-​practices dissemination window for all cities.17 This portal had e-​learning modules for frontline officials on various aspects of public health with certification systems for completion of those modules. It had model tender and contract documents for a large number of procurements. It also consolidated the various best practices and prepared small documentary multimedia on innovation approaches. Interested cities had all the technical support required to move forward with a locally prioritized SBM implementation agenda. Finally, it also promoted competition among cities through the annual Swachh Survekshan rankings. Such flexibility will invariably cause disequilibrium and disruption. There will doubtless be failures. But it stands a far better likelihood of throwing up sustainable public service delivery models than the current top-​down, norms and components–​based strategies. Further, it would be in line with the central government’s stated policy of a cooperative federalist relationship. To some extent, such ‘marginal revolutions’ can be calibrated and expedited by setting in motion competition among participants. Over a period of 10–​15 years, this process can generate a virtuous cycle of cross-​fertilization of innovative solutions where cities learn from and



17 Swachh Bharat Mission Urban (http://​swachh​bhar​atur​ban.gov.in).

266  Possible Solutions emulate each other’s smart city interventions. As an illustration, we feel this strategy can be adopted for the likes of the following:

1. 2. 3. 4. 5.

Ease of doing business improvements Municipal governance reforms Improving learning outcomes Health care services delivery E-​governance for statutory services delivery.

The following illustrative action plan for implementing this approach: 1. Finalize a list of interventions. 2. Develop a high-​quality interactive web portal. The portal should have a provision for a dashboard/​microsite for each city/​district/​ state as well as host a solutions exchange. 3. Prepare multimedia materials explaining the concept and implementation plan in multiple languages. 4. Prepare and make available model documents applicable for procuring services and contracting for interventions, where required. Also make available open-​ source software wherever possible. Benchmarks for services levels and outcomes can also be published. 5. Disseminate the intervention communication materials through the portal. 6. The portal could also have short e-​learning modules and videos explaining the common problems and challenges. 7. Cities/​districts/​states may register their intervention and implementation plans on their dashboard/​microsite. 8. The cities/​districts/​states could leverage financing from various public programs, besides their internal resources, as well as forge public-​private partnerships. 9. Annual awards to cities/​districts/​states on performance of the implementation of the interventions. This strategy would be in line with the most efficient leveraging of resources and expertise. The GoI is best placed to support with high-​level technical assistance in the form of exposing the implementing agencies to best practices and innovations and then supporting them with its

Addressing policymaking challenges—II  267 implementation. Interested implementation agencies could choose to implement innovations which are meet their requirements and are also suited to their respective contexts.

The Million Marginal Revolutions of China

The conventional wisdom on development is that institutions are what determine national development and economic growth. Institutions set the rules of the game, either formal or informal. And the path to progress is to work within the scope of those institutions. While this path is hardly linear, the institutional guardrails determine the broad trajectory of development. At some level, this assumes at least two things. One, there are some rules of the game which are handed down and which are most important to success with the endeavour. Two, there is something certain and perhaps permanent about the thing called an institution. What if both are questionable? In a fascinating study of the spectacular economic development of China, Yuen Yuen Ang unsettles this conventional wisdom* Ang argues that development is a ‘co-​evolutionary process’, where institutions of the society, policy, bureaucracy, and the economy constantly interact to change each other. In particular institutions and markets interact and evolve together. This in turn means that effective co-​evolution requires adaptability within institutions. Ang shows that China’s success lay in the fact that at the start of its modern development trajectory, the country benefited from ‘weak’ institutions. In fact, this means getting institutions ‘wrong’, and having an appetite for change and adaptation, is perhaps better than getting them ‘right’. The ‘right’ institutions come in the way of change. And the ‘wrong’ institutions allow for change! As Lant Pritchett wrote in the blurb, ‘harness weak/​wrong/​bad institutions to create markets’! Ang uses three processes of co-​evolution—​variation (how does the system throw up alternatives and options to deal with particular problems?); selection (how does it select among the variants to form new combinations?); and niche creation (how does it craft new, distinct and valuable roles within the system). Allow a million flowers to bloom, bless those that survive, and create rules of the game to support them. Echoing Ang, Ronald Coase and Ning Wang too trace the evolution of China’s economic growth.** They examined the country’s growth

268  Possible Solutions since the late seventies and refuted the conventional wisdom that it was driven by an omnipotent Communist Party through tight central planning and benign leadership of a group led by Deng Xiaoping. Instead, they point to a decentralized and flexible model of growth, which allowed experimentation with ideas and encouraged competition among cities and provinces. Indeed most of these initiatives failed, but not closely yoking them to high-​profile central programs allowed the space for experimentation and risk-​taking. This was essential for iterative adaptation that is critical for refining the implementation design of large-​scale policy interventions. In fact, each of China’s three biggest socio-​economic interventions—​agricultural de-​collectivization, township and village enterprises (TVEs), and special economic zones—​ emerged iteratively from bottom-​up, marginal initiatives. But once the initial TVEs and Shenzhen emerged unobtrusively from such marginal initiatives, the central government stepped in to guide the course of their development and scale up the model. *

Yuen Yuen Ang, ‘How China Escaped the Poverty Trap’, Cornell Studies in Political Economy, 2016. ** Ronald Coase and Ning Wang, How China Became Capitalist, Palgrave Macmillan, 2012.

11.6  Simplifying Policy Design There are several reasons for the popular perception of rigidity associated with administrative regulations. This section focuses on two attributes of policy or program design, which are both responsible for considerable inefficiencies and are also largely avoidable. One stems from the urge to retain control of execution processes. In Chapter 10, we talked about the issue of control in the context of execution. This is about control in the design of the policy itself. They include eligibility or assessment norms for public benefits, permissions and licensing conditions for various activities, and common statutory certificates. Another arises from the urge to adopt some universal and aspirational standard. They are mainly focussed on regulations on safety and other preventive safeguards and standards and specifications with equipment and facilities.

Addressing policymaking challenges—II  269 An important consideration for policymakers formulating regulations and program guidelines is to retain control over execution. At a fundamental level, it is a basic objective of any administrative rule. So, the issue is not about the inherent merits of any such requirement. Instead, it is about where to draw the line between what is necessary and what is avoidable. The eligibility requirements for welfare or other benefit programs are often shrouded in indicators that are too many and/​or difficult to assess or measure. The motivation is to limit false positives, or inclusion errors. While this is laudable, beyond a point the requirements turn counter-​productive and detract from the other important consideration of limiting false negatives, or exclusion errors. Examples vary from the requirements to access a ration card to loan guarantees under a stimulus program for a particular group of beneficiaries or housing loan under an affordable housing program or giving export refunds under GST. Complicated and excessive eligibility requirements, coming from a mindset of limiting inclusion errors, have over the years been the bane of many well-​intentioned programs which have failed to take off. Incidentally, one of the reasons why Collectors by and large across states do a reasonably good job of managing disasters is the simplification of procedures, delegation of powers, and the system’s tolerance for some degree of leakage and pilferage in relief activities. Similarly, the requirements for permissions and licenses are often excessive and onerous in the form of documentation, procedures, and physical inspections. Again, the idea is to ensure that the permits or licenses are conferred only to those who have the capacity to perform the activities and are in conformity with requirements. Driving licenses, mining permits, building permissions, trade licenses, and so on are examples. Perhaps the most draconian examples of such control focus are found in the erstwhile labour laws. The Factories Act 194818 in Chapter III contained some safeguards on health which were some of the most prescriptive anywhere. It had provisions for the frequency and nature of painting of walls, quality of water used for humidification, height of internal walls with glazed tiles in bathrooms (90 cm), and even spitoons. One provision even talked about preventing ‘the formation of shadows to such an extent

18 The Factories Act, 1948.

270  Possible Solutions as to cause eye-​strain or the risk of accident to any worker’. And these are not just part of the regulations but are part of the Act itself. Consider the regulations governing the licensing of contractors supplying contract labour.19 Contractors have to obtain multiple licences for each work order within a state and separate licences for operating in various states (even if it’s for the same organisation). The employers have to give contractors a certificate for each work order. The tenure of the licence differs among states. Every time there is a change in the headcount, the contractor has to go back to authorities to update the licence owing to the fee structure attached to it —​a process as good as applying for a fresh licence.

In stark contrast, in countries like China and Vietnam, minimum qualification requirements are defined and contractors licensed. Unlike these, far from safeguarding labour from the contractors, the Indian regulations end up hurting the interests of labour. The intent behind such prescriptive regulations is generally positive, with the interests of workers in mind. But given their very nature and the context in which these regulations operate, the end up adversely impacting the very stakeholders they intend to protect. Instead, a more effective alternative may be to define certain basic standards and associated eligibility requirements and then trust the stakeholders while complementing them with effective oversight. Finally, the commonest source of such red tape is the regular statutory certificates issued by public agencies. Excessive, and often superfluous, documentation requirements are a common source of frustration among applicants. What makes this very important is that all citizens have to access these certificates, thereby making them a ubiquitous source of discontent. They include the likes of birth and death registrations, land records, and passports. They also include categories like income certificates, which are perhaps issued only in India as a statutory service by at least certain state governments. 19 Somesh Jha, ‘India’s Contract Labour Laws Are Riddled With Incentives for Corruption’, Business Standard, 17 June 2020 (https://​www.busin​ess-​stand​ard.com/​arti​cle/​compan​ies/​india-​ s-​contr​act-​lab​our-​laws-​are-​ridd​led-​with-​inc​enti​ves-​for-​cor​rupt​ion-​120​0616​0186​1_​1.html) (Accessed 3 October 2020).

Addressing policymaking challenges—II  271 As mentioned earlier, the fundamental objective with all these three categories is to retain control. However, it needs to be acknowledged that control is a double-​edged sword. Too much of it is as much inefficient and ineffective as too little. They are therefore among the biggest contributors to the perception of excessive regulations that are often described as a feature of doing business in India. The adoption of universal or aspirational standards in legislations and policies often creates large distortions and anomalies. In the case of safety and preventive issues, public policy often takes a maximalist view, whereby regulations are designed to avoid any damage or offer all protections. These are most often borrowed from developed countries. They are not grounded on prudent cost–​benefit assessments and are also inappropriate for the local context. For example, the de jure safety regulations for many types of commercial facilities are drawn based on those in more developed contexts. However, the de facto compliance in all these cases is minimal. This, apart from the legal issue of non-​compliance, also engenders a moral hazard of permissiveness—​everyone knows that the standards are too onerous as to be effectively meaningless. Another set is the standards and specifications on services, products, and infrastructure facilities. These too are often borrowed from developed countries or international benchmarks. It can be legitimately argued that the curriculum standards expected in the earlier grades in schools in India are prohibitively high. Or the occupational certification for vocational skill jobs places too excessive a focus on college education qualifications. In the case of infrastructure facilities and equipment, a common problem is that of excessive technical specifications (larger land or built-​ up area, more material strength, higher bandwidth, greater storage capacity, etc.). These are most often driven by vested interests involving nexus between politicians, officials, consultants, and private suppliers. In these cases, it needs to be borne in mind that more is not necessary when less is enough. Some regulations are both aspirational and seek to control. The licensing requirements for higher education and medical education are governed by the regulations issued by the University Grants Commission, the All India Council for Technical Education, and the Medical Council

272  Possible Solutions of India. These regulations are characterized by both unrealistic and rigid standards on a variety of inputs like land availability, built-​up space, faculty qualifications, associated infrastructure, and so on. Further, they also micro-​manage with tight prescriptive norms for ratios of different categories of faculty, instruction time for courses, and small details of coursework requirements. In either type of case, it is important to acknowledge that higher requirements and standards come with costs, which raises access or entry barriers and also incentivize efforts to evade them. In each of them, a cottage industry emerged, specializing in various ways to skirt around the regulations. In other words, the very set of requirements and standards aimed at realizing certain objectives end up detracting from those objectives. There are different ways in which some of these problems can be addressed. Technology and e-​governance solutions have great promise in addressing some of these concerns. Another approach would be to rely on self-​assessments and self-​certifications and thereby dispense with the onerous documentation proof requirements. Ultimately, the issue of controlling is also a reflection of mindsets. In particular, two important requirements among all stakeholders in this regard are trust in stakeholders and an appropriate enough level of tolerance for failures and leakages. The latter is especially important since no solution or system will be watertight. Over a period of time, systems can potentially move over to a more efficient equilibrium.

12 Becoming Better Policymakers—​ Behavioural Change Policymaking structures and processes are important. However, the skills, attitudes, and behaviour of those who make or advise on policy also make a big difference to the quality of policymaking. This chapter makes suggestions on how middle-​level and senior civil servants can become better at policymaking. The unique career progression of IAS officers poses a challenge in terms of developing the right approach to policy work. IAS officers typically start their careers as field-​level functionaries entrusted with leading the implementation of various government programs and schemes. In these situations, they have clearly laid down rules and guidelines that define the implementation path. They are program executors and need to get things done by following pre-​defined rules of the game. After a decade or so of such sub-​state level implementation, they move to the state or federal government secretariats to function as policymakers. In this role, they have to formulate policies and design programs and schemes. In other words, they have to transition from being a program implementer to being a policymaker. This abrupt transition assumes that program implementation endows officers with all the requisite capabilities to become effective policymakers. This arises from the belief that policymaking is about policy implementation on a bigger scale. But in fact, as we shall see, the behavioural attributes necessary to succeed with execution may not only be inadequate but can often become stumbling blocks in being effective as policymakers. In simple terms, program execution is about objects and agents, whereas policy design adds the layer of the system within which the

State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0012

274  Possible Solutions Program Implementation Vs Policy Making Skills

Program implementation skills/attributes required for policy making

Program implementation skills/attributes

Additional skills/attributes required for policy making

Policy Making skills/attributes

Figure 12.1  Program implementation vs. policy making skills

objects and people are located (Figure 12.1). This addition creates several complications. In this context, Lant Pritchett’s metaphor of ‘baby ontology’ is instructive.1 Children relate with objects and people. Objects have properties and people have preferences. But they struggle with understanding the system in which they live. Similarly, as program implementers, young officers relate quickly with both objects and agents but may struggle to appreciate the system. In fact, they anthropomorphize even organizations. For example, people have internalized organizations as an agent—​so you have the ‘head’ of the organization. Whereas objects and agents can be ‘controlled’ through some supervisory and monitoring systems, the system is an endogenously evolving entity and defies efforts at ‘control’. Just as not all children ‘grow up’, there is nothing pre-​ordained about good program implementers becoming good policymakers. The escape from the baby ontology is not automatic and demands learning and equipping oneself with a set of skills. Being conversant with objects and agents is critical to be able to engage with the system. But this is only a necessary but not a sufficient condition.

1 This draws on a discussion with Lant Pritchett.

Becoming Better PolicyMakers—Behavioural change  275 The focus here is on the behavioural attributes and practices that can be, especially but not exclusively, relevant in program design. In particular, this section is concerned about specific behavioural traits that can enhance the quality of program design as well as certain others found in implementers, which can be detrimental in policymaking. Certain commonly observed biases that detract from objectivity and professionalism are also discussed. While written with IAS officers primarily in mind, given their idiosyncratic career progression, these attributes are relevant as much to all civil servants occupying policymaking roles. There is a rich and growing body of research on behavioural psychology, which illustrates that human beings are vulnerable to several cognitive and other biases. Such biases form and reinforce patterns of deviation from rational judgement about people and situations based on prior personal experiences. They include generalizing from a handful of data points, overestimating one’s own judgements, being anchored to personal experiences or anecdotes, reluctance to give up inferences formed from personal experiences, psychological comfort with logical consistency overlooking practical considerations, and plain hubris. All these invariably ‘cloud’ our judgements. Needless to say, the authors too have been afflicted on numerous occasions by all these failings. To some extent, this is therefore an autobiographical consolidation of failings and learnings from those experiences.

12.1  Bridge the ‘Trust Deficit’ We need to have detailed guidelines, with unit costs, so that we can be sure that the financial allocation is not misused. We need to have a majority of direct nominees of the Government in the University’s Board so as to ensure that we exercise control.

There is a very rich body of literature that has examined the importance of trust in the performance of societal organizations. Trust is described as the ‘propensity of people in a society to co-​operate to produce socially

276  Possible Solutions efficient outcomes and to avoid inefficient non-​cooperative traps.’2 The same trust is as important to the effective functioning of bureaucratic organizations as societal and private organizations. A characteristic feature of bureaucratic organizations in many developing countries is the near-​complete absence of trust of both its own personnel as well as those seeking to be regulated. Officials at each level, in general, are taught to feel that their subordinates cannot be trusted to honestly discharge their responsibilities. And the bureaucracy as a whole works on the assumption that their clients—​citizens, businesses, private institutions, etc.—​cannot be trusted to conform to prevailing rules and regulations. The bureaucratic rules of the game are accordingly tailored to limit discretion. Apart from coming in the way of delegation, this also engenders an urge to regulate away discretion in the hope that it would lead to discipline. It is therefore no surprise that civil servants, steeped in this ethos, generally struggle to trust their sub-​ordinate systems and personnel. Unfortunately, this stifles local initiative and creates unforeseen perverse incentives besides prohibitive costs. At the level of the regulated, citizens are forced to validate their documents on any application through a series of arcane processes, businesses are mandated to submit elaborate compliance reports, and private institutions are intensely micro-​regulated. All these inflict enormous economic and social costs, both in terms of enforcement and compliance costs. Consider the requirement for attestation of copies of a certificate by a gazetted officer or a notary to access any government service. It is possible to view the problem in two ways. There is an entrenched belief that in the absence of this safeguard, at least some citizens are likely to forge certificates and submit fraudulent copies. So, from the government’s perspective of deterring fraud it appears reasonable to mandate this requirement. But from the citizen’s perspective, the associated additional cost is large. It appears all the more unreasonable given that the vast majority of citizens who are honest are being penalized for deterring a minuscule minority. And the deterrence is itself deeply questionable, whether in 2 Rafael La Porta, Andrei Shleifer et al., ‘Trust in Large Organizations’, AEA Papers and Proceedings, May 1997, pp. 333–​338 (http://​scho​lar.harv​ard.edu/​files/​shlei​fer/​files/​trust.pdf) (Accessed on 18 December 2018).

Becoming Better PolicyMakers—Behavioural change  277 discouraging those likely to indulge in such fraud or in punishing those certifying them (not to mention the inefficiencies and distortions from the emergence of a market in attestors and touts). A cost–​benefit assessment is unlikely to come out in favour of attestation. It was only in 2015 that the government took the leap of faith to dispense with attestation and accept self-​certification. The former approach is purely regulatory, making everyone suffer for the likely actions of a handful, whereas the latter approach is facilitatory, trusting the citizens and easing access to public services while finding other means to prevent fraud. The trust deficit entraps the system in a deeply sub-​optimal equilibrium. The lack of trust within the bureaucracy and on the clients stifles local initiatives and increases regulatory layers. But reposing greater trust is impeded by the strong belief that lighter regulation will lead to corruption and widespread abuse. Interestingly, this assumption is more a function of entrenched convictions than of any evidence. There is a rich and growing body of evidence that leaders who exhibit trust and are able to delegate responsibilities are more likely to develop cohesive teams that can perform at high levels of efficiency. In the case of governments, reposing trust in citizens and pruning down regulatory layers would reduce harassment and lower transaction costs: a good example of sound economics and sound politics. Therefore, a more effective strategy would be to fix accountability, align incentives, ensure transparency, and delegate to allow discretion at the cutting edge. Things will possibly go wrong. That cannot be helped. But incorporate active feedback loops, put in place response systems that allow for continuous implementation design refinement, and strongly encourage the bright spots that emerge.

12.2  Enabling Legitimate Private Benefit Is Good Public Policy If we don’t ask them to produce attested certificates, there is likely to be a large number of fraudulent applicants.

278  Possible Solutions There is a problem with how those within the government sometimes view those outside the government. The latter includes both the for-​profit private sector and the non-​profit sector. The stereotype is that of a zero-​sum game between public and private benefits. Private benefit is thought to come at public cost. This can be characterized as shown in Figure 12.2. In this paradigm, private benefit necessarily happens outside the space of public policy. Any public policy is viewed from the lens of whether it is benefiting any private individual or entity, and if so, restrictions are imposed to eliminate, or at least limit, the private benefit. So, any regulation has the explicit objective of limiting potential damage by the private interest. Laws get enacted to catch the deviant. Accordingly, any citizen seeking a service is viewed as a private benefit–​seeking individual and therefore engaging in a transaction where private benefits have to be limited. This is irrespective of whether the individual is seeking a service that he/​she is rightfully entitled to, and the public agency’s real mandate is to deliver this service to the citizen. What makes this worldview even more corrosive is that it can tend to view public benefit in terms of denying private benefit. Even where there is a law or rule in place, avoiding its subversion takes precedence over upholding its delivery. So, a land acquisition award is less about following the law and giving the land loser the rightful amount for taking their land and more about

Public policy space (As often viewed now)

Public Benefit

Private Benefit 1

Private Benefit 2

Private Benefit 3

Figure 12.2  Perception of public policy space

Becoming Better PolicyMakers—Behavioural change  279 minimizing public expenditure. Or issuing a certificate or license is less about ensuring hassle-​free access to a statutory service and more about limiting fraud or wrongful claims. Or even a single-​window application interface to improve ease of doing business soon becomes overtaken by the desire to put in place controls to screen ineligible applicants. This worldview is given credence and amplified by two things: First, the general behaviour of many private citizens and private companies (and some of their advisers, lawyers, consultants, accountants, etc.) who tend to seek private benefit at public cost. In repeat games, these behaviours, in turn, shape expectations and push all sides into a self-​reinforcing bad equilibrium. Three centuries of colonial rule (by a government without moral legitimacy) may perhaps have contributed to a view where cheating the state is morally acceptable. Second, this worldview allows certain public officials and representatives the fig leaf for rent-​seeking. And this, in turn, perpetuates the rent-​seeking structure that is pervasive in public engagements across levels. Third, for thirty years, from 1988 to 2018, the law (Prevention of Corruption Act) provided for ‘pecuniary gain to any person without public interest’ to be treated as a criminal act of corruption. Thankfully, this was repealed in 2018; but a whole generation of civil servants, auditors, and prosecutors have grown up internalizing its import. It may take a while for the effect of the change to sink into the collective psychology of doers and checkers. An alternative worldview is depicted in Figure 12.3. From this perspective, any public policy or public engagement is explicitly acknowledged as an interaction that may create public and private Public policy space (ought to be) Private Benefit 4

Private Benefit 1

Private Benefit 3

Public Benefit Private Benefit 2

Figure 12.3  Desirable view of the public policy space

280  Possible Solutions benefits. Governments exist to enable private actors, individuals and entities, engage in all forms of legitimate activities. Public policy is most often enacted to enable or facilitate legally mandated activities that create private wealth. In this view, public and private benefits go together, and there is no tension between public policy and private benefits. In fact, the promotion of legitimate private benefit becomes an explicit objective of public policy. Accordingly, one may provide for deemed approval if an application is still pending beyond a certain time. Or, more explicitly, public policy may seek to catalyze specific markets or economic activities, or deregulate certain activities or ease documentation requirements on the premise that the net economic costs far outweigh the public finance benefits. Or one may just promote public policy options that offer the widest spectrum of private beneficiaries. However, in a society where the private sector is often stigmatized for crony capitalism and cutting corners, and where rent-​seeking involving public officials and private interests is not uncommon, getting even the honest government officials to internalize this worldview is a massive challenge. The legal change brought about by the government in 2018 is no doubt a huge positive step in that direction.

12.3  Logic Is Sometimes an Enemy of Prudence For every complex problem, there is an answer that is logical, simple, and wrong! —​slightly rephrasing H L Mencken

Logical solutions provide a level of psychological comfort for officials that can lead them to gloss over the practical problems likely to arise from the adoption of such solutions. Logic can blind one with an illusion of success. The success of some important IT initiatives has led to a belief among many officers that e-​governance can be a panacea for persistent development challenges, but this is an example of logic overwhelming prudence. It is now commonplace to have suggested IT solutions to problems everywhere—​from schooling to transportation.

Becoming Better PolicyMakers—Behavioural change  281 In many cases, while the information is necessary to get us to the starting line, it is hardly sufficient to realize the desired outcomes. The latter is contingent on addressing hard constraints and complementing information with smart and persistent monitoring. In fact, paradoxically, it is likely that the psychological comfort associated with the availability of information at fingertips provides an illusion of control that prevents us from exploring the more serious implementation challenges. This is not to suggest that logical solutions are always bad or that new systems must not be attempted. Rather, the prudent approach is that, when faced with radically different and logically neat solutions, it is useful to • consult widely with actual field functionaries (even if they are suspected to be deliberately resisting change); and • to test, as far as possible, the logical solution in field conditions through pilots or phased implementation with close monitoring and correction.

12.4  The World Is Second Best, at Best3 Best-​practice institutions are, almost by definition, non-​contextual and do not take account these complications. Insofar as they narrow rather than expand the menu of institutional choices available to reformers, they serve the cause badly . . . real-​world reformers operate in a second-​best environment of their own, which means they need to keep an eye on how proposed solutions affect multiple distortions. —​Dani Rodrik4

Logically consistent and idealistic models of public policy are mostly unlikely to work in the real world. They need to be subjected to the filter of real-​world constraints—​social norms, political acceptability, financial 3 Avinash Dixit, Governance Institutions and Economic Activity, Presidential Address to the American Economic Association, 4 January 2009 (https://​www.prince​ton.edu/​~dixi​tak/​home/​ PresAd​_​F1.pdf) (Accessed on 18 December 2018). 4 Dani Rodrik, Second-​Best Institutions, NBER Working Paper No 14050, June 2008 (https://​ www.nber.org/​pap​ers/​w14​050.pdf) (Accessed on 18 December 2018).

282  Possible Solutions and physical resources, and administrative feasibility. More often than not, each one of these constraints poses insurmountable barriers to the implementation of an ideal model of the policy initiative. Similarly, best practice models of public policy are rarely ever successfully replicated. What works here is very unlikely to work with a similar effect elsewhere. Context matters greatly in public policy, and no two places are identical. Even when the idea underlying the best practice is replicable, its implementation design will need context-​specific adaptation. Such adaptation will most often involve certain compromises on the ideal implementation design to accommodate contextual constraints. Best practice examples of public policy interventions are generally either deeply rooted and dependent on the specific local context or the outcome of an exceptional individual initiative by a champion leader. In the case of the latter, the initiative is unlikely to survive upon the transfer or exit of the champion. Such models are therefore inherently not readily replicable. But the clamour to replicate successful solutions is commonplace. Sample this: The District Collector of X initiated a program where all the public facilities in a Gram Panchayat were brought under a single campus. This enhanced administrative efficiency and oversight, reduced absence, increased safety for women, and improved overall outcomes over the past two years. We should strive to replicate this model elsewhere.

Or this: We are developing new building bye-​laws like in the US, which define standards, on par with the best in the world, on everything related to the building construction.

Technology initiatives are often used as best practice models for emulation. Geographic Information System (GIS)-​based mapping of property appears an elegant and logically sound strategy to improve property tax assessments and collections. The attraction arises from the popular narrative of large numbers of unassessed and defaulting properties and the absence of granular data on such properties. The technological solution

Becoming Better PolicyMakers—Behavioural change  283 appears compelling and neat—​geo-​reference all properties, identify non-​ payers, assess, and collect. In the real world, especially in the larger cities, it is not so much the unassessed as the under-​assessed and uncollected, and the low tax rate itself, that are the challenges. The completely unassessed are more likely to be in slums and squatter settlements where assessment raises practical difficulties. Technology offers less help in these cases. Formidable state capacity constraints hinder progress in these areas. None of this is to deny the undoubted potential benefits of such technology solutions. For example, GIS tagging is a promising and useful technology. But those benefits can be realized only if they are complemented with other measures. As another example, a biometric attendance system cannot be fully effective without diligent follow-​up action based on its information and its integration into a functional performance management system. There are thus two lessons. One, best-​practice technology (and other) interventions are not substitutes for addressing fundamental state capability weaknesses. Two, public policy problems being intimately influenced by context, best-​practice solutions, most likely tailored around their specific context, are unlikely to be amenable for simple replication.

12.5  Comfort With Holding Multiple Hypotheses About an Issue The test of a first rate intelligence is the ability to retain two opposing ideas in the mind at the same time and still retain the ability to function. —​F Scott Fitzgerald We support a higher minimum wage since we feel it would improve overall incomes, trigger increased consumption, and boost private investment, thereby creating more jobs. But it is possible that the minimum wage would stifle investment and even cause job losses by reducing labour demand. The evidence we have today supports the former view. But we shall be open to feedback and if facts available change, we’ll change our views.

284  Possible Solutions Public policy initiatives are generally complex by their very nature. It is commonplace to find more than one causal mechanism or outcome, often contradictory, for each initiative. Similarly, implementation outcomes vary widely based on the context. For example, as with the minimum wage, cash transfers can lead to potentially divergent outcomes. One theory of change postulates that individuals respond rationally to incentives and know what is in their best interests. Accordingly, cash transfers constitute an incentive-​compatible and efficient means to achieve welfare objectives. But a contrasting theory of change would argue that individuals behave irrationally and are likely to fritter away such transfers. There is no way of reliably assessing ex ante which of the two opposing theories of change is likely to be realized in any particular context. Policymakers need to be cognizant of this and be able to mentally juggle such contradictions. They should cultivate the ability to analyse dispassionately two contradictory propositions or likely theories of change about the same issue and then exercise judgement in favour of one without ever becoming a captive of that proposition and losing sight of the other. Further, they should constantly keep an eye out for tell-​tale ‘signatures’ of those contradicting or alternative theories of change. This assumes significance in light of the uncertain dynamics of any initiative, wherein unexpected or less than desired outcomes may materialize instead of the originally desired outcomes. Unfortunately, in such instances, the strong mental anchoring around a specific program design and certain outcomes often become cognitive barriers for policymakers in revising that policy or program design. For example, once a program disappoints, instead of also revisiting the first principles of the program design, they tend to restrict their search for explanations to execution failures.

12.6  Comfort With Ambiguity We feel that there should be a set of clear and prescriptive program components to improve learning outcomes that are implemented nationwide. We cannot abdicate our responsibilities and leave program components and their design to the discretion

Becoming Better PolicyMakers—Behavioural change  285 of the district or school. How can we implement such an open-​ ended policy across the country?

The conventional wisdom about policymaking is that there are facts and propositions involving those facts, and based on them policy can be made with certitude. So we talk unambiguously about the right and wrong policies. But this narrative has very little resonance with the real world of public policy. In that real world, there are very few facts, much fewer propositions drawn from them, which can provide decision-​support on important decisions with a high degree of certainty. Instead, administrators must learn to be comfortable in dealing with ambiguity and decision-​making under uncertainty. They need to accept the reality that the vast majority of public policy challenges do not have a set of readymade prescriptions that can be simply taken off the shelf and implemented. Most often solutions are sui generis and emerge from context-​specific local design, although based on some universal principles of program implementation. They are not amenable to the reassurance of prescriptive implementation guidelines and neat end-​to-​end monitoring. In fact, such a straitjacket strategy may be counter-​productive and result in many failures, wastage, and even egregious corruption. The policymaker, while having mechanisms in place to mitigate such implementation risks, should be able to accept such turbulence. Many development challenges have no neat, tangible, and comforting solutions. Lant Pritchett makes the distinction between thin and thick activities.5 The former are inputs and logistics focussed and amenable to easy monitoring based on quantifiable parameters. Common examples include infrastructure projects, increasing school enrolment, recruiting doctors, and financial inclusion. In contrast, the latter are quality-​focussed initiatives where the outcomes are critically dependent on the nature of cutting-​edge human engagement. Learning outcomes, effective health treatment, bank account utilization, and all types of behavioural changes are not amenable to routine one-​size-​fits-​all execution plans and top-​down monitoring. 5 Lant Pritchett, The Risks to Education Systems From Design Mismatch and Global Isomorphism, CID Working Paper No 277, February 2014 (https://​proje​cts.iq.harv​ard.edu/​ files/​bsc/​files/​277_​pr​itch​ett.pdf) (Accessed on 18 December 2018).

286  Possible Solutions Policy formulation and implementation on thick activities (and even certain thin activities) generally require multi-​ pronged approaches where not only are some of the prongs diffuse and indirect in their effect but also there is considerable uncertainty about their outcomes. They, therefore, demand an acceptance of the reality of uncertainty in outcomes and a willingness to accommodate multiple approaches. They defy the ‘emerge-​from-​a-​presentation-​with-​assured-​solutions’ nature of conventional problem-​solving. This requires an appetite for ambiguity.

Becoming Better PolicyMakers—Behavioural change  287

12.7  Detachment From ‘ Achievements’ When I was . . ., I did . . . So the policy on . . . should be . . .

It is a common feature that people are strongly attached to their self-​perceived ‘achievements’. It means that their comprehension of the particular public policy issue is very deeply influenced, or clouded, by their prior experience. Behavioural psychologists point to the certain biases that afflict the human mind. The availability bias anchors the human mind to evaluate a topic based on the examples available and entrenched in our memory. Similarly, over-​estimation bias makes us repose far higher subjective confidence on our own judgements than their objective accuracy would suggest. Policymakers should be able to analyse a problem on its merits without being excessively encumbered cognitively by the specific experiences from their earlier postings. Personal achievements are, almost by definition, the exception and not the norm. One only remembers a handful of ‘successes’ from among the several initiatives that one would have tried during a career. It is also a fact that one forgets one’s failures very fast! Almost always, the contribution of the individual “me” to its failure gets glossed over. The median bureaucrat and business-​as-​usual bureaucratic system may not be able to get the initiative past the post. Further, personal achievements invariably suffer from over-​estimation bias, whereby the individual becomes blind or at least overlooks the generally more nuanced actual implementation outcomes. The ‘successes’ get amplified even as the ‘failings’ are diminished. The more effective approach may be to develop policy based on objective principles and then filter it based on anecdotes of achievements. A reversal of the order can, more often than not, be counter-​productive.

288  Possible Solutions

12.8  Not Generalizing From a Few Experiences or Data Points .

We did cash transfers in PDS in our district. It was a remarkable success. We should therefore generalise and immediately scale up across the state.

Human beings suffer from representative and anchoring biases, which make us crystallize our views on an issue from a few personal experiences and the limited examples that immediately come to our mind while discussing that issue. Such biases prevent us from realizing that our ‘prior experience’ is but just one data point and the particular outcome deeply influenced by its specific context. In the process, we gloss over the contextual and other factors that may have contributed to the specific outcomes in our very particular experiences. For example, our personal experiences are circumscribed in place, time, supervisory effort, and context. Given this scale of variance, it may be presumptuous to generalize from our experiences. Consider the (hypothetical) example of Utopia Electricity Distribution Company. Its new Chairman introduced performance pay, with sub-​stations as cost-​centres, to incentivize officials to do rigorous energy audits and lower their distribution losses. This became a major success and Utopia lowered its distribution losses to the lowest in the country. After two years, he got transferred and posted as Director of the State’s Education Department. Glowing from his success with performance pay, he decided to introduce the same for teachers, despite stiff opposition from teacher unions. The tenth class pass percentage became the preferred metric. In two years, the initiative degenerated into a massive failure as teaching for the exam and copying became the norm. Clearly, the two contexts are vastly different. In the former, it was possible to collect and disseminate information that was a credible measure of the objective sought to be realized. But, in the latter, in the absence of similarly credible metrics, the teachers responded by gaming the data collected and transferred. In fact, it may be very difficult to even translate the same success with performance pay in one Distribution Company to another.

Becoming Better PolicyMakers—Behavioural change  289 Personal experience and prior data points are valuable and can make the difference between the right decision and the wrong decision with the same information. But, as in the case of personal achievements above, a simple thumb rule may be to abjure from making individual personal experiences the basis of broad policy inferences. Instead such experiences should only be used to validate policy prescriptions.

12.9  Top–​Down vs. Bottom–​Up Perspectives on State Capability We can effectively implement the ICDS activities in a district by having the GIS co-​ordinates of all the Anganwadis, SMS-​based attendance monitoring, random inspections by the supervisors, periodic training, and regular monthly reports. This is a best-​ practice and shows the way ahead!

The conventional wisdom on state capability views it as a supervision and monitoring challenge—​how do we design systems and rules that enable more effective monitoring and supervision? An alternative approach to building state capability is the creation of an ecosystem that facilitates and encourages effective implementation at the cutting edge—​what can be done to create an environment where the various stakeholders are incentivized or encouraged or nudged into effectively discharging their respective responsibilities? There is a fundamental difference between the two approaches. The former is about superimposing a top-​down monitoring apparatus on the existing system, oblivious to the capacity of the system to deliver on the desired outcomes or the practical realities of economic and social contexts. In contrast, the latter is a bottom-​up organizational re-​design that seeks to create a network of incentive-​compatible accountability relationships that would increase the systemic capacity to achieve the desired goals. The natural inclination of policymakers is to design policies with the former approach. The longer time generally required with the latter approach naturally discourages administrators with their limited tenures. Further, the former engenders an illusion of control, which makes it more

290  Possible Solutions amenable to accommodate bureaucratic guidelines and provides safeguards against ex post facto audit and vigilance enquiries. (Admittedly, this illusion of control may even be insisted upon by approval authorities as well as audit and vigilance agencies.) Finally, the latter involves the application of several levers of influence, most of which require considerable local discretion and cannot therefore be reduced to simple prescriptions, which can then be issued as guidelines. It does not require much insight to appreciate that too many regulations and excessive enforcement are a recipe for more harm than good. For one, the reality of weak state capacity means that regulations are more likely to remain on paper and open up the likelihood of selective enforcement. Rent-​seeking invariably follows.

12.10  Difference Between Partial and General Equilibrium The State’s experiment with standardized tests in secondary schools did not succeed as expected. In the first couple of years, the test results were credible and it created competition. Subsequently, the teachers started gaming the tests, thereby significantly eroding its credibility. It highlights the perils of excessive reliance on testing and the challenge with the implementation of any performance management solution.

Most public policy interventions have an immediate result that is often very different, even contrasting, from its long-​term impact. This applies to both pilot projects as well as the initial stages of a project, the outcomes of which are often at wide variance with those from its expansion in scale and evolution over time. Our cognitive bias or reluctance to engage more deeply with the problem blinds us to mistake the intermediate output for the final outcome. It is also the case that the short tenures of officials encourage them to pay greater attention to the immediate benefits and overlook longer-​ term costs. Many subsidies have large-​ scale unintended macro-​ economic effects. Consider the example of subsidized farm power. When originally

Becoming Better PolicyMakers—Behavioural change  291 conceived, it was thought that it would enable farmers, especially in areas without assured irrigation sources, to draw groundwater to irrigate their fields and increase their productivity. But over time, political dynamics took charge, and the subsidy steadily increased till farm power became free. Free power encouraged skewed cropping patterns and over-​exploitation and the depletion of groundwater, all of which further increased power consumption, and so on. Much the same applies to performance payments. If implemented without credible performance measurement metrics and poor implementation mechanisms, as is most likely in many cases, there is a strong likelihood that the performance system will not only be gamed but also slip into becoming an entitlement system. We are likely to end up with a business-​as-​usual system with higher expenditure on the establishment. In fact, the dynamics of scaled-​up implementation creates effects that often take away from the successes of pilot implementation. Primarily, unlike the focussed and closely monitored implementation of small pilots, any scale-​up runs into the state capacity and other standard constraints of implementation. Invariably the results are far less encouraging than that of the pilot. Another example comes from a recent study on the introduction in some American states of ‘ban the box regulations’. These regulations prohibit initial job applications from asking candidates to tick a box indicating whether they have a criminal record. The assumption is that candidates can better explain their personal history in an interview. But this, in turn, created perverse incentives. Preventing employers from screening out candidates with criminal histories at the start of application process did not change their underlying desire to avoid hiring such candidates. Denied the screening, they started using race as a proxy for criminal history and increased discrimination against all low-​skilled African-​Americans.6 This case is representative of the difficulty in anticipating all possible emergent outcomes when designing policy. Given the central role played 6 Sendhil Mullainathan, ‘Ban the Box? An Effort to Stop Discrimination May Actually Increase It’, The New York Times, 19 August 2016 (http://​www.nyti​mes.com/​2016/​08/​21/​ups​hot/​ ban-​the-​box-​an-​eff​ort-​to-​stop-​dis​crim​inat​ion-​may-​actua​lly-​incre​ase-​it.html) (Accessed on 18 December 2018).

292  Possible Solutions by human agents in such transactions, and the numerous possibilities associated with human response to incentives, it is only natural that there be similar surprises most often.

12.11  Beware of the Law of Unintended Consequences Fervently supported by Baptists, early last century, some American states promulgated laws to ban Sunday sales of alcoholic beverages at legal outlets. Surprisingly, the biggest supporters of this legislation were the Bootleggers, who benefitted from the reduced availability and the opportunities to black market. In simple terms, the Baptists and their support for prohibition keep the Bootleggers in business! —​Bruce Yandle7

Consider these examples. Treated bed-​nets which are distributed to protect people from malaria end up being used as fishing nets. Toilets come to be used as storerooms. Bins distributed to encourage people to segregate garbage get used as buckets to store water. Free farm power leads to increased groundwater usage and depletion of water tables. Road widenings merely provide temporary relief and exacerbate the situation by drawing more vehicles into the road. Well-​intended urban regeneration projects end up gentrifying those areas and displacing the poor away from their livelihoods. The larger the policy objective, the greater is the likelihood of unintended consequences. So macroeconomic policies—​ new financial market regulations, monetary policy decisions, trade liberalization, environmental deregulation, and so on—​invariably generate unintended consequences. If not adequately addressed, some of the unintended consequences end up dwarfing the originally intended consequences and the policy itself backfires. 7 Bruce Yandle, ‘Bootleggers and Baptists—​The Education of a Regulatory Economist’, AEI Journal on Government and Society, May/​June 1983 (http://​pir​ate.shu.edu/​~rotth​oku/​Libe​rty/​ On%20Boot​legg​ers%20&%20B​apti​sts.pdf) (Accessed on 18 December 2018).

Becoming Better PolicyMakers—Behavioural change  293 This should not come as a surprise: most policies benefit some at the cost of others and, to that extent, have distributive consequences. The winners and losers adapt to the change in the status quo to maximize their gains and minimize their losses. In addition, those who see opportunities in the new regime are attracted. How all these dynamics play themselves out is difficult to predict with any degree of certainty in the case of large systems. Even the most well-​intentioned regulations generally inflict significant costs on legitimate business transactions. For example, instead of lowering the cost of formality, regulations aimed at addressing the problem of informality may end up actually increasing the cost of economic transactions. Consider the example of real estate market regulation. In its ideal form, it would have many consumer protection standards, including complete transparency in transactions, title protection, and escrowing of the amounts collected from buyers. While undoubtedly laudable, they are also likely to increase the cost of development for developers, who are most certain to pass on the costs to the buyers, thereby forcing up prices. Taken to its extremes, its impact on the objective of making housing affordable may be less than benign. Hence the need for balance in designing regulations. Or consider the case of progressive labour market regulations like mandatory employer-​funded labour protections. Take the example of domestic and construction workers, two categories who suffer from extremely poor working conditions. It is the natural response of well-​intentioned people to demand regulation of working conditions and higher standards. Most often, if not always, they advocate state-​of-​art labour protections for these workers. But in extremely price-​sensitive markets where the margins are very small (the numerous layers of sub-​contracting in construction reduces margins at all levels), the cumulative cost of these protections can be prohibitive for the employers. In such circumstances, the result of higher standards is likely to be regressive—​workers forced out of the market by mechanization, more informality, more harassment, and more corruption. As both the examples of real estate and labour market regulations show, informality and its ills cannot be regulated away. Since most policies end up having unintended consequences, as with the previous issue, concurrent evaluation of policy implementation

294  Possible Solutions (whether so called or not) and course corrections thereon are as important as the policy’s design.

12.12  Policymaking Is Often About Iterating With a Draft Solution We want to implement the odd-​even road-​rationing experiment in X city. But we are conscious of the contextual problems. We also want to avoid too many disruptions. So we are adopting an iterative approach, where we propose to experiment with an implementation design, then constantly refine it based on emergent evidence. We hope to get the program design right in due course.

Horst Rittel and Melvin Webber described many social policy challenges as ‘wicked’ problems.8 They argued that such problems have no ‘solutions’ in the sense of definitive and objective answers, much less ‘optimal solutions’. Given the complex nature of reforms in most areas, with several uncertain elements with the potential to stifle economic activity, a more appropriate strategy may be to embrace an approach of iteration and adaptation in response to emergent problems to the original implementation plans. Most persistent development challenges require using multiple interventions (or program components) and with varying intensity depending on the context. Consider the example of a program to improve learning outcomes in schools or lower traffic congestion or make housing affordable. Learning outcomes require interventions that enhance the quality of classroom transactions, empower and equip teachers with requisite capacity, and ensure effective monitoring of the interventions and assessment of learning outcomes. Apart from infrastructure, reducing traffic congestion would require interventions that encourage the use of public 8 Horst W.J. Rittel and Melvin M. Webber, ‘Dilemmas in a General Theory of Planning’, Policy Sciences 4, 1973 (https://​link.sprin​ger.com/​arti​cle/​10.1007/​BF0​1405​730) (Accessed on 18 December 2018).

Becoming Better PolicyMakers—Behavioural change  295 transit, lower commute requirements, and restrict private vehicles usage. Affordable housing demand interventions to both increase supply and support demand. It is impossible to just brainstorm and design an effective program with all the component interventions and their respective intensity and prioritization. For a start, program features may need to vary across contexts. Further, implementation can surface unanticipated problems. In fact, these problems and the attendant public discontent may often provide precisely the right opportunity to mobilize the consensus required to take decisions that would otherwise have been difficult. As we know, there is no better opportunity to resettle riverbed encroachers than when there is flood. In these cases, foisting a tightly scripted program is unlikely to be effective. The program needs to evolve during implementation. Some things can make this evolution efficient. One, the first version of the program should be robust enough (but not the last word) and have sufficient flexibility for adaptation and improvement. Two, there should be tight feedback loops to gather and process information and sufficient capacity to incorporate and revise the original program design. In management-​speak, the first version of the program or regulation (in case of a new law) is equivalent to a minimum viable product (MVP), the best possible version that could have been formulated given all the multiple stakeholder concerns. But once the MVP rolls out, the problems of implementation invariably emerge. As the saying goes, the best response to such situations may be to ‘keep the powder dry’! The basic premise here is that the success or otherwise of the roll-​ out of most complex development projects and public policy initiatives (programs or regulations) is not such much the original product, but the ability of the system to respond swiftly and effectively to emergent challenges. This is as true of legislation as it is to a newly commissioned airport terminal or metro railway project. There is no substitute for field-​ testing and responding swiftly to emergent problems. This insight applies as much to program implementation as to program and policy design. Uncertain elements are always likely when we start out on the journey. Figuring them out is possible only during the journey.

296  Possible Solutions

12.13  Negotiating Policy Choices When faced with a public policy challenge, the academic researcher uses theoretical frameworks to figure out the ideal solution. The bureaucrat in turn jumps straight into the solution, which is most immediately apparent and politically expedient. The former suffers from the solution being impractical, and the latter from that of being inefficient. So how does one get to a solution which is both practical and efficient given the objectives of the political executive? One approach would be to figure out how to move from the ideal to the practical and efficient. Assuming a solution spectrum, how far from the ideal solution should we travel to reach the practical and efficient solution? What are the costs of moving away from the ideal solution? The objective should be to arrive at a solution that minimizes costs and is politically acceptable. The process of arriving at that solution involves the bureaucrat negotiating and bargaining with their political masters. So, for example, to address the problem of farm distress, the political executive may prefer a subsidy. Let’s assume that the ideal solution may be the elimination of price controls. But political economy dictates the ideal solution is off the table. The challenge then is to design the least costly subsidy. It may be a gradually declining cash transfer with a sunset provision. Or something else. Or in other cases, the political choice may be for a health insurance, when the value for money choice should have been to strengthen the public health care system. The challenge then is to design an insurance system that also goes some way towards strengthening the public health system and negotiating to arrive at a politically acceptable compromise. A good civil servant should understand the ideal solution and the objectives of the political executive. He or she should highlight the costs, both short-​and long-​run, associated with deviating too much from the ideal solution. The endeavour should be to negotiate with the political executive to choose the least costly solution (Figure 12.4). In a democratic polity, the acceptable solution to any public issue is a matter of political choice. Such choices invariably impose their efficiency costs. Good policymaking is about minimizing the welfare losses due to political economy constraints. Negotiations assume great significance

Becoming Better PolicyMakers—Behavioural change  297 Negotiating to the practical and efficient solution

Efficiency

Ideal Solution Practical and efficient

Efficiency cost–to minimize this

Spectrum of negotiations Expedient and easy

Practical

Figure 12.4  Negotiating to the practical and efficient solution

here. To this extent, good policymaking in a democracy is as much about negotiating skills as it is about professional competence.

12.14  Policies Generally Generate Probabilistic and Not Binary Outcomes Many issues in development—​ reducing littering, increasing learning outcomes, lowering traffic congestion, making housing affordable etc –​are a long journey. There is nothing absolute about judging it. In fact, the standards of assessment of even the degree of progress made varies over time and with the initial conditions.

There are rarely, if any, outright successes among social policy initiatives. Like the proverbial hourglass, the success or otherwise of policies and programs can be judged only in degrees and not in any unqualified manner. For a start, most development challenges are unlikely to ever disappear. Poverty (at least in relative terms) is unlikely to be ever eliminated (as happened with smallpox) but only alleviated and lowered (as with influenza). Public policy can only minimize the degree and extent of the

298  Possible Solutions problem. Success, therefore, is a matter of degrees and not the realization of some absolute destination. Further, most often, the success of the same policy varies widely geographically or among population or institution categories. In fact, it may be impossible to have a perfect program design that is able to accommodate all these multi-​dimensional variations. The resolution of many complex development challenges is inherently likely to display wide heterogeneity. Furthermore, most development problems are complex and specific policy initiatives are only an intermediate step in a long journey. So, for example, financial inclusion is about enabling access to a formal financial institution, offering appropriate credit and savings instruments, helping overcome the behavioural biases that prevent utilization, and providing sufficient financial literacy to ensure prudent use. Each of the four prongs is critical to the realization of financial inclusion and they are part of a long-​drawn collective effort. This also means that any initiative is, almost by definition, vulnerable to criticism. The critics are quick to seize the opportunity, notwithstanding the impossibility of reaching the destination in finite time. Instead of reflex opposition to such criticism, however ill-​informed and unfair, it may be prudent to use it as valuable feedback and improve the implementation design of the initiative. In fact, announcements like child labour free or litter free should be seen as provocations to elicit valuable feedback, which can be used to further improve the program outcomes. The increasing use of challenge competitions to test the validity of such claims may also help internalize the probabilistic nature of outcomes associated with these initiatives. In simple terms, initiatives should be judged in probabilistic terms—​ the degree of success or failure. To what extent has the initiative succeeded or managed to stave off failure? Then one must work towards constant improvements.

12.15  Willingness to Make Bayesian Updations Situations change constantly, rendering earlier assumptions meaningless. Twenty years back the challenge in school education was to get inputs in place—​build infrastructure, recruit

Becoming Better PolicyMakers—Behavioural change  299 teachers, ensure student enrolment and attendance etc. Today, while these are in place, student learning still remains elusive. School education policy has to pivot to reflect the changed realities.

Thomas Bayes, an 18th-​century British mathematician and theologian, made a seminal contribution to the field of logical reasoning. He formulated statistical rules based on the probability of event occurrences that allow us to combine prior beliefs with new evidence and improve the quality of the original hypothesis. In simple terms, Bayesian updation enables us to constantly refine our worldview or theory of change based on new evidence. The basic concept underlying Bayesian updation carries large, but deeply under-​ appreciated, significance in public management. Policymakers, for whatever reasons, generally make judgements that are based on certain theoretical principles or data that are made available to them or advice given by experts. Accordingly, policies are formulated and programs designed. But the uncertainties posed by the implementation context (in terms of time, place, people, and resources involved and other environmental constraints), commonplace in public policy, are not typically given adequate consideration or are under-​estimated in the decision-​making process. This generally manifests as program or policy flaws that contribute to their failure. Since it is almost impossible to anticipate all contingencies and design policy accordingly, the best that can be done is to be receptive to the likelihood of failings and provide adequate flexibility to accommodate those emergent concerns. Policymakers should pay close attention to detect and exhibit willingness to update their ‘priors’ (i.e. prior assumptions), if needed, based on emerging evidence. Consider the example of India’s experience with public–​private partnerships (PPPs) in the roads sector. At the turn of the millennium, the country embraced what was then claimed as the best incentive-​compatible contract structure of Build Operate Transfer (BOT) Toll for its national highways. In such contracts, the concessionaire put in all the investment and recovered it through toll revenues. Soon, the risks from failure of traffic realization and construction delays surfaced, raising questions about the commercial viability of the BOT Toll model. This

300  Possible Solutions Bayesian updation led to a preference for the BOT Annuity model, where the government took the commercial risk and made fixed annuity payments at pre-​defined periodicity. But this still left construction risk with the concessionaire. In a country where right-​of-​way acquisition is fraught with all sorts of uncertainties, it was a matter of time before cost over-​runs battered developers. Another Bayesian updation led to the conclusion that it was prudent to mitigate at least some part of the construction risk. The result was a hybrid annuity model (HAM), where up to 40% of the construction cost was paid upfront. In conclusion, decisions under uncertainty should be made based on available data, which is filtered through the lens of knowledge about the context, and be constantly updated based on any new evidence.

12.16  Follow Rational Scepticism Water conservation and management measures have been my main focus. The entire district administration was galvanized around the program. Our hydrologists say that in the last two years the water table has risen by 5 m. But I’m not very sure. We need to have truly independent assessment of the outcome.

There are rarely fool-​ proof solutions for non-​ technical problems. Policymakers should refrain from peddling them and also take such claims with a bucket of salt. In general, it may be rational to view performance data in any field with care and have mechanisms to validate them through independent agencies. It has become commonplace to hire consultants to evaluate a policy or intervention. Unfortunately, such exercises often end up being favourable documentation rather than a rigorous assessment. In today’s world of social media-​based information dissemination, exaggerated positive claims and deliberate exaggeration of negatives are common in both conventional and social media. The retweet or likes of these positive and negative exaggerations becomes a negative social externality. A cautious attitude to such claims—​both the positive and the negative—​can therefore be socially valuable. More fundamentally, as

Becoming Better PolicyMakers—Behavioural change  301 responsible public officials, civil servants should both refrain from making as well as promoting positive claims which are not grounded on rigorous enough evidence and contradict non-​rigorous negative claims. There is a Tamil proverb that says: ‘What you see with your own eyes is false. What you hear with your own ears is false. Only that which you have thoroughly verified is true.’ This is good advice for civil servants when dealing with media-​generated fads.

12.17  Tolerate Failures and Allow ‘a Million Marginal Revolutions’ Action fuelled by doubt allows for failures to be left behind . . . . Instead of asking: what benefits [has] this project yielded, it would almost be more pertinent to ask: how many conflicts has it brought in its wake? How many crises has it occasioned and passed through? And these conflicts and crises should appear both on the benefit and the cost side, or sometimes on one—​ sometimes on the other, depending on the outcome (which cannot be known with precision for a long time, if ever).”—​ Albert Hirschman9 “Cross the river by feeling the stones.”—​Chinese proverb

One of the unfortunate unintended consequences of excessive transparency and media scrutiny of policy implementation has been to shrink the space available for policy experimentation. Faced with the possibility of being criticized for implementation shortfalls, always likely given the complex nature of most development challenges, governments have become naturally excessively risk averse. Policy gets designed to avoid certain headline failures, even at the risk of compromising on more important underlying factors. Perverse incentives are easily engendered. Given the lower threshold for failures, officials are loath to experiment for fear of accompanying retribution. Worse still, they become more bureaucratic by refraining from 9 Malcolm Gladwell, ‘The Gift of Doubt’, New Yorker, 24 June 2013 (https://​www.newyor​ker. com/​magaz​ine/​2013/​06/​24/​the-​gift-​of-​doubt) (Accessed on 18 December 2018).

302  Possible Solutions delegation, preferring one-​size-​fits-​all policy design, and adding procedural layers to minimize discretion and maximize top-​down control. After all no one ever lost his job for following the norm. Also, officials are incentivized to engineer numbers and outcomes to present a more favourable picture. Given the likelihood of right to information (RTI) requests, officials are now less willing than before to allow independent evaluations, lest it shows up with less than comfortable findings which in the past would have been used for internal correction but will now result in external flagellation. In contrast, if one were to zero down on the distinguishing features of Chinese economic growth over the past three decades, the importance of experimentation and tolerance of failures (admittedly in a country without the RTI or writ petitions or freedom of the press) cannot be over-​stated. In a nutshell, be it de-​collectivization, township and village enterprises, special economic zones, price reform, or share trading, the government allowed low-​profile experimentation, tolerated failures, and came forward to bless successes once they were demonstrated and had them scaled up. In their excellent exploration of Chinese capitalism, Ronald Coase and Ning Wang have this to say10: When a novel practice was first experimented with, particularly in a regulation-​free environment, it was bound to go awry, giving its political opponents sufficient justification to close it down. To allow this strategy to work, the government had to be tolerant, open-​minded, and learn to act quickly—​either setting up a regulatory framework to allow the practice to develop or shutting it down when the experiment turned sour. The policymakers would be guilty of halting promising practices too soon, as well as neglecting to end bad practices . . . For those who wanted to push reform forward by introducing new and viable practices, the most effective method was to experiment as much as possible to find something workable within a limited time window, while recognizing the risks that more mistakes were bound to occur when more

10 Ronald Coase and Ning Wang, How China Became Capitalist?, Palgrave MacMillan, London, 2012.

Becoming Better PolicyMakers—Behavioural change  303 experiments were run . . . this was logic behind Deng’s aphorism, “Don’t argue; try bold experiments and blaze new trails”.

If there is one major policy takeaway from China for a country like India, it is perhaps this strategy of ‘crossing the river by feeling the stones’, and that too in a low-​profile manner, one which encourages policy experimentation in a low-​stakes environment.

12.18  Avoid the Predecessor–​Successor Syndrome The e-​governance program to redress grievances started by Mr ABC in X district began to flounder immediately after his transfer due to lack of adequate attention by his successor.

Max Weber’s classic formulation of a bureaucracy presented it as a system of impersonal management. Accordingly, a bureaucracy is an institutionalized system of administration, indifferent to those administering it. However, the conventional wisdom has it that far from being impersonal, India’s bureaucratic system is deeply personalized. It is widely accepted that a large share of the blame for the failure of initiatives can be laid on the absence of their institutionalization—​successors end up neglecting, even sabotaging, their predecessors’ initiatives. The program falls through the cracks in the transition. This trend is not the exclusive preserve of public bureaucracies. All leaders try to impose their agenda on the organization. It is just that the incentives to junk an ongoing initiative and embrace a new one are more skewed in government where tenures are much shorter. In this context, institutionalization of initiatives assumes great significance. As a first-​order starting point, civil servants would do well to make the distinction between the personal and the professional while initiating new projects or adopting innovative approaches to implement programs. While personal initiative doubtless may have helped a new idea or project take off, we need to bear in mind that it has to be anchored around the traditional norms of bureaucracy. The program has to bear the imprimatur of impersonality and be governed by the rules of the bureaucratic game.

304  Possible Solutions An important pathway to institutionalization of local initiatives is to have the State Government Department or Central Ministry (as the case may be) as a partner in the initiative. A state-​level steering committee can create the right level of stakeholder support and institutionalization required to sustain the initiative beyond the tenure of officials.

12.19  Personal Effort Is Not Systemic Improvement In my two years as District Collector, I have constructed more than a fifty public toilets across all municipalities in the district through PPPs. I have also documented it here. All that needs to be done is to scale it up throughout the country by issuing guidelines based on this document.

Often, the success of the initiative is a result of the disproportionate personal attention of the District Collector and the impact it brings to bear on the initiative. Unfortunately, not only are most initiatives a result of personal effort, but they often come at the cost of institutionalization. The impatient officer micro-​manages the program and fails to nurture personnel capacity (except among a handful of his coterie, who are most likely to drop away when a successor arrives), thereby crowding out the development of institutional capacity. Such success is most likely to be transient and not survive the officer’s transfer. Therefore, any serious effort at sustainable reform should seek to cultivate organizational champions and institutionalize change. The touchstone for sustainable design is the question—​what elements should be in place such that the policy or program does not require any particular officer’s presence? And then work to incorporate those elements into the policy design and implementation. This assumes far greater relevance in the case of policy formulation than with program execution. In the case of the latter, it is often possible to substitute systemic improvements with raw personal effort to get stuff done. But with policy formulation, it is critical to have institutionalized processes and incentives alignment to ensure that the policy implementation is effective.

Becoming Better PolicyMakers—Behavioural change  305

12.20  ‘Do No Harm’ Principle I saw my most important role as the Chairman as Council of Economic Advisors not so much in making reform suggestions, as in killing off bad ideas. —​Greg Mankiw, Chairman of Council of Economic Advisers for U.S. President George HW Bush

In the venture capital (VC) industry, it is said that the mark of a successful VC fund manager is perhaps more about quickly identifying losing bets than picking winners. Just as losing bets bleed money, bad policies, especially given their vast jurisdiction of implementation, can do enormous damage and, therefore, if not avoided, need to be just quickly nipped off. On many occasions, given the context, reforms may not be advisable. In such contexts, the policymakers, in their anxiety to reform—​seen to be ‘doing something’—​end up promoting solutions whose long-​term effects are questionable. It is important to bear in mind that if you cannot improve or reform the situation, at the least one should abjure from worsening the situation. Bad policies can have profound and long-​lasting implications. When policies like free power were first introduced, little would we have known how endemic and systemically distortionary they would end up becoming. Not only do they cause wasteful public spending, but they also are a form of assault on incentives, distorting the incentives of a generation of citizens. Therefore, a simple orders of magnitude–​based economic cost–​benefit analysis can be a valuable decision criterion before starting any new intervention or new regulation. A bureaucrat who does the job of highlighting this and contributing to burying a bad proposal would have done just as much as one who contributes to a good initiative.

12.21  Just Improve Basic Administration Most often, all we need is to do the simple things and not search for innovations and new initiatives. Administer the ongoing

306  Possible Solutions programs well using the existing bureaucracy and things will be largely alright.

It has become a cliché, a cop-​out, to look for innovations as a panacea for any public policy problem. Process re-​engineering, outsourcing, privatization, and so on have emerged as apparently magic pills to address persistent development failures. The simple assumption is that the problem has remained intractable for a lack of innovative strategies, and with innovations it will be automatically resolved. Unfortunately, in the vast majority of public policy situations, there are no such magic pill innovations. And the real reason for failure is often bad governance surrounding its implementation. In general, public systems in developing countries like India are entrapped in a low-​level equilibrium, exacerbated by state capacity weaknesses. The immediate objective should be to move such systems from a bad or unsatisfactory state to a satisfactory or good one. Most often, such transitions can be achieved with simple improvements in governance, or even more specifically, basic good administration. However, transforming a good system into a great one goes beyond good governance and would demand innovations. But that is an agenda for a later day. In moving public systems to deliver to reasonable standards, plain and simple good administration or management, difficult though it would be, is more than adequate and cannot be substituted with innovation. But the pursuit of innovation often crowds out good administration by diverting the limited systemic resources and personal bandwidths of public leaders. Instead of searching for innovations everywhere, Collectors who set their priorities, take routine decisions promptly, empower their teams, put in place simple but robust feedback and monitoring mechanisms, conduct random field inspections, enforce basic administrative discipline fairly, provide guidance, work hard, and innovate tactically are most likely to succeed. The most effective development intervention in the world is perhaps plain, simple persistent monitoring and appropriate follow-​up action. The reader may ask why good administration figures under better policymaking rather than better implementation. The reason it is included here is that good administration is a major part of good governance; an emphasis on good administration is one of the simplest and most effective

Becoming Better PolicyMakers—Behavioural change  307 policies a government can pursue. It carries low political risks and can bring substantial benefits. If good administration is carried out across the whole range of departments and offices, instead of just a few (and assuming the policies are not intrinsically bad), it results in improved policy and program effectiveness.

12.22  Development Is Very Hard, a Little Humility Would Do No Harm! “There is no one right answer to address the problem of deficient learning outcomes or malnutrition or traffic congestion or poor sanitation in developing countries. There are no magic pill solutions to these problems. In each case, context matters. And solutions have to evolve gradually. We need to be humble enough to accept this reality.”

Consider this. The problems of getting people to abjure open defecation or to save money or getting school systems to improve learning outcomes have been a constant concern across time, places, and policymakers. In other words, numerous policymakers, including some obviously smarter than you, have in thousands of administrative units across the world, and over many decades been grappling with these problems. These are an enormous number of data points. It still elides a neat and replicable solution. So, maybe, there are no such universal and easily adoptable solutions. We need to accept this and move forward. It is commonplace for generalist administrators who have joined a new posting to be deceived by the illusion of low-​hanging fruit from logically consistent and attractive solutions that are presented to them. And there is likely to be no dearth of such simple solutions and people peddling them. It is also commonplace for them to embrace some of these solutions and make grand announcements committing to solve complex problems whose resolution have elided their predecessors for decades. This would also be despite the failure of similar attempts by some other predecessor. Almost always such intent comes wrapped in the form of a more vigorous profession of commitment and a new program, the primary

308  Possible Solutions differentiator most often being the scale of the ambition in terms of targets and time within which it is sought to be achieved. This alone, it is assumed, would ensure the success of the endeavour. A two-​part smell test is suggested below to assess whether ‘this time is any different’. 1. What is being done now that was not part of earlier efforts and how will it increase the likelihood that this time is different? (The perennial suggestion to create more ‘Vigilance Officers’ in organizations that do not have them to reduce corruption is an example—​the existence of Vigilance Officers and Chief Vigilance Officers has arguably not made any difference to the organizations that have them; some would even argue that the Central Vigilance Commission is redundant!) 2. What has been done to improve state capability in the execution of the program? (The creation of a ‘new policy’ is very popular with civil servants—​but was it the earlier policy that was poor or was it the implementation? If the latter, what steps are you taking to make sure the new policy is implemented?) Very often, the suggested solution will not pass the smell test. Development is really hard!

12.23  Making the Behavioural Changes Happen Making officers follow these principles is not something that can be done by fiat. How can officers be motivated to adopt them? Fortunately, there are two important features of IAS and the nature of their work that can be leveraged to inspire motivation. One, even though significantly diminished, the IAS still possesses a very high degree of esprit de corps. The relatively small size of the cadres and the nature of the recruitment and training processes and service conditions are important reasons for its endurance. The associated social capital is rare among similar bureaucracies. This makes peer pressure and moral suasion effective instruments to change behaviours. Second, the IAS possesses a convening power that is perhaps unmatched for any comparable bureaucracy in another country. While

Becoming Better PolicyMakers—Behavioural change  309 this exists at all levels, it is especially so at levels of the state government and lower. Many individual heads of departments and Collectors and Municipal Commissioners evoke both authority and respect, some, especially at the sub-​state level, even admiration. These two are invaluable resources to capitalize on in any effort to imbibe some of these elements and transform governance in India. The esprit de corps can help motivate and the convening power help channel it towards realization of the objectives. There is a rich body of management research around the idea of individual initiatives in realizing significant organizational changes. Advocates of the taking charge behaviour (TCB) school talk about ‘voluntary and constructive efforts by individual employees to effect organizationally functional change’.11 In fact, the very popular Japanese management approach of Kaizen, though focussed on multitudes of incremental changes, also owes to employee initiative. One study was conducted to determine the impact of the TCB approach on public service motivation (PSM) and behaviour change among the German federal police force. The study finds that TCBs are associated with transformational leadership and that PSM mediates the relationship between perceived transformational leadership and TCBs.12 Given their convening power, the freedom for taking charge and the potential for realizing change from such actions is much higher in the case of IAS officers. The principles outlined earlier can be both individually inculcated and organizationally internalized at the initiative of individual officers at all levels. It is for the self-​motivated among IAS officers to take the lead and embrace some of these principles in their administration and management. They also have a role in influencing at least some of those within their networks to adopt these principles. Peer groups and civil service associations have an important role to play in promoting motivation and ‘taking charge behaviours’. For example, the state and central level service associations could exercise moral suasion and discourage the successor–​predecessor syndrome. Or senior officers 11 Elizabeth Wolfe Morrison and Corey C Phelps, ‘Taking Charge at Work: Extra Role Efforts to Initiate Workplace Change’, Academy of Management, 30 November 2017 (https://​journ​als. aom.org/​doi/​10.5465/​257​011) (Accessed on 18 December 2018). 12 Fabian Homberg et al., ‘Public Service Motivation and Extra-​Role Performance: Taking Charge Behaviour at Police Services’, Public Management, May 2019.

310  Possible Solutions could encourage experimentation by their subordinates and tolerate associated failures. Many of these principles, and more, can be codified as a set of desirable behaviours for IAS officers and issued as part of a Code of Conduct booklet by the Central IAS Officers Association. It would be applicable just the same for any other service association. Similarly, senior officers could encourage some of these practices by calling them out and commending them in the likes of social media. The peer effects associated with such praise can be significant. Finally, there could be a category for good bureaucratic principles in awards, like the Prime Minister’s Excellence Award, that covers some of these principles. But ultimately, the responsibility for imbibing them rests with individual officers.

13 Addressing Execution Challenges—​I The preceding chapters dwelt on ways of improving policymaking and program formulation. This and the next two chapters deal with how to improve the execution of policy and programs.

13.1  Strengthening Supervision 13.1.1  Improving Traditional Methods of Supervision One of the most important determinants of successful implementation is its supervision. While there are several reasons for the inadequacies in implementation, a common underlying thread is ineffectual supervisory guidance and lack of rigorous monitoring.1 Unfortunately, in the search for solutions to implementation failure, not much attention has been paid to the critical role of supervisors. Any bureaucracy can be divided into broadly two parts—​one undertakes field-​level execution and the other supervises. For example, while teachers carry out classroom instruction, headmasters and school inspectors guide and monitor the quality of this instruction. Within the supervisory bureaucracy itself, the bottom half directly supervises field-​level execution, while the other half does higher-​level supervision. Typically, the sub-​district level officials belong to the former category, while those at the district and above are in the latter. The efficiency of field supervision is a major contributor to the successful execution of any policy and implementation plan. Weak supervision of auxiliary nurse midwives (ANMs) by health supervisors and 1 Gulzar Natarajan, ‘Supervisory Deficit in India’, Mint, 6 April 2012 (https://​www.livem​int. com/​Opin​ion/​1TF​2uI0​UfJl​GUh7​tyJM​APN/​Supe​rvis​ory-​defi​cit-​in-​India.html) (Accessed on 20 December 2018). State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0013

312  Possible Solutions doctors weakens our primary health care system. The poor quality of supervision of ration shop dealers by revenue inspectors is a major contributor to the weakness in the public distribution system. Inadequate supervision of building constructions opens the doors for the proliferation of illegal structures that dot our cities. And so on. What is responsible for this supervisory deficit? For a start, as discussed in Chapter 6 in the context of the ‘thin’ state, supervisors, at all levels, are over-​burdened with responsibilities, in both their functional and geographical spans. Their bandwidth is constrained either by having too large a jurisdiction or too many people to manage or too many functions to supervise. Often it is all three. Second, field-​ level supervisors, cutting across departments and throughout the country, function primarily as data collection agents. They generally make inspections to collect data and hold meetings to consolidate them. And all this data then flows upward, most often meaninglessly, to satisfy routine supervisory and monitoring requirements. In the scramble to send reports before periodic deadlines, they lose sight of their primary responsibility of oversight and guidance. It is rare to find field-​level supervisors spending more than a small portion of their time on the quality of implementation. Another critical area is the lack of sufficient empowerment. Supervisors have limited powers to hold errant field officials accountable. Even when they have some de jure powers (especially performance assessments), de facto they are blunt tools. Politicization exacerbates the problem. Extracting work and delivering on objectives without the requisite administrative powers is well nigh impossible. There are no easy solutions to these problems. However, there are certain fundamental principles that should underpin any effort to bridge the supervisory deficit. First and foremost, supervisors should move away from being mere data collectors and use that data to support decision-​making. Data collection should preferably be done directly through field functionaries by leveraging technology, such as tablets and mobile phones. It would also help improve the accuracy of the information collected and increase the likelihood of institutionalizing data collection systems. In this context, digital interfaces—​computers or mobile phones—​that combine analytics software and vivid data visualization techniques can

Addressing execution challenges—I  313 provide highly effective decision-​support. They distil the essence of the data collected into clear user-​directed messages that amplify supervisory efficiency. For example, school or hospital inspectors should be able to reliably identify bright and dark geographical locations, or functional areas of failure and success, and target their daily inspections accordingly. Second, they should be liberated from attending repeated and mostly meaningless meetings and be allowed to focus on their primary responsibility—​ monitoring field work. Unnecessary bottom-​ up periodical reports and low-​importance meetings and reviews need to be systematically identified and culled. Third, instead of rambling over the entire spectrum of their duties, supervisors should be guided into prioritizing work and focussing on critical areas. This assumes greater significance in light of the functional over-​load that field officials face. Strategic supervisors should guide field supervisors in finding this balance without compromising on objectives. Fourth, routine inspections and meetings should be partly replaced with substantive monitoring that pays attention to critical implementation processes and quality of interventions as measured by outcomes. Given the large jurisdictions—​geographical and functional—​inspections and meetings should be made to count with rigorous and continuous follow-​up. Finally, supervisors at different levels should coordinate responsibilities in a manner that prevents duplication and enables complementary monitoring and follow-​up. Fundamentally, supervisors should re-​focus their energies on guiding and improving the quality of implementation. Needless to say, this will require significant effort at capacity building before supervisors can function as effective public managers. Yet, when compared to many other possible improvements, it is the authors’ view that simple improvements such as those described above have far greater potential (at far lower cost) than many more sophisticated solutions.

13.2  Electronic Monitoring Systems Electronic systems are another way of improving supervision and monitoring. Advances in digital and communications technology have created

314  Possible Solutions a great interest in management information systems (MISs) and, in particular, MIS dashboards. State governments, for instance, Andhra Pradesh, have pioneered the extensive use of IT systems to improve monitoring and the effective implementation of government programs. Andhra’s Real Time Governance System (RTGS) uses comprehensive dashboards for decision-​support and outbound telephone calls to validate with feedback. Other states are exploring establishing similar systems. Institutional funders and donors too are exploring opportunities to establish elaborate Chief Minister dashboards that track indicators across Departments. Technology solutions have been implemented in recent years across all levels of the government to enhance the effectiveness of monitoring the delivery of public services. Today most departments in many states and some at central government levels claim to have dashboards. But few of these solutions are being used at anywhere near their respective optimal efficiency levels. More importantly, there is little evidence that these systems have improved the effectiveness of public services delivery. Notwithstanding the various accolades garnered, few have survived the frequent transfers of officials and become institutionalized organization-​wide. In even the more lasting implementations, critical last-​mile gaps, especially around reliability, relevance, and rendition of data, hamper their effective use.

13.2.1  Conceptual Challenges In the private sector, usually causal factors are more limited and clearly identifiable, and outputs are more easily and credibly measurable. Even if they are not, the ‘ultimate bottom line’ of the profit motive provides clear guidance in most situations: if products or services are selling in increasing volumes and profits are good, that by itself shows that value is being created. The underlying concept behind such monitoring systems—​measure something proximate to the outcome and you can control implementation—​is appealing. It comes from the belief that what has worked in the private sector should work in governments too. Thus the following assumptions underlie the push towards digital dashboards:-​

Addressing execution challenges—I  315 1. There are some things that are proximately causal to the desired outcome, 2. We know what those things are, 3. We can quantify and measure them credibly, 4. Actionable insights on them can be presented in a simple and cognitively striking manner, 5. The system has the intent and capacity to execute those insights as required, and 6. The associated monitoring can ensure effective execution. Unfortunately, most development challenges defy these assumptions. Many of the assumptions struggle to hold in the case of the activities of governments. Arguably the biggest challenge lies with attaching causal attribution to even the most obvious factors. So more and better teachers do not necessarily generate improved learning outcomes. Similarly, building more toilets and even making water available is no guarantee for toilet usage. Or putting in more litter bins and harsher fines is no guarantee against littering. Or getting real-​time information about public grievances does not mean those grievances will be resolved. Any intervention consists of several inputs or processes, a particular combination of which, itself mostly context dependent, generates the desired level of outputs and outcomes. Take the example of agricultural extension services or school education. The actions of the extension officer or the teacher are expected to respectively enhance farm productivity and student learning outcomes. But each is a function of several factors, including many which are beyond their control. For example, the realization of student learning outcomes depends on several proximate inputs—​teacher availability, teacher and student attendance, classroom instruction time, pedagogic approach, class size, physical school infrastructure, provision of textbooks, teacher training, periodic credible assessments, school leadership, etc. This is not to mention enabling requirements like teacher recruitment and their management policy and parental education and engagement. So, if the objective is the improvement of student learning outcomes, what should be monitored at the various levels? Monitoring inputs and process indicators is difficult since bandwidth is scarce, and inputs are

316  Possible Solutions several. It is therefore tempting to argue that while the frontline functionaries and their immediate supervisors need to be cognizant of all these inputs and processes, the headline monitoring at higher levels should focus largely on outputs, or at most on a couple of critical inputs. But this, in turn, poses another set of problems: it is not easy to measure learning outcomes with any credibility. This makes credible top-​down monitoring based on outputs or outcomes difficult or even impossible. After all, when faced with such uncertainty, how can you hold officials accountable enough? Any attempt to prioritize and monitor a couple of inputs runs into the problem of uncertainty about their relative contributions. Granted, education outcomes are perhaps among the hardest to measure and hence not typical. But headline performance monitoring on even something as simple as the progress of building or road construction cannot, with any degree of satisfaction, capture measures about the quality of the construction. Even something as routine and simple as attendance monitoring can get gamed in completely unexpected ways.2 Human ingenuity on these has no bounds. Further, focus on inputs, especially by way of high-​level monitoring, can generate incentive distortions. As will be seen later, this can have far-​ reaching consequences. Furthermore, weak state capacity means the ability of bureaucracies to act, even when information is provided, is seriously limited.3 So even where it becomes possible to credibly measure outputs, other systemic constraints soon show up. For example, a push to reliable monitoring of ante-​natal check-​ups, institutional deliveries, and immunizations will quickly reveal how over-​burdened the typical field nurse is. Placement-​ related outputs associated with short-​term skills training programs expose the practical problems related to labour market preparedness and the jobs market. In all these cases, hitherto unrecognized acute state capability constraints become evident, opening a new front for engagement, and so on.

2 Iqbal Dhaliwal and Reema Hanna, The Devil With the Detail: Successes and Limitations of Bureaucratic Reform in India, NBER Working Paper No 20482, September 2014 (https://​www. nber.org/​pap​ers/​w20​482) (Accessed on 20 December 2018). 3 Devesh Kapur and Aditya Dasgupta, ‘The Political Economy of Bureaucratic Overload: Evidence from Rural Development Officials in India’, American Political Science Review, 114, no. 4 (2020): 1316–​1334.

Addressing execution challenges—I  317 Even without accounting for the weakness in state capacity, there is the issue of enforcement (or forced compliance) being effective only at the margins. In any functional system, it is expected that the majority of workers discharge their responsibilities without the need for any external prop. A minority are likely to be shirkers and therefore require some form of external oversight. Appropriate monitoring can ensure compliance by the latter. But if this balance is reversed and shirking becomes the norm, then monitoring loses its effectiveness. This highlights the reality that the compliance effect of monitoring and enforcement is meaningfully effective only at the margins or in ensuring compliance when adherence is the norm.4 In simple terms, strong monitoring is more likely to be successful in deterring the exceptions, not in changing the norm. When the overwhelming majority shirks, then deviance becomes the norm, i.e. the de-​ stigmatized convention. Such behaviours have a scientific basis. Behavioural psychologists have shown how positive messages that highlight conformity to rules or social objectives elicit far greater responsiveness than negative messages, which either highlight the magnitude of the problem or warn about the consequences of deviation. Accordingly, messages which highlight that 90% of people are paying taxes are more likely to be effective than those which claim that half the people do not pay taxes or which warn of exemplary punishments.5 Clearly, the realization that you are deviating from the norm is very effective in discouraging potential violators. Further, in more practical terms, when the majority shirk, enforcement becomes expensive. It goes beyond the administrative capabilities of a functionally and geographically over-​stretched field-​level bureaucracy. For example, it is impossible for even well-​intentioned and committed officials to uniformly enforce law with the existing resources in any meaningful manner. It is no surprise that numerous such efforts over time to enforce compliance on apparently simple but pervasive problems

4 Gulzar Natarajan, ‘The Challenge With Law Enforcement’, Governance Agenda, Pragati, 1 March 2013 (http://​prag​ati.natio​nali​nter​est.in/​2013/​03/​the-​challe​nge-​with-​law-​enfo​rcem​ent/​). 5 Laura Haynes et al., Test, Learn, Adapt: Developing Public Policy with RCTs, UK Cabinet Office, Behavioural Insights Team, 14 June 2012 (https://​www.gov.uk/​gov​ernm​ent/​publi​cati​ ons/​test-​learn-​adapt-​dev​elop​ing-​pub​lic-​pol​icy-​with-​ran​domi​sed-​con​trol​led-​tri​als).

318  Possible Solutions like littering, not wearing helmets or seatbelts, public urination, etc., have very rarely succeeded.

13.2.2  Design Challenges The design of many MISs prevents them from being effective as decision-​ support for officials. Most MIS processes are conceived as large, top-​down software solutions implemented in rigid systems in tight compartments where problems and solutions and actors and their roles are defined and unchanging. They are tailored predominantly to address the imperatives of higher-​level monitoring. And this is as much a global problem as India’s, not exclusive to governments, and pervasive across the implementation of software projects in general. An analysis of around 50,000 software development projects of various kinds and sizes from across the world by the Standish Group found that only 29% of projects were considered successful and 19% were considered failures.6 Incidentally 62% of those considered successful were small projects. In contrast, of those considered successful, just 2% and 6% were grand and large projects, respectively. Further, the report also tells that projects with iterative agile implementation vastly outperform those with linear waterfall implementation. It also found that inadequate user involvement and incomplete specification of requirements were the primary reason for the failure of information systems.7 Another example comes from education. A 2017 World Bank report on building effective educational MIS (EMIS) stated that performance ‘fell short of expectations, with widespread deficiencies that ranged from unclear definitions and understanding of EMIS to ineffective implementation and utilization.’8 The report found that such projects were

6 S. Hastie and S. Wojewoda, (Oct 4 2015). Standish Group 2015 Chaos Report—​Q&A with Jennifer Lynch. Retrieved 19 July 2017 from https://​www.infoq.com/​artic​les/​stand​ ish-​chaos-​2015. 7 ‘A Look at 25 Years of Software Projects. What Can We Learn?’, Speed & Function, 7 January 2017 (https://​speed​andf​unct​ion.com/​look-​25-​years-​softw​are-​proje​cts-​can-​learn/​). 8 Husein Abdul-​Hamid, Namrata Saraogi, and Sarah Mintz, ‘Lessons Learned from World Bank Education Management Information System Operations: Portfolio Review 1998–​2014’, 2017 https://​openkn​owle​dge.worldb​ank.org/​bitstr​eam/​han​dle/​10986/​26330/​978146​4810​565. pdf?seque​nce=​2.

Addressing execution challenges—I  319 predominantly focused on developing a centralized EMIS as a tool for planning by the Ministry officials. Frontline officials therefore saw them as simply tools to report upwards to the central government and with limited relevance for them. It found hardly any example of EMIS that meaningfully informs a school inspector about which among his/​her 30 schools requires his/​her immediate, focussed attention and provides actionable insights for his/​her inspections to those schools, or which informs a headmaster about learning outcomes quality in his/​her own school. Take the persistent problem of technical and collection losses in electricity. More than a fifth, perhaps even a quarter, of the power generated is unaccounted for. Granular spot-​billed consumption and payments data, amenable for rigorous energy audits at various levels, would be of great value. However, few distribution companies use this data for frontline supervision and related decision-​support. Howsoever sophisticated the analytics and user-​friendly the interface, the sustainability of any measurement system depends on its institutional adoption. As reference points, there are many examples of e-​governance applications developed by District Collectors and other officials, which very rarely outlast their tenures. In sum, it is rare to find a well-​functioning and durable electronic MIS within governments at any level. Another deficiency is with decision-​support, even where there is no measurement problem. While governments collect large volumes of granular information with some periodicity, it is usually part of a top-​ down MIS and provides little or no real-​time decision-​support for field supervisors. Even the best field supervisors largely function with a combination of make-​shift and ad-​hoc monitoring arrangements and cognitive familiarity. By the time the information flows up the top-​down MIS and is consolidated and reported to higher-​level officials, it is often dated. More importantly, the level of actionability of such information is limited.

13.2.3  Incentive Distortions Honig and Pritchett, in an interesting and useful play of words, point out the limitations of accounting-​based accountability and contrast them

320  Possible Solutions with account-​ based accountability. ‘Accounting-​based’ accountability is based on accounting—​rules and processes to measure through quantitative indicators (including financial and physical account books).9 As we have seen above, such accounting has its set of problems, thereby creating limits to such accountability. This is contrasted with account-​ based accountability. Here the ‘account’ has a different meaning: the description or narrative (somewhat as in the phrase ‘she gave an interesting account of her visit’) which officials use to justify their actions. In most governance situations, accountability requires a combination of accounting-​and account-​based monitoring—​i.e. quantitative and qualitative accountability—​with their relative importance depending on the nature of the task. Pritchett makes the distinction as follows10: The fundamental notion of accountability is an account, which is a narrative, a story we tell to those to whom we feel we owe a justification (which can included both authorizers who have provided us with resources and levers to act, professional and occupations peers with whom we share identify and seek approval, and those on behalf we were supposed to have acted, and social peers) about why what we did was the right thing to do (or not) in the circumstances to advance the shared objectives and within the accepted norms. Our account may have some hard numbers and data—​what I call ‘accounting’—​but the fundamental issue is the account. The question is whether accountability can be fully exhausted by accounting. With Moore’s law increasing by orders of magnitude the power to do transmit series of zeroes and ones (information in the Shannon information theory sense) it is increasingly asserted accountability can be reduced to hard, objective, computer storable, numbers and that technology will help governments improve accountability.

9 Dan Honig and Lant Pritchett, ‘The Limits of Accounting-​ Based Accountability in Education’, Center for Global Development Working Paper 510, May 2019 (and Far Beyond) (https://​www.cgdev.org/​sites/​defa​ult/​files/​lim​its-​acc​ount​ing-​based-​acc​ount​abil​ity-​educat​ion-​ and-​far-​bey​ond-​why-​more-​acc​ount​ing-​will.pdf). 10 Lant Pritchett, ‘Account Based Accountability and Aid Effectiveness’, Building State Capability Blog (https://​buil​ding​stat​ecap​abil​ity.com/​2018/​06/​28/​acco​unt-​based-​acc​ount​abil​ity-​ and-​aid-​effect​iven​ess/​) (Accessed on 30 December 2018).

Addressing execution challenges—I  321 Dashboards take care of only the accounting part, and that too with all the problems discussed above. Worse still, they may contribute to marginalization of the account-​based accountability dimension as well as other qualitative assessments. There are numerous examples of governments and officials at different levels from across Indian states being entrapped within their dashboards even as different reality plays out in the field. Even as the underlying figures and data workflow get manipulated by entrenched vested interests, the form of the dashboard continues to offer the illusion of control. Given the inherent difficulties of quantifying and capturing performance measures, when the monitoring stakes go up in this age of publicity-​driven governance, there is often a temptation to ‘game’ the data. Faced with intense public scrutiny, officials and governments, in particular, are loath to paint unflattering portraits of government programs. In these circumstances, instead of helping with performance management, dashboards degenerate into helping with looking good.

13.2.4  Possible Solutions The first way to address these problems may be to invert the design of the information management system and prioritize the concerns of the frontline supervisors. Digital supervision and monitoring systems have to combine equipping frontline managers with functional decision-​support and higher-​level managers with robust monitoring. Systems need to be designed from the perspective of the frontline supervisors and then work upwards to meet the requirements of higher-​level monitoring. Further, instead of collecting new data, it may be useful to limit the objective to making better use of existing administrative data. Information presented on such dashboards has to be both relevant and actionable. The relevancy requirement depends on the functional responsibility of the official concerned.11 The actionability requirement demands that the

11 Lant Pritchett and Gulzar Natarajan, ‘One Report Card Is Not Enough’, The Indian Express 15 March 2014 (https://​indian​expr​ess.com/​arti​cle/​opin​ion/​colu​mns/​one-​rep​ort-​card-​is-​not-​ eno​ugh/​).

322  Possible Solutions information is actionable not only in theory but also given the political economy and state capacity constraints. A second way is to have MISs and dashboards adopted as the only information management system within a department or the government. In other words, they should be the exclusive basis for all the reviews about the particular activity or program at all levels of the government. All reporting about the activity should be extracted from the MIS. Currently, even where MISs are in place, performance reviews continue to rely on a mix of the MIS and separate physical reports, which often do not converge. This dilutes the importance of the MIS and comes in the way of efforts to improve it. This assumes great practical relevance since if supervisors and monitoring officials at all levels discard all parallel physical reporting systems and start using the MIS to monitor performance, then it creates the incentives and imperatives to iterate and improve the quality of the system. It then becomes a matter of time before the system gets institutionalized. Theoretically, another way to institutionalize MISs would be to integrate such a data management system into the individual performance appraisal of personnel at all levels (i.e. the appraisal of their performance for personnel decisions like pay and promotion). However, the history of public performance management points to at least four formidable challenges.12 First, public service delivery, especially its quality, is not often amenable to credible quantification. How do we quantify cleanliness or public health or learning outcome improvements which lend themselves better to what Pritchett calls account-​based accountability? Second, performance management raises the stakes associated with operational data from public services, thereby generating strong incentives to game the data management process. How do we ensure that the declaration of open defecation-​free status or student test scores are not gamed? Third, with many development activities, context is important and varies widely, and success is contingent on too many factors beyond the control of the officials. How can we hold medical officers accountable when some are hobbled with over-​burdened ANMs and large numbers of vacancies and 12 Prajapati Trivedi, ‘The Rise and Fall of India’s Government Performance Management System’ (http://​asq.org/​gov/​2018/​02/​the-​rise-​and-​fall-​of-​ind​ias-​gov​ernm​ent-​perf​orma​nce-​ man​agem​ent-​sys​tem.pdf).

Addressing execution challenges—I  323 others are not, or compare those in rural areas with their urban counterparts? Finally, outcomes are often the result of combined actions of multiple Departments. How can Health Department officials be held responsible for the outbreak of an epidemic when the Municipality was responsible for cleanliness? These are formidable challenges and have no easy answers. Credible and comparable measures of performance are scarce and its application in performance management is fraught with practical difficulties. In the circumstances, a prudent start would be to start with a small number of easily quantifiable and robust metrics which are non-​contentious and amenable to credible comparisons and avoid all else. Direct monetary rewards or punishments on the basis of such measures should be eschewed. Further, the aggregates for the performance management system should be drawn directly from a data workflow starting with the most granular base data. This would limit opportunities for manipulation and gaming of the process. If the environment is not right and these conditions cannot be met, then (however attractive in theory) the link between the MIS and personnel appraisals should not be attempted. Furthermore, in light of the weak state capacity, any such MIS-​linked performance management has to be accompanied by high-​ quality coaching and mentoring. An effective performance management system should be more about creating the enabling conditions to enhance competitiveness and less about trying to capture the shirkers and the inefficient. Capacity building is critical to the success of any performance management system. The kind of high-​stakes performance management (with strong ties between measured performance and personnel decisions) that many outside critics of public systems desire is not immediately feasible—​and may not even be desirable because of its unintended consequences. Putting all this together, Figure 13.1 summarizes the elements of an effective information management system that meets the requirements of monitoring and supervision. To conclude, implementing an effective electronic MIS to monitor the implementation of government programs and initiatives requires careful attention to design and detail. As a checklist for public officials seeking to develop MIS, a few questions are in order:

324  Possible Solutions Data Management in Public Services Delivery Performance Management System

Functional Use

Higher Officials

Feedback High level monitoring (Expost) Direct supervision (real-time)

Supervisors

Implementation Support (real-time)

Implementors Data work-flow

Figure 13.1  Data management in public services delivery

• What is the purpose of the system? • Who are its primary users? • How much of it responds to the concerns of frontline supervision and how much to that of higher-​level monitoring? • How relevant and actionable is the information rendered? • Is the data management part of a unified workflow? Finally and separately comes the question of how the MIS is integrated into the personnel performance appraisal of the administrative unit. This is even more complex and should only be attempted if the pre-​conditions are in place. A properly designed MIS can improve supervision and outcomes even if not integrated into personnel appraisal systems.

13.3  Delivery Units A popular idea for improving the state’s monitoring capacity is the establishment of ‘delivery units’ (DUs) or ‘delivery labs’ at high levels of the

Addressing execution challenges—I  325 government. The consultancy firm, McKinsey, has even formulated a management approach, Deliverology, around such units.13 Irrespective of its design variations, the underlying premise of DUs is that relentless top-​down high-​quality performance data analysis delivering actionable insights can enhance the effectiveness of program implementation. In the hands of the top political leaders it can be a useful tool when regularly used. The logical appeal of DUs can be irresistible. For governments desperate to be seen doing something proactive in delivering on their promises, such units are attractively salient. It is therefore unsurprising that the concept has become something of a fad in development. Many donors too are keen to support governments establish such units. The pioneer with DUs was the British government under Prime Minister Tony Blair, which created one in 2001 within the Cabinet Office to monitor the progress of implementation of the Prime Minister’s core priorities. The Prime Minister’s Delivery Unit (PMDU) had a combination of civil servants and private individuals sat outside the Ministry structure and reported directly to the Prime Minister. It analysed data and presented progress reports to the Prime Minister on his important priorities. In 2009, the Malaysian government set up a Performance Management and Delivery Unit (PEMANDU) to lead the effective implementation of its national transformation programs. It was set up in the Prime Minister’s Office (PMO) and was headed by a Chief Executive Officer, leading the Government Transformation Program and Economic Transformation Program. Its areas of focus included crime prevention, reduction of corruption, and improving rural infrastructure. In more recent times, the Punjab Government in Pakistan spearheaded its education sector reforms by establishing a DU within the office of the Chief Minister, Mr Shahbaz Sharif. Sir Michael Barber, the first head of the British PMDU, worked closely with Mr Sharif to establish the unit.

13 Michael Barber, Paul Kihn, and Andy Moffit, Deliverology: ‘From Idea to Implementation’, McKinsey and Company, February 2011 (https://​www.mckin​sey.com/​ind​ustr​ies/​pub​lic-​sec​tor/​ our-​insig​hts/​deliv​erol​ogy-​from-​idea-​to-​imp​leme​ntat​ion).

326  Possible Solutions Similarly, in Tanzania President Jakaya Kikwete established DUs in six Ministries to monitor their programs.14 Some of these DUs, like in Malaysia and Tanzania, have also sought to become policy design labs, bringing together all stakeholders to quickly brainstorm and design policies in response to emergent issues. Several other countries too have established such units, though little remains know about their actual operational effectiveness. They have been established within Departments and even within local governments. Jen Gold identifies around 25 such units, though some of them are units only in name.15 In several cases, the DUs closed down with the exit of the respective political leader at whose instance it was established. In fact, many units that were created with great fanfare have been closed down, and several others are just limping along. While comprehensive assessment of DUs is not available, there is enough evidence to suggest that the results have been mixed. Apart from a handful of successes, most of the DUs have little to show for. Sample this assessment by the UK’s Institute for Government16: Our report, Tracking delivery, traces what’s happened to delivery units over the past 16 years. It shows that governments can pay a heavy price for ineffective units, and not simply in terms of wasted resources. When units lose influence, for instance, their continued existence can cultivate a false sense of security that projects and programmes are being properly monitored, sometimes with devastating results . . . Or they can institutionalise confrontational relationships between centres of governments and line departments.

14 Andy Brock, ‘Pulling the Lever: Can Delivery Units Deliver?’, Development Policy Blog, 12 March 2014 (http://​www.devpol​icy.org/​pull​ing-​the-​lever-​can-​deliv​ery-​units-​deli​ver-​20140​ 312/​). 15 Jen Gold, ‘Tracking Delivery: Global Trends and Warning Signs in Delivery Units’, Institute for Government, 27 April 2017 (https://​www.ins​titu​tefo​rgov​ernm​ent.org.uk/​sites/​defa​ult/​files/​ publi​cati​ons/​Glo​bal%20D​eliv​ery%20rep​ort.pdf). 16 Jen Gold, ‘How to Make Delivery Units Work’, Institute for Government, 17 May 2017 (https://​www.ins​titu​tefo​rgov​ernm​ent.org.uk/​blog/​how-​make-​deliv​ery-​units-​work).

Addressing execution challenges—I  327 Some common features of such DUs are evident. For a start, all of them stand on political authority, with strong backing at the highest level of political leadership. In fact, the primary users of such units are the President or Prime Minister themselves. They are decidedly not bureaucratic monitoring systems. Second, they sit outside the line departments and provide independent data analysis for monitoring. Third, they consist of both civil servants and private sector experts. In India, too, there have been such units. In May 2007, the Government of Tamil Nadu, under the late Chief Minister Mr M. Karunanidhi, set up a Monitoring Unit with very similar objectives, headed by a Secretary reporting directly to the Chief Minister (who happened to be one of the authors of this book). The unit monitored the delivery of key government programmes and important citizen services, providing an independent source of information on actual results on the ground. As an example, during a period of power shortages, it directly obtained independent weekly feedback on actual hours of availability of power from a sample of villages to cross-​check data coming from the Electricity Board, which enabled corrective action. It was dissolved after the change of government in 2011. A Delivery Monitoring Unit (DMU) was established in the PMO in 2009. The DMU was responsible for rigorous high-​level monitoring and close coordination of flagship programs and certain prioritized projects and initiatives. The Ministries and Departments whose programs and projects were being monitored were asked to submit quarterly status reports in prescribed formats. Its mandate included ensuring effective delivery of selected programmes through steady monitoring for stated output by the Ministry concerned; fast-​tracking implementation and ensure trouble-​shooting through periodic reviews; informing the Prime Minister on a quarterly basis on performance under these programmes; ensuring the evaluation of the impact of selected programmes; and ensuring that information on these programmes, to the extent possible, is placed in the public domain of the Ministry concerned with responsibility for the information being that of the Ministry concerned. The DMU consisted of a Committee of Secretaries headed by the Secretary to the Prime Minister. Monitoring format and milestones were prepared for each program/​project/​initiative

328  Possible Solutions and quarterly reports were uploaded on the website by the Ministry/​ Department concerned. A recent DU-​like initiative was the Real Time Governance (RTG) initiative and CM dashboard in Andhra Pradesh under the Chief Minister, Mr Chandrababu Naidu. The Pro-​Active Governance and Timely Implementation (PRAGATI) initiative was started in 2014 by the Government of India under Prime Minister Mr. Narendra Modi. It is not a DU in the conventional sense but has several similar features and objectives and, in some cases, goes further than a DU. It is not a data collection, analysis, and monitoring system. Instead it uses digital data, video-​conferencing, and geospatial technologies to help monitor the progress of identified projects and programs and, importantly, help de-​bottleneck projects and programmes with special attention on those that have been stuck for a long time due to coordination failures. Lack of coordination between central Ministries, between states, and between centre and states has been a perennial problem with large projects and schemes. An important component of PRAGATI (which is not typical of DUs) is that the Prime Minister directly reviews pre-​identified programmes and projects with Secretaries of the Government of India and the Chief Secretaries of states through video-​ conference. PRAGATI has been a major success, especially successful in expediting the progress of long-​pending infrastructure projects. Many projects languishing for years or even decades have been completed. It has also been important in ensuring the delivery of key government programmes and in the redressal of common public grievances. In the first 35 PRAGATI meetings, 290 projects costing Rs.13.60 lakh crores, along with 51 programmes/​schemes, and grievances pertaining to 17 different sectors were intensively reviewed.17 The top-​level commitment of the Prime Minister and the specificity of its focus have been critical to the success of PRAGATI. DUs are a form of the more generic project management unit (PMU) but with a mandate that is broader. They monitor several programs as well as helping to surface and resolve implementation problems associated

17 https://​www.pmin​dia.gov.in/​en/​news_​upda​tes/​pm-​cha​irs-​36th-​prag​ati-​meet​ing/​ (Last accessed on 30 April 2021).

Addressing execution challenges—I  329 with them, including design improvements. The one clear difference is that unlike DUs, PMUs are clearly bureaucratic units established to assist with regular program monitoring. The key purpose of a DU to the political executive is the improved quality and speed of feedback on the success or otherwise of initiatives, which is important for a government wanting results before the next election. If the officials manning the DU have the breadth and maturity to do this without second-​guessing or supplanting line management or seeking personal authority, DUs can help considerably with this objective: but strong executive backing from the elected chief executive is essential. A Chief Minister’s DU without the explicit backing of, and relentless use by, the Chief Minister is unlikely to be effective; equally a DU with officers who seek to give direct orders to line departments or take over decisions that ought to be taken elsewhere is also bound to do more harm than good. DUs need to be headed by officers with wide knowledge and small ego and populated by a small team of young and energetic officials.

13.4  Greater Emphasis on the Qualitative Dimension of Accountability A comparative study of school education bureaucracies in Himachal Pradesh and Uttarakhand found interesting differences in both the operational norms and outcomes between the two states.18 While the former was found to rely on norms that promote local initiative and outcomes, the latter has been found to rely on inputs and processes based on legalistic norms and compliances. In terms of realization of student learning outcomes, Himachal outperformed Uttarakhand significantly. ‘Account-​based’ accountability could help align the behaviours and norms of officials in the realization of systemic objectives. Achieving

18 Akshay Mangla, ‘Bureaucratic Norms and State Capability in India’, Asian Survey, 55, no. 5 (2015): 882–​908 (https://​www.hbs.edu/​facu​lty/​Publ​icat​ion%20Fi​les/​Asian_​Sur​vey_​5505​_​03_​ Mang​la_​2​19b4​d54-​2a4f-​4a43-​a0ab-​d6326​28f8​fae.pdf) (Accessed on 20 December 2018).

330  Possible Solutions this cultural shift is difficult, but instilling this insight among newly recruited government officials at all levels is perhaps the best starting point. While local initiative and motivation is undoubtedly important, one must not ignore the importance of the procedural and legal norms. These norms may appear perverse and irritating but their basic purpose is to achieve fairness between different people, keep records to demonstrate accountability, and ensure probity in the use of public funds. All of these are essential and laudable and cannot be achieved without some degree of what is pejoratively called ‘red tape’. One could reasonably agree that in India, the legal norms and documentation requirements are excessive and local initiative and empowerment deficient. But instead of an either-​or situation, what is needed is a combination of the two—​a bureaucracy that is committed and motivated and adheres to procedural requirements. Meeting both the requirements demands action at two levels—​right-​sizing the procedural requirements and figuring out a way to instil and entrench outcomes-​focused norms and encourage local initiative.

13.5  Enhanced Delegation and Discretion In Chapter 10, enhanced delegation was recommended in the context of policymaking, but (as noted there) it is a measure that can also improve execution. Delegation does not merely mean enhanced financial powers. A closely related area for improvement is a qualitative increase in delegation through enhanced discretion and empowerment (within legal and procedural norms). The Indian administrative system is often over-​centralized and ‘more delegation’ is probably a good rule of thumb to apply in most situations. The system is also overly prescriptive in terms of process and form; what it needs is a reduced emphasis on form and procedure with more attention to content and substance.

Addressing execution challenges—I  331

Studies have found that discretion available to frontline bureaucrats enhances their effectiveness.19 First, discretion influences client meaningfulness because street-​level bureaucrats are more able to tailor their decisions and the procedures they have to follow to the specific situations and needs of their clients. Hence, discretion gives street-​level bureaucrats the possibility to apply their own judgments when dealing with the needs and wishes

19 Lars Tummers and V.J.J.M. Bekkers, ‘Policy Implementation, Street-​Level Bureaucracy and the Importance of Discretion’, Public Management Review, 16, no. 4 (14 April 2014): 527–​547 (https://​pap​ers.ssrn.com/​sol3/​pap​ers.cfm?abst​ract​_​id=​2289​338&downl​oad=​yes).

332  Possible Solutions of citizens . . . At the same time, the positive effect that discretion has on the bureaucrat’s perception of client meaningfulness can be seen as a condition for the second effect: more willingness to implement the policy. When street-​level bureaucrats perceive that their work is meaningful for his/​her clients, this strongly influences their willingness to implement it. This is in line with the notion that street-​level bureaucrats want to make a difference to their clients’ lives. Furthermore, the results also point to another, more autonomous, effect that discretion directly influences willingness to implement; hence, discretion is inherently valued by bureaucrats.

This aspect assumes even greater relevance in the context of frontline or street-​level bureaucrats. Ethnographic studies have shown that while they have a considerable degree of operational autonomy, they also face serious constraints in discharging their functions.20 They struggle with inadequate resources, an ever-​growing demand for their services, vague or conflicting organizational expectations, and challenges to performance management. Complicating matters, given the government’s exclusive statutory role, they offer services to ‘captive’ clients, which too distort incentives. This description of the problem and its consequences from the last revised version of Lipsky’s seminal work is very appropriate21: Street-​level bureaucrats—​from teachers and police officers to social workers and legal-​aid lawyers—​interact directly with the public and so represent the frontlines of government policy . . . these relatively low-​ level public service employees labor under huge caseloads, ambiguous agency goals, and inadequate resources. When combined with substantial discretionary authority and the requirement to interpret policy on a case-​by-​case basis, the difference between government policy in theory and policy in practice can be substantial and troubling. The core dilemma of street-​level bureaucrats is that they are supposed to help

20 Michael Lipsky, Street-​Level Bureaucracy: Dilemmas of the Individual in Public Service, Russel Sage Foundation, New York, 1980. 21 Michael Lipsky, Street-​Level Bureaucracy: Dilemmas of the Individual in Public Service, Russel Sage Foundation, 30th Anniversary Edition, New York, 2010.

Addressing execution challenges—I  333 people or make decisions about them on the basis of individual cases, yet the structure of their jobs makes this impossible. Instead, they are forced to adopt practices such as rationing resources, screening applicants for qualities their organizations favour, “rubberstamping” applications, and routinizing client interactions by imposing the uniformities of mass processing on situations requiring human responsiveness. Occasionally, such strategies work out in favor of the client. But the cumulative effect of street-​level decisions made on the basis of routines and simplifications about clients can reroute the intended direction of policy, undermining citizens’ expectations of even-​handed treatment.

There cannot be any substitute to equipping the building inspectors and school inspectors with requisite resources and capability and empowering them to lead the process of delivering public services in their jurisdiction. Therefore, sustainable efforts to address implementation deficits have to involve attempts to revive the moribund systems at the frontline of implementation. In short, it requires an attempt at a transformation of frontline systems. A practical step that senior officers or Ministers can take is to carry out a systematic ‘zero base review’ of delegation in each organization, i.e. revisiting all existing orders of delegation, to see if delegation can be enhanced.

13.6  Use of Social Media Social media has opened up opportunities in information dissemination, awareness creation and social mobilization, and setting the agenda. It can disseminate information about social and other issues quickly among many people. It can also shape behaviours and attitudes. Social media campaigns have contributed significantly to revolutions and regime shifts in different parts of the world. It has the potential to mobilize large numbers of people within a short time and channel their energies to fight a common cause. Government officials and agencies could leverage this potential to benefit their activities. Social mobilization, awareness campaigns,

334  Possible Solutions behavioural change, and information transmission on public policy issues can benefit from social media engagement. The former External Affairs Minister, late Mrs Sushma Swaraj, and the Ministry of External Affairs received much acclaim for their adeptness with the use of social media in enabling access to common citizens. The Ministry was quick to put in place mechanisms to respond immediately and effectively to social media tagging of the Ministry and Minister by citizens on the spectrum of services offered—​visas to passport and other services. A tweet of an initiative of a District Collector by the Secretary to the government can create a scramble for emulation by peers. Similarly, a tweet by the Collector of an initiative by a Block Development Officer (BDO) can trigger a similar dynamic. The power of such competition should not be underestimated. In a few states social media has become the preferred medium for even official communications. District Collectors are enrolled into several WhatsApp or Telegram groups and receive real-​time instructions from different Heads of Departments. However, there are disadvantages too. Social media encourages quick wins and short-​term initiatives. This favours the more publicity-​ conscious officers over those who prefer to hunker down and painstakingly work on problems, who are also those more likely to make a meaningful long-​term contribution. Building toilets or schools and inaugurating them, or dashboards with latitude-​longitude facilities or biometric attendance of teachers offer great tweetable pictures. But diffuse and long-​drawn diligent action involving interventions to improve classroom instruction is not tweetable. Superior officials are more likely to tweet the former, thereby aligning incentives within the system to emulate them. The important could get marginalized by the tweetable! This can set in motion the practice of mindless replication of the so-​ called best practice interventions. The official who subscribes to too many WhatsApp groups can also become overwhelmed by the sheer overload of information. It is easy for anyone to shoot off 280 characters, especially if it also serves the purpose of both making one feel better as well as score social brownie points by signalling one’s commitment and interest in public

Addressing execution challenges—I  335 issues. Even better, there is no cost or accountability associated with the tweet. Like talk, tweet is cheap! Perhaps the most pernicious problem for the civil service is social media’s role in blurring the lines between the official’s personal and professional selves. Social media provides an unparalleled platform for people to reach out, build networks, and project themselves. The social media account becomes the personal Public Relations Officer (PRO). Social media may become a stage for personal popularity contests. Often officials struggle to resist the temptation to play to the galleries and boost their personal popularity. Since many officials also use their personal social media accounts for their professional lives too, the personal and professional get merged. Officers tweet their personal preferences, which immediately get the attention of those following the person for his or her official position. The question immediately gets raised whether this represented the personal or the official view. Most often, even before clarifications could be issued, the damage would have been done. Thus, social media can influence the nature of engagement with development problems. A cost–​benefit assessment of social media may tip to the negative in many instances.

13.6.1  Suggestions on Managing Social Media Officers engaging with social media should engage officially only through their official accounts. The personal account should not be used for discussing official issues, much less for official purposes. As regards the personal account, the same set of conduct rules on the public expression of opinion should govern their private social media engagements. The conduct rules may have to be elaborated to provide greater clarity on this, especially also to address the propensity of officials using social media accounts to compete for popular attention and glory. The time has come to ring-​fence the personal from the official in the virtual domain. Second, it is desirable to issue guidance about officials publicizing their official successes as personal achievements. This is important given that a fair share of social media engagement revolves around officers publicizing their successes. But official successes are rarely ever, if at all,

336  Possible Solutions exclusively the result of individual effort. It may therefore be appropriate to have some restraints on the use of personal accounts to disseminate official achievements. Further, it may be desirable to revive some of the standard practices of media engagement. Most government agencies have a PRO who is entrusted with external communications management. In the age of social media, many officers have themselves assumed the role of PRO. Besides engagement without any rules, this is also unproductive in so far as it displaces significant time and effort from the regular duties of the officers. Bringing back the PRO to mediate the social media engagement, too, albeit with a more liberal set of rules of engagement to account for the nature of the media and the changed times, may be good. Official accounts should be used where benefits are clearly evident and not indiscriminately, much less for personal publicity.

13.7  Use of Consultants and Advisors Faced with a thin state and newer types of responsibilities, it has become a common (and understandable) practice, in the Government of India as well as in state governments, to hire consultants to assist with policy formulation and implementation. Accordingly, their services are procured for clearly defined tasks—​a PMU, preparation of project reports or policy drafts, managing with public relations, etc. These activities require specific expertise and it is only natural to hire the services of those with the requisite expertise. The consultants’ role is to undertake deep dives and offer recommendations for reforms. The competent government officials within the bureaucracy are supposed to then consider them, exercise judgement and process the recommendations for decision-​making at appropriate levels. Even when related to core functions of the government, the consultants are expected to only provide inputs to support government officials in their decision-​making process. Further, the procurements have generally been confined to consulting services delivered by consultancy organizations. Individual consultants were not commonly hired to deliver such services.

Addressing execution challenges—I  337 However, in recent times, government agencies have started hiring consultants to undertake the core functions of the government itself. And instead of hiring consulting organizations to deliver specific services, individual consultants are being hired, effectively as contract employees to also do recurring administrative activities. For example, documentation of consultation and decisions through files within government agencies is considered a core activity. This was considered the exclusive domain of permanent employees of the state or central secretariat service. But today, in many state governments as well as some Ministries in the Government of India, consultants are being used, formally and informally, to prepare files. In ministries like Industries and Infrastructure, consultants have taken over the origination and documentation of government processes on economic projects. In many cases, consulting organizations have been hired not as service contracts but effectively as labour contractors—​say, deliver x number of people for 2 years. The Secretaries to state and central government, and Commissioners and Directors in states extensively use the services of consultants in their own offices. Several state Chief Ministers’ offices are filled with consultants. Important intermediate management positions in Departments—​ Assistant, Deputy, Joint, Additional Directors, etc.—​are being filled with individual consultants recruited without following the due process. The practice of hiring consultants has complemented another trend of appointing Advisors in Departments and Directorates of state governments. These Advisors occupy much higher positions and are most often conferred the rank of at least Secretaries to government, and often even Cabinet Ministers of the state government. Advisors are typically drawn from the pool of retired IAS officers and reputed persons from the industry and academia. As the name suggests, their role is to advise the government on their area of expertise. They work closely with the Secretary and the Minister in the activities of the government. While they are not part of the file processing system, they generally have enough sway in influencing the decision-​making process. There are at least four factors underlying this trend of using consultants. One, as we discuss in Chapter 6, governments face acute personnel scarcity. The rapid expansion in government’s activities has not been

338  Possible Solutions accompanied by anything close to a commensurate increase in its manpower. In fact, in many cases, the strength of government personnel has declined. This has been exacerbated by the rise in vacancies across levels. Consultants become a backdoor around the hiring freeze. Two, there is a similarly acute skill and expertise deficit. The eligibility and other recruitment conditions of candidates being appointed to various posts have not changed over time with the latest requirements. Existing personnel too have not acquired the requisite professional skills and domain expertise to meet today’s requirements. Three, many officials and politicians have come under the thrall of a narrative that prioritizes the expertise of consultants over that of government officials. In fact, hiring consultants to get work done has even become a signal of being progressive and efficient. Consultants offer the imprimatur of expertise and efficiency. Multilateral agencies and external funders invariably demand that consultants be hired to support project implementation. When all these factors come together, even the more circumspect officials and politicians prefer to go along with hiring consultants to get their work done. The factors that drive the appointment of advisors are different from that of consultants and are three-​fold. First, advisors are expected to bring rich experience and high-​quality domain expertise that can add significant value to policy formulation. While this is certainly the case in some cases, the additionality may perhaps not be significant in most cases. Second, rewarding loyal people with advisorships is a good opportunity to confer patronage. Chief Ministers, in particular, feel the urge to offer such sinecures to the senior Secretaries who have served the government on their retirement. The possibility of such positions helps in making them more amenable to the requests while in service. Third, an appointment of an advisor is also perceived as a signal that the government is committed to addressing the problems facing that Department. Often, advisors have specific mandates to help with the implementation of a project or a scheme. Consultants have indeed eased some of the work burden on governments. Many of them have also added significant value to policymaking and program implementation processes. However, while consultants are often valuable, the practice of using them indiscriminately and

Addressing execution challenges—I  339 undertaking core responsibilities of the government is very deleterious to state capability. Foremost, it goes contrary to some of the basic features of the Weberian bureaucracy. The classic bureaucracy is characterized by permanent bureaucrats who (for all the criticisms) embody experience and institutional memory which are critical to the effectiveness of their bureaux and who are also accountable for their actions. In contrast, even with file recording, consultants, who flit around for a couple of years, carry important institutional memory with them, leaving knowledge vacuums within the bureaucracy. Besides, private consultants in the bureaucracy are sometimes the rough equivalent of mercenaries in warfare, with limited means to enforce accountability. This moral hazard is highest with unaffiliated consultants. Even with affiliated ones, in case of non-​performance or misdemeanour, the long route of accountability between the consultant and the consulting organization and the organization and the public agency is tenuous at best. This is a matter of increasing concern across government levels where consultants are an important presence. Second, the revolving door of consultants provides an opportunity for all sorts of vested interests and lobbyists to influence policymaking surreptitiously. Given that consultants are making policy documents and also scrutinizing and circulating the relevant files for decision-​making, the possibility of serious conflicts of interest and subversion of government processes exists. In infrastructure sectors, where the role of consultants within and outside the government is most salient, small tweaks to policies could confer large potential benefits on private parties. This is a big risk when consultants are entrusted with the complete responsibility of procurements, including defining critical project specifications and eligibility qualifications. The possibility of corruption and subversion is never far away. In fact, anecdotal evidence points to this emerging as a matter of serious concern across sectors. Third, consultants have sometimes become trojan horses to disseminate new ideologies of questionable value within public systems. For example, consultants bring in the tools and techniques of mainstream management, which focusses on concerns of efficiency, value for money, cost-​cutting, and so on, often at the cost of those like fairness, equity, and

340  Possible Solutions even genuine political choices of democratically elected governments. Or more specifically, they have become the vanguard or even marketing arms of technology solutions providers and policy ideologies. Perversely, and exacerbating the problem, the same set of consultants often act on all three sides of the market—​contributing to hyping up the idea with thought leadership articles and research insight studies, advising private suppliers and solutions providers, and being consultants to governments procuring those services. Conflicts of interest can generate serious adverse consequences.22 Fourth, policymaking is not just about domain expertise. It is also about knowledge of the policy context and administrative processes. Consultants, even those with expertise, invariably struggle with the latter and may end up subtracting value. For example, it is not uncommon to have Ministries channelling consultant’s context-​agnostic general knowledge as considered Ministerial opinion on Cabinet Notes. These, in turn, not only detract from the quality of policy decisions but also end up delaying the Cabinet Note approvals process. Fifth, excessive use of consultants contributes to enfeebling of the civil service itself. Instead of strengthening the state, it gets weakened. The practice of hiring consultants ends up minimizing government but also minimizing governance. Finally, the pervasive practice of using expensive consulting organizations to do routine administrative activities like recording minutes, follow-​up of decisions, consolidating material for internal presentations and processing of files, and so on is a poor use of scarce public resources. The consequences of appointing advisors are also mixed. While some advisors offer genuinely high-​quality expertise and rich experience, they are likely to be very few in number in most states. This is especially true with IAS officers since the more enterprising among them often end up gravitating to the Government of India and retire there. They have access to a wider pool of more powerful post-​retirement positions. In one sense, the case for ex-​civil servants as advisors is even weaker than for consultants who bring specialized outside knowledge. Most

22 Carey Baraka, ‘The Failed Promise of Kenya’s Smart City’, Rest of World, 1 June 2021 (https://​ rest​ofwo​rld.org/​2021/​the-​fai​led-​prom​ise-​of-​ken​yas-​smart-​city/​).

Addressing execution challenges—I  341 advisors are simply older and more senior equivalents of the Secretaries themselves and their ‘specialized’ knowledge is highly questionable. Also, unlike consultants, advisors are likely to have a constraining effect on the system. Secretaries to Departments will invariably be looking over their shoulders to advisors. Not only are they likely notionally higher in rank, but they are also seniors to them in the IAS cadre. The authority that comes from these can be constraining for Secretaries. As it is, managing a two-​way relationship between the Secretary and Minister is non-​trivial and a not uncommon cause of stress and paralysis within Departments. Adding an advisor to the mix could potentially compound the problem. Tales abound from states of advisors constraining Heads of Departments, even marginalizing them. Further, the likelihood of advisors representing vested interests is a matter of concern. This assumes greater relevance with the appointment of private individuals. The possibility of revolving doors and influence peddling is real. The fact that these advisors sit right at the top of the decision-​making system amplifies the concerns and demand robust safeguards.

13.7.1  A Balanced Approach Clearly, consultants can fill capability gaps by bringing in much-​needed specialist and outside knowledge and by augmenting quantitative capacity for temporary or short-​term needs. When capability gaps exist, the long haul of recruiting and training cannot be the only answer. The right approach involves the careful and selective use of consultants to fill gaps rather than using them as a substitute for routine government work. There will always be new and specialized tasks in every era and every sector where in-​house expertise will not be enough to obtain the best results. Consultants—​and occasionally advisors—​can and should be used in these contexts. However, when engaging consultants, if the task is not purely temporary, there needs to be a clear plan for knowledge transfer and training as part of the consulting assignment itself. This needs to be part of the terms of reference of the consulting firm with specific financial fees earmarked for this process along with credible performance measures. The

342  Possible Solutions Department engaging consultants will also need to assign promising junior civil servants as understudies to the consulting team (if necessary, reversing the typical hierarchy where the officer sits on top of the consultant). Consulting assignments and advisory positions must, above all, not be substitutes for internal qualitative capability and quantitative capacity. There is no substitute for regular recruitments to perform core activities of the government. The core functions can neither be outsourced nor digitized. The bureaucracy will also have to develop expertise in newer areas like engaging with the private sector. For performing routine activities of government, even if there is a shortage of personnel, direct contracting of qualified personnel instead of expensive service contract with consultants should be preferred. Merely hiring more people will not do. The recruitment conditions—​ eligibility and qualifications, recruitment processes, service conditions, etc.—​should be adapted to the changed requirements. Civil servants need exposure to the latest in domain expertise and technology trends. But more importantly, they should be sensitized to the demands of the modern state, with its greater transparency, increased public expectations, and rights-​based service delivery requirements. Training curricula need to be revamped significantly to meet these requirements. As regards advisors, it is necessary to have clearly defined mandates for them, explicitly making the distinction between that of advisors and heads of departments. And the greater challenge is to adhere to the distinction. Advisors, like consultants, should be prohibited from being part of the file circulation process. No doubt it can be subverted (for example, file note being written by the consultant but circulated by the official concerned), but explicit prohibition has a deterrent value and helps in fixing responsibility.

13.8  The Enforcement–​Harassment Trade-​Off An important source of corruption is the interaction between public officials and citizens on the enforcement of laws. The stereotype of harassment by inspectors across public services is well known. The problems with labour department inspections led to the demands and introduction

Addressing execution challenges—I  343 of a series of measures to limit the possibility of corrupt practices during inspections.23 Given that the enforcement role of government cannot be avoided, the challenge is to design enforcement regimes that limit the likelihood of harassment and of corruption. There is a useful distinction to be drawn between harassment and corruption. In this chapter: • ‘harassment’ means regulatory costs and inconvenience imposed on a law-​abiding person due to false suspicion of wrong-​doing; and • ‘corruption’ means the extraction of pecuniary value by public officials; corruption may be either to condone wrong-​doing or simply to avoid harassment even though no wrong has been committed. Over-​zealous honest officers may harass without being corrupt; corrupt officers may extract bribes from someone who actually has evaded the law without involving harassment. The following examples illustrate the difference: • A person has not evaded taxes. A suspicious tax officer unreasonably summons him for repeated hearings and orders a special audit but does not expect or demand a bribe—​he is just over-​zealous in trying to achieve revenue targets. This is harassment. It creates an economic loss. • A person has not evaded taxes. A dishonest officer behaves exactly like the officer in the previous example but offers to drop the proceedings if a bribe is paid. The person (fed up with the costs of the proceeding) agrees to pay just to avoid the nuisance. This is corruption with harassment. It creates an economic loss greater than the previous example. • A person has evaded taxes. An officer demands a bribe to overlook the evasion. This is corruption, but not harassment. Note that though the enforcement action is legitimate, the government may not get the full revenue due. However, the deterrent effect on the evader is 23 Transparent Central Labour Inspection Scheme for Random Inspection of Units, Ministry of Labour and Employment, Government of India (https://​lab​our.gov.in/​tran​spar​ent-​cent​ral-​ lab​our-​ins​pect​ion-​sch​eme-​ran​dom-​ins​pect​ion-​units) (Accessed on 18 December 2019).

344  Possible Solutions partly achieved because a cost is imposed on him. If the sum total of the bribe and the tax paid is identical to the tax due, there is a financial loss to the government, but no economic loss since the deterrent effect on the lawbreaker is unchanged. In practice, corrupt officials and tax evaders look for win-​win deals where the total of the bribe plus tax is less than the tax due; in such cases there is both an economic and a financial cost to society. From a purely economic point of view, corruption with harassment is the worst, mere harassment is second, and corruption against the lawbreakers is the least costly. Public agencies engaged in the enforcement of the law, including enforcement of tax law, face a challenging trade-​off. On the one hand, they need to either minimize regulatory violations or maximize revenues, whereas, on the other, they are also expected to maximize citizens’ welfare and provide honest, courteous, and good service to those they interact with. Tax collectors are expected to be tough on evaders and nice to honest taxpayers; traffic police are expected to be strict with rule-​ breakers but kind to those who make an innocent mistake. But the steps, procedures, or attitudes which help with one objective may ipso facto hinder the other. This conflict is not unique to India, but it is made more difficult by the seemingly greater ‘propensity to evade’ which is present in the country. The Goods and Services Tax (GST) is an example of such trade-​offs. Getting a business registered under the old Value Added Tax administered by the states was perceived to be a source of much corruption. Applicants had to wait for a tax officer to visit their premises and approve their application, and very often a bribe was involved. Under GST, the procedure was made much easier—​the entire process became online and virtually automatic; unless the tax officer objected within 7 days, registration was automatically granted. But the experience from 2017 to 2020 showed that a large number of non-​existent businesses got registered, which were used to generate spurious invoices that were used to take credit for taxes not paid. Literally thousands of such cases with tax evasion of billions of rupees were detected. Eventually in 2020 the government decided to re-​introduce new procedures requiring physical presence at an application centre and/​or inspection by a tax officer. This

Addressing execution challenges—I  345 helps prevent tax leakages through fake invoices—​but may also re-​introduce the possibility of corruption in registration. Table 13.1 tries to capture the trade-​off involved in enforcement actions. Tighter procedures reduce the likelihood of evasion but increase the likelihood of false positives (i.e. suspicious cases that are really innocent) and of corruption. Harassment (e.g. ‘catching’ someone, summoning them to a government office, and checking their records when they have actually done nothing wrong) is a cost to the economy, which reduces output and deters investment. When harassment is accompanied by corruption it is worse. Looser procedures reduce corruption and economic costs of false positives but increase the chances of evasion of the law with resultant economic costs and/​or fiscal loss to the exchequer. The GST registration case was one where there was tangible evidence that a taxpayer-​friendly system was being abused and that the financial losses were significant. In many cases, Indian law enforcers lean too much on heavy-​handed enforcement, oblivious of the costs to honest citizens. A similar issue is of, say, tax refunds to exporters. One approach would be to view the primary objective as incentivizing exports. This would suggest a procedure that is simple and easy for claimants but backed up by risk analysis and swift, targeted action against abuse. This would minimize corruption and delay but may mean tolerating a higher level of evasion. Another view could be to view the prevention of tax evasion as the primary objective. This would require a very thorough process Table 13.1  Harassment-​Enforcement Trade-​Off Matrix Severity of enforcement actions/​procedures Low Probability of Harassment/​corruption Probability of evasion of the law

High

Less

More

More

May be less or unchanged: • If corruption is low, tighter enforcement will reduce evasion. • If corruption is high, evasion may not diminish with severe enforcement actions.

346  Possible Solutions of checking and approving the vouchers supporting the claim. It would mean slower claim processing and more opportunity for bribery and probably less tax evasion. But genuine and honest taxpayers would suffer much more under the second system. And over a period of time, because of corruption, dishonest taxpayers and dishonest tax collectors may collude so that public revenue gets diverted into private (bribe) revenue. In such a situation, the revenue under the ‘tight’ system may not be much higher than under the ‘liberal’ system and a lose-​lose outcome (low revenue with high harassment) may arise. Such lose-​lose equilibria are frequently found in Indian public administration. Tightening of enforcement occurs through two sources. The first is through well-​intentioned honest officials trying to reduce evasion. The second is through lobbying by corrupt officials seeking opportunities for bribes, using public interest as a cover. Often, even well-​intentioned officials do not fully consider the trade-​offs pointed out in Table 13.1 and come up with enforcement regimes that are sub-​optimally strict.

13.8.1  Data Analytics: Weak vs. Strong Signals In this context, data analytics presents an interesting conundrum when applied as regulatory and enforcement decision-​support. On the one hand, more data generates more insights about the trends associated with underlying activities. On the other hand, paradoxically it also opens up possibilities for harassment and corruption. In other words, the effective use of data analytics becomes increasingly challenging with the increase in the depth (volume) and breadth (categories) of data available. The job of an expert data analyst is to capture all indicators of violations and fraud. But however good the indicators are, there will inevitably be false positives. Therefore, administrators who use this as decision-​ support need to use their judgement to prioritize among signals and use them accordingly. But far too often, given the ease of generating electronic notices, enthusiastic officials are likely to be inclined to issue them indiscriminately whenever a ‘deviation’ is flagged by the data analytics unit. This results in harassment, if not corruption. In this context, it may be useful to make the distinction between strong and weak signals of likely violations and deviances. While a weak signal

Addressing execution challenges—I  347 could have been generated either by a legal or illegal transaction, a strong signal is one where the probability of guilt is high. Strong signals minimize false positives. Under GST, payment of most of the tax liability through input credit is associated with many cases of fake invoicing. But it is also associated with exports which are tax free! So the indicator of ‘low proportion of cash payment’ is a weak signal. But an indicator of ‘low proportion of cash payment with no exports and no exempted turnovers’ may be a strong signal. In terms of enforcement action, there should be a premium on strong signals and a discount on weak signals. Actions carefully targeted based on strong signals of violations are both cost-​effective and minimize the likelihood of harassment. Issuing electronic legal notices to all commercial taxpayers who fail to submit their returns in a month is an example of acting on a weak signal since it would also include those with genuine problems beyond their control. Besides, this trend is also a reflection of ‘lazy administration’. But issuing notices to those who have consistently failed to submit their returns over several months is about acting on a strong signal. Factory inspection protocols for effluent emission compliance or adherence to labour standards which are fully or mostly universal are both expensive as well as run the risk of harassment from inspecting officials. The Labour Ministry’s use of ‘risk based objective criteria’ to target inspections is another such example.24 An air bombing campaign may end up maiming and killing innocent people; air attacks should seek to minimize collateral damage. The enforcement-​harassment trade-​off is similar and must seek to minimize costs to those who are law-​abiding. The primary value of enforcement actions lies not in the specific cases detected but in their deterrent value. In fact, if the majority of people default, enforcement actions are most likely to be both blunt and impractical. Enforcement actions are useful and effective when deviation is marginal and compliance is the norm.25 Further, in such instances, the value of enforcement is in its deterrence signal to prospective defaulters.



24 Ibid.

25 Natarajan, ‘The Challenge With Law Enforcement’.

348  Possible Solutions

13.8.2  Lessons for Enforcement Systems Design The preceding discussion yields the following lessons for the design of enforcement activities: a) In many situations, a slightly higher level of evasion with less harassment or corruption may be better for society than an over-​strict regime that imposes significant economic costs. b) Processes for the design of enforcement procedures and activities and administrators in charge of enforcement functions must actively and explicitly consider the cost to society of false positives and should keep in mind the trade-​off—​‘maximizing’ strictness is almost always sub-​optimal and promotes corruption and harassment. c) Performance assessment procedures for enforcement staff must explicitly assign a substantial weight for avoiding harassment and false positives—​failure rates of enforcement actions (i.e. enforcement not actually detecting violation of law) must be monitored and treated as a negative indicator of performance. d) Data analytics is useful—​but needs to be used with discretion. A clear distinction between strong and weak signals must be made. e) Purely system-​driven actions must not be resorted to except on the basis of well-​chosen ‘strong signals’. f) Focussing on strong signals economizes on regulatory resources and increases effectiveness.

14 Addressing Execution Challenges—​II Personnel, Procurement, and Permits

Arguably the three biggest traditional contributors to the erosion of accountability and rampant corruption across government bureaucracies are failures in personnel management, public procurement, and issuing of permits and licences—​which (to coin the kind of short form beloved of modern management commentators) can be called the ‘3 Ps’. The popular narrative of neta-​babu raj is largely built around these activities. All three often suffer from political interference (defined in Chapter 6 and simply summarized as ‘politicians taking or influencing decisions that by law and/​or procedure ought to be taken neutrally by civil servants’). The deeply politicized and rent-​seeking nature of much of recruitment and transfer, procurement and licensing, and permits, especially in, but not limited to, the lower levels of bureaucracy, convey a very high level of institutional decay. Each of these is discussed below.

14.1  Personnel Management: Transfers and Postings Bribery in transferring government officials at various levels, ranging from teachers and village officers to senior tax officials, is perceived to be widespread across many state governments. Corruption in recruitment is also believed to be common. (To be fair, in both recruitment and transfers, there are also commendable islands of integrity.) Frequent and arbitrary transfers of officials are a feature of personnel management within government agencies across India. There are many instances where political leaders place their favourites in important positions to exercise influence over the Department as well as to further their State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0014

350  Possible Solutions personal rent-​seeking ambitions. The former is essential to dispense political patronage through preferential and extra-​legal access to public services. The latter generates the rents from political power.

Any meaningful efforts to reform personnel deployment and procurement have to focus on limiting such political interference. Personnel deployment reforms should strive for at least three essential features. One, transfers in each Department should become institutionalized in the form of a fair transfer policy. It should clearly define those eligible, the criteria, and the process. Any deviation should be only under permissible circumstances. The Department should preferably audit all its annual transfers within a defined time. The Head of Department should be made

Addressing execution challenges—II  351 accountable for revoking deviant transfers, initiating and finalizing disciplinary action against those responsible for them, and making public the action taken to report or even tabling it before the State Assembly. Unfortunately, though many departments across states now have a policy, its compliance is poor. Second, the entire process of actual transfers should be made transparent on a standardized, and preferably online, platform. The seniority list and duration of tenures of all officials should be formally available always on the public domain. The criteria for transfers should be simple and be clearly outlined and publicized and deviations should be minimized. At least 90% of all transfers should be done strictly only during the transfer season, except for those caused by retirements or resignations, in which case too they should follow clear guidelines. The process of internal re-​deployments within a district by local heads to meet various contingencies should also be governed by a similar transparent policy. Several elements of this have been adopted by a few states in the annual teachers’ transfer process with significant positive impact. As low-​hanging fruit, adoption of this, with even a reasonable level of adherence in implementation, can result in a step-​change from the opaque processes in many states for large categories of employees. Third, stability of tenure, by itself, creates accountability. It takes a few months for any official, howsoever experienced, to become familiar with his or her operational jurisdiction. When officials are transferred once a few months, they have no incentive to plan and implement programs. More importantly, it becomes easy to shift blame to their predecessors or to extraneous factors. If an official has been in place for at least two years, he cannot either feign ignorance or evade responsibility about problems in his jurisdiction. In this context, both the Supreme Court and the Administrative Reforms Commissions have repeatedly recommended that governments formulate a policy to ensure the stability of tenure. Its implementation would significantly enhance administrative efficiency.

14.2 Procurement The procurement of goods and services by government agencies is the other important source of governance failures. Procurements range from

352  Possible Solutions students’ hostel provisions and office supplies to outsourcing hospital cleanliness and awarding large engineering contracts. As with personnel deployment, political leaders are often attracted by the patronage and rent-​seeking opportunities that come with favouring one supplier over another. They enlist the services of collaborators within the bureaucracy to game the procurement process. Poor quality goods and services are procured. Efficient and superior contractors and service providers get crowded out. The system gets entrapped in a vicious spiral of inefficiency and squandering of public funds. Procurement differs from personnel management inasmuch as there are often serious inefficiencies even when there is no political interference or corruption. However, corruption and political interference in procurement are arguably the more serious concern and hence discussed first.

14.2.1  The Rule Framework The General Financial Rules are the primary source of rules on procurement at the Union level in India. They are ‘a compilation of rules and orders of the Government of India to be followed by all when dealing with matters involving public finance.’1 They are executive instructions that are binding on all central government departments, as well as attached and subordinate bodies. They do not have statutory backing but, having been first issued in 1947, have the authority of more than 70 years of custom and practice and are equivalent to delegated legislation. Violation of the General Financial Rules (GFRs) is taken seriously, especially by the CAG but also by anti-​corruption agencies, and conscious or outright violation of the GFRs is extremely rare. The GFRs do not apply to Central Public Sector Units (CPSUs), which make their own procurement policies. However, these policies are often aligned in their essential aspects with the GFR. This is for two reasons. Firstly, government officials are invariably represented on the boards of public undertakings and they tend to encourage alignment with the government’s own rules on the assumption (not always valid) that the government’s time-​ tested rules are the best. Secondly, the Central Vigilance Commission does 1 Ministry of Finance, Department of Expenditure, General Financial Rules 2017, New Delhi, 2017.

Addressing execution challenges—II  353 have jurisdiction over government companies and corporations; its Chief Technical Examiner and his staff are generally drawn from the ranks of central government departments such as the Central Public Works Department, and therefore tend to view the GFRs as a benchmark. State governments have their own financial rules, sometimes referred to as ‘financial codes’.2 These are similar to the GFR, though, until the revision of the GFRs in 2017, some of them were arguably more advanced in their coverage of procurement than the GFR.

14.2.2  Can Corruption Be Reduced by Legal/​Procedural Reform? Commentary on public procurement from those outside government usually focusses on procedural reform. One of the suggestions made very often is the introduction by the government of procurement legislation.3 Compared to a mere executive instruction, a law could theoretically have the following advantages: • A law could provide for, and make mandatory, the following of principles and procedures for fairness and transparency. • By putting those principles and procedures into a law, it becomes more difficult for government to circumvent; rules like the GFRs can theoretically be relaxed through ministerial or cabinet approval (and this may not be possible with a law except to the extent the law itself allows it). • A law could provide explicit remedies to aggrieved tenderers who feel the procurement process has been unfair. The United Nations Commission for International Trade Law (UNCITRAL) model provides for a challenging procedure. Procurement laws may also provide for a ‘standstill’ period for representations by losing tenderers. • A law could provide for penalties for violation, thereby deterring wrongdoing. 2 See, for instance, the Tamil Nadu Financial Code, partially available at (http://​www.tn.gov. in/​docume​nts/​dept/​9/​2010-​2011) (Accessed on 14 August 2021). 3 This section draws substantially from Bibek Debroy and T.V. Somanathan, ‘Procurement Principles—​Statute or Process?’ in On the Trail of the Black, Rupa, New Delhi, 2017.

354  Possible Solutions On a close and rigorous examination, it is not clear that these benefits hold up in the Indian context. Judicial interpretation in India has already made it essential to follow the basic principles of fairness which are part of, say, the UNCITRAL model law. Indian courts have unambiguously held that government cannot be arbitrary in its procurement decisions and needs to follow principles of fairness. The GFRs too require the observation of these principles. There is little evidence that government officers or departments deliberately flout the GFR. The Indian Constitution already provides a writ remedy—​where citizens including tenderers can take the government to court even where there is no contractual right—​and the courts have widened it by interpretation to cover all possible scenarios. Unlike many of the countries where such laws have been enacted or proposed, Indian courts are very activist, do not go by the strict wording of the law, and readily entertain challenges even if there is no explicit provision therefor. Many tenderers can and do challenge tender awards and even obtain stays on the tendering process. In Indian conditions, explicit provisions of new remedies in addition to the existing ones are likely to delay the completion of procurement and encourage vexatious litigation. Deterrent penalties for procurement irregularities or failing to follow procedure (even the procedure contained in executive instructions) are already present in practice because India’s anti-​corruption laws are very strong and there is little a new law can usefully add. Indeed, it has been credibly argued that India’s anti-​corruption law was counter-​productive because it punished even honest decisions if, by hindsight, it could be shown that a tenderer or other private party gained by a decision.4 In short, in a country like India, where judicial review and anti-​corruption laws are already very strong, the marginal benefit of a procurement law in the respects of mandating principles of fairness, providing remedies, and prescribing penalties is limited or non-​existent. Of the four benefits mentioned above, the only real advantage that remains to be considered is the second one: that unlike executive orders such as the GFR, a law can bind not only lower officials but even the highest levels of government. In examining this further, it is necessary to 4 See for instance Box 2.2 in the Economic Survey 2015–​16, Ministry of Finance, Government of India, pp. 48–​49.

Addressing execution challenges—II  355 distinguish between parts of a ‘law’ that can be modified through delegated legislation and those which can only be modified with the approval of the legislature. To the extent the ‘law’ is contained in delegated legislation (most of the US procurement law is in this form), it is almost as easy for the government of the day to modify it as it is to issue executive orders. However, delegated legislation has to be placed in Parliament whereas mere executive orders do not. This means, both in theory and in practice, that there is a slightly but distinctly greater level of rigidity attached to rules. Portions of the law that require legislative approval are of course difficult for the executive to modify in the short run. Greater rigidity, i.e. reduced ability of government to relax procedure, is the only real difference, which a procurement law would have in India over executive instructions. However, this very rigidity can be a cause of inefficiency. As it is, the requirements of transparency and fairness put the government at a disadvantage versus a private buyer in a game-​theoretic sense. Information is power. The government is not allowed to conceal information and, under the Right to Information Act, can be compelled to disclose everything it knows sooner or later. It must follow a pre-​set procedure. It cannot act unfairly. A private buyer, on the other hand, can simply refuse to deal with a vendor it dislikes for reasons of past bad performance in order to induce a better price or indeed on a mere whim and is not required to state reasons. A private vendor to the government is not required to disclose anything beyond what is asked for and is not expected to be transparent about its costs. Nor is it, as a private citizen, held to any norms of fairness. A vendor to government can litigate repeatedly without becoming ineligible for future procurement. Poor performance is very difficult to establish to a judicial standard of proof and judicial processes often take years. If public administration were honest and competent and there was no risk of corruption (and if the notion that ‘fairness as a good thing in itself ’ is ignored), transparent and fair public procurement procedures could actually end up costing the taxpayer money by weakening the bargaining power of the taxpayer’s agents. Greater rigidity in procedure (the one true benefit of a law over instructions) can worsen this competitive disadvantage. It would raise the bar in terms of government’s obligations to show adherence to procedure. While price is quantifiable, quality differences between suppliers often cannot be ‘proven’ objectively to a judicial standard; suppliers often

356  Possible Solutions produce quality ‘good enough for government.’ Substantiating poor or delayed execution to a judicially acceptable standard is also difficult, and therefore suppliers feel they can provide bad service without suffering in future procurement. Rigid rules can create difficulty in responding to changing circumstances or special situations. It was the absence of rules that enabled President Thomas Jefferson to purchase in 1803 (on a negotiated basis with no ‘transparency’) 828,000 sq. miles of land from France (the so-​called Louisiana Purchase), doubling the size of the then United States, a decision which was questioned at the time. The absence of a law enabled the Government of India to respond within days to the procurement challenges of the Covid-​19 lockdown in March 2020, through executive instructions. Thus while rigidity may reduce the scope for abuse, it also can be a constraint on acting in the best interest of the public in genuine cases. If, to avoid rigidity, a provision for relaxation is included in the law, it allows for genuine cases but also lends itself to misuse for an improper reason and gets one back to the situation with executive instructions—​a paradox which cannot be easily resolved. Moreover, sophisticated forms of corruption are based on manipulation without procedural deviation and can co-​exist with a good law. Few of the instances of corruption in procurement have involved violations of procedure. There is no evidence—​empirical or anecdotal—​that corruption in procurement in Tamil Nadu or Karnataka (which have procurement laws) is significantly less than at the central level or in states without a procurement law. In practice, therefore, the existence of a procurement law is neither a necessary nor a sufficient antidote to corruption in India. In India, where the existing rules are quite sophisticated, corruption co-​ exists with procedural sophistication and typically occurs without violation of the rules. It is sometimes argued that there are certain other benefits that a procurement law might bring, which are not related to the issue of corruption. To the extent legislation introduces new and modern forms of procurement not common in the country, it can have a beneficial impact on improving the knowledge base and acceptability of these new forms. For example, the use of framework agreements—​well known in India for decades as ‘rate contracts’—​is not as widespread as it should be. Till recently, the GFRs did not have explicit provisions for electronic tendering.

Addressing execution challenges—II  357 The revision of the UNCITRAL model law was mainly done to allow for modern electronic means of procurement. New procedures could well be introduced through executive instructions; a law may have some advantage insofar as it reduces the likelihood of a new procedure being challenged in the courts. A major problem in Indian public procurement is the inability to exclude bidders who are known to have performed badly in the past (especially in terms of delay) or who are unduly litigant. This allows poor performers to go unpunished and continue to win new tenders. Theoretically, executive instructions could be issued to establish a system to rate vendor performance and exclude bad suppliers. Nevertheless, an explicit statutory provision for this may make it easier to enforce such a system and reduce the likelihood of judicial challenge. Many high-​profile scandals or controversies do not relate to procurement—​they relate to the sale or disposal of valuable rights by government. The use of the term ‘procurement’ in the title of most procurement laws, and the definitions of procurement given, may lead to judicial challenge on the applicability of a procurement law to disposals, especially of abstract rights. Some east African countries (under UNDP guidance—​ Uganda, Kenya, and Tanzania) have covered ‘disposals’ also under the law but the provisions are sketchy. The Government of India appointed a Committee on the Allocation of Natural Resources, which recommended a series of measures for fairness and transparency in natural resource allocation.5 The Supreme Court has also clarified the broad principles applicable to the allocation of certain types of natural resources.6 The court recognized that different circumstances would require different approaches. A general law relating to disposal would be extremely difficult to draft and implement unless it were at the level of very general principles, in which case its anti-​corruption effect would be insignificant. In short, in Indian conditions, a procurement law is unlikely to produce any major impact on probity and, unless kept at the level of broad 5 Report of the Committee on Allocation of Natural Resources, Cabinet Secretariat, Government of India, May 2011. 6 Supreme Court of India, Natural Resources Allocation, In re, Special Reference No.1 of 2012; (2012) 10 SCC 1.

358  Possible Solutions principles, can hamper efficiency. Its effects on sophisticated forms of corruption are likely to be negligible; on corruption, the main effect is thus cosmetic. Further, a procurement law will not necessarily address issues relating to the disposal or sale of tangible or intangible rights. Procurement and disposals are core executive functions; they require discretion in the proper, judicially approved, sense of careful application of mind to relevant criteria in public interest. These functions cannot be reduced to rigid rules, especially in the midst of rapid technical change, without seriously impairing the efficiency of governance. Introducing rigidity will not solve the problem of probity but will create new sources of delay and buck-​passing in an already risk-​averse and slow system. Going beyond the issue of procurement law, a more general point is that reducing corruption in procurement through procedural rigidity is theoretically appealing but practically both difficult and not effective. Many well-​intentioned procedural solutions with the aim of curbing corruption have ended up with no reduction in corruption but an increase in rigidity, thus resulting in a lose-​lose solution! Most Indian governments already have good procurement procedures in terms of form—​but the substance may be different. For instance, open electronic tendering and a norm of accepting the lowest responsive bid are almost universal. But there may be informal restrictions on who actually bids. This is particularly true of works contracts. Bidders feel that if they bid against the wishes of, say, the Minister in charge of the Department, then even if they win, they may face serious problems in contract execution. On the other hand, some projects are generally considered to be successful and cleanly and well executed, like the first phases of the Delhi Metro under Mr. E. Sreedharan witnessed procedural violations which—​ had they occurred in a different organization—​might have attracted odium. Thus rigid procedures may end up obstructing honest officers trying to be innovative while allowing corrupt officers to make money by appearing to strictly follow procedure. In short the core integrity problems that affect procurement in government are not because the rules of the game are merely rules and not a law; they will not be solved by legislation. Legislation may reduce efficiency without reducing corruption. There are no easy procedural solutions to the integrity aspect of procurement.

Addressing execution challenges—II  359

14.2.3  Efficiency Issues The efficiency issues in procurement are primarily due to the absence of clear procedural guidelines for procurements that fall outside the traditional 19th and 20th century tendering framework of ‘goods’ or ‘public works’. The general norm for procurement is what is commonly known as ‘L1’, i.e. choosing the lowest quotation or the lowest tender or bid. There has been much criticism of the ‘L1 culture’ whereby the lowest quotation or bid is accepted without commensurate attention to quality which is more difficult to specify or assess.

360  Possible Solutions Till recently, the GFRs had many gaps. For instance: i) They did not adequately provide for the procurement of services where non-​price considerations are crucial and where ‘quality’ is not a pass-​fail test but something to be optimized; ii) They did not contain provisions dealing with electronic means of procurement; iii) They were not adequately supported by practical guidance or ‘good practice notes’ that could enable procuring personnel to apply and implement broad principles correctly and consistently when dealing with complex projects in a modern economy.

States like Tamil Nadu and Karnataka, where there are procurement laws, had an advantage in this respect; this is because many of the ambiguities in the Central GFRs had been made clear by the rules under these laws, since those rules were issued much more recently than the GFRs. (However, this beneficial impact did not stem from their statutory backing—​it arose because they had updated their practices.) The central government has rectified the weaknesses in procurement policy and practice. The GFRs were thoroughly revised and re-​issued in 2017 with greater clarity and with explicit provisions for modern methods of procurement. They were accompanied by both a Manual for the Procurement of Goods7 and a Manual for Procurement of Consultancy and Other Services.8 However, there is always scope for further improvement. Some areas of potential improvement are discussed below.

14.2.4  Possible Improvements in the Procurement Procedure 14.2.4.1 Reducing Procurement Delay Through Indecision Currently, there is a natural predilection to delay or defer a procurement at the first sign of a problem. The implicit and sometimes unrecognized 7 Ministry of Finance, Department of Expenditure, Manual for Procurement of Good 2017, Government of India, New Delhi, 2017. 8 Ministry of Finance, Department of Expenditure, Manual for Procurement of Consultancy & Other Services, Government of India, New Delhi, 2017.

Addressing execution challenges—II  361 assumption is that doing a fresh procurement will improve matters; even if the assumption is recognized to be untrue by the executing agency, the fact that it is the standard assumption of auditors and Vigilance authorities deters decision-​makers from doing what is right. The reality is that delay incurs costs, too: first, the service is not provided and the public suffers thereby; second, there may be lost revenues; and third the cost of the project is likely to increase with inflation (though this may not apply to equipment where prices often fall). The GFRs should provide that if the decision takers decide to re-​tender despite procedurally compliant bids being available, they should explicitly consider the costs of non-​acceptance and resulting delay and record reasons. Auditors and vigilance authorities need to issue internal guidelines recognizing these factors. Similarly, there is a lack of clear guidance on how to proceed when a civil contractor abandons a multi-​year contract midway for reasons of solvency or any other causes. Clear guidelines allowing a very quick method of procurement in such situations—​e.g. by entrusting the work on a single-​source basis or on a limited tender—​would help a lot in speeding up many stalled public projects. 14.2.4.2 Reducing Disputes Arising From Indecision Complex construction projects awarded by government agencies often involve risks that are not fully under the control of either party—​delay in land acquisition, interim stay orders of courts, delays in issuing permits by various statutory authorities like Pollution Control or other environmental authorities. In many situations, there are also some delays by the contractor. The tendency in government is to always blame the contractor and not recognize the portion of the delay attributable to factors not in the government’s control. Partly this stemmed from the fear of the old (since repealed) Section 13(1) (d) (iii) of the Prevention of Corruption Act 1988 that could result in an officer being penalized for any pecuniary gain to a private party if it was not in public interest. Proving that it is in public interest to pay more to a contractor than was originally agreed in the contract could be very difficult. Therefore whenever anything needs to be paid to a contractor that is not explicitly in the contract, civil servants often feel it is safer to raise a dispute and allow some other forum to decide. And if appeals from that forum are possible, it is always safer to appeal until no

362  Possible Solutions further appeal is possible. This leads to an enormous number of unnecessary disputes where the government side knows there is substantial truth in the contractor’s case but will still dispute the claim for years until it is decided by the Supreme Court or at least the High Court. The delay also has the ‘benefit’ that the initial decision-​makers might have retired, given the slow speed of judicial proceedings in India. There is a need for clear mandatory rules and guidelines on settling such cases and for creating incentives and disincentives that encourage officers to settle genuine claims rather than litigate them. 14.2.4.3 Fixed Cost Selection The Fixed Cost Selection (FCS) method needs to be introduced in public procurement for certain kinds of services and possibly some kinds of simple works. Under this method, the tendering authority (perhaps with the aid of specialized consultants) prepares an estimate (including milestones and timing of payment) of what would be the reasonable cost. That then becomes the fixed cost of the contract. Bids are invited to disclose the fixed cost and evaluated purely on their technical merits without discussion or evaluation of price because the price is fixed in advance. Regardless of which firm is selected, the amount payable remains the same. A margin of, say, 10% for physical contingencies/​additional work arising during execution may be allowed. The method is suitable for repetitive types of work that can be quantified well; an example is the preparation of detailed project reports for, say, sewerage projects in 20 different towns of approximately similar size. It eliminates the complications of price comparison. The scope for corruption through pricing changes is eliminated and the decision maker cannot cause a financial loss or gain to the government. Thus the selection is free from fear of allegations of causing ‘avoidable loss’ or ‘infructuous expenditure’ from the audit angle or ‘wrongful loss without public interest’ from the anti-​corruption angle. The only remaining issue is fairness among the bidders, which is less susceptible to controversy when no financial loss is involved. 14.2.4.4 Rate Contracts A variation on the theme of fixed cost selection is the rate contract (also known as ‘framework agreement’) which applies to the purchase of

Addressing execution challenges—II  363 goods. Rate contracts provide a very useful service by allowing purchase by government departments of various things at a fixed price without tenders. They can be very useful in things like computers, scanners, copiers, phones, and office supplies and in many similar cases. Rate contracts in goods are no longer available since the winding up of the Directorate General of Supplies and Disposals (DGS&D). They need to be re-​introduced. They can play a useful role not only in goods but even in services. ‘Classes’ of labour could be notified along with monthly/​ hourly/​daily rates for each kind. Even higher-​level consultancy services can be incorporated in this by creating a separate classification. 14.2.4.5 Centralized Earnest Money A system could be introduced where service providers can register with one central authority, get classified by size or technical area, present/​update their credentials, pay an annual earnest money deposit (EMD), and then be eligible to submit bids to any government tender (state or central) without requirements of either verification of their qualifications or payment of EMD. Software companies, consultancy firms, and so on could be divided into, say, three or four classes for each category based on their size, turnover, and so on, and pay a lumpsum EMD reflecting a reasonable security amount for the number of bids they may be simultaneously involved in at one time. Registered contractors should be automatically qualified for all contracts of the relevant class/​size without any separate verification of their qualification criteria or payment of EMD. Only unregistered tenderers would go through detailed scrutiny of eligibility and payment of EMD. The main benefits of such a system would be speed (shortens or eliminates verification stage), reduced overheads for government (less work to verify credentials, preserve and refund EMDs), reduced costs for bidders (who need not lock up money in EMDs in multiple projects all over the country and chase refunds), and consequently lower prices for the government. 14.2.4.6 Shared Service Centres or Agencies Apart from the absence of adequate guidelines, another major problem is the lack of procurement expertise, especially in the non-​engineering

364  Possible Solutions departments where procurement is done only intermittently. A remedy sometimes suggested in this regard is the establishment of common shared services centres (SSCs) to support procurements of several departments. A UK National Audit Office (NAO) report summarizes the case for SSCs,9 The principles of reducing costs through using shared services are straightforward and widely understood. They combine two key elements. One element is to standardise processes and services so that they can be provided in a consistent and repeatable way, in high volumes, by a single provider. This often involves moving to a common IT system (operating platform). The other element is to outsource operations to an organisation that can specialise in providing a service and, through economies of scale, can offer the service at a lower cost.

But shared services involve some trade-​offs. Individual processes offer the convenience and advantages of choice and flexibility for departments, which help them customize solutions to their requirements. Standardized services offer cost-​efficiency, professionalism, and data analytics possibilities. It can be argued that this trade-​off is a serious issue only in a few cases, and the vast majority of procurements of departments are standardizable. Only in the case of a handful of departments are these likely to be a major challenge. However, its practical experience has not been so encouraging. In 2014, the UK Government established two independent SSCs to provide back-​office functions for fourteen departments and their agencies. The centres were outsourced to two private providers. However, due to poor implementation, the assessment of its performance by the NAO has not been satisfactory. In particular, the report found a lack of clarity in procurement standards and processes, poor management of the standardization process including failure to make evident the business case for SSCs, and delays in the integration and migration of internal departmental back-​office workflows into that of the SSC providers.

9 National Audit Office, Shared Service Centres, 20 May 2016 (https://​www.nao.org.uk/​wp-​ cont​ent/​uplo​ads/​2016/​05/​Sha​red-​servi​ces-​cent​res.pdf).

Addressing execution challenges—II  365 In India, a similar example is an experience with the DGS&D, a former central government agency that was responsible for centralized procurements of certain categories of goods. It came to be associated with rampant corruption and poor quality of service, which left client departments most often deeply dissatisfied. It was disbanded in 2017. Given the financial stakes involved with centralized procurements, the risks of capture by vested interests are very high. The peculiar political, bureaucratic, and judicial culture of the Indian administrative system only increases the likelihood. Another example is that government departments are often compulsorily required to get building construction done by the Public Works Department; there are often delays and the user department often complains of poor quality. While the compulsory use of a shared service agency is usually not desirable, giving non-​specialist departments a choice to hand over procurement to specialized procurement agencies is a good idea because it gives departments better control and makes the specialist agency accountable. This system already exists in some form of works contracts: in certain circumstances, central departments can entrust civil work to either the Central Public Works Department or to a few other designated agencies like the National Building Construction Corporation, who then becomes responsible for the procurement process. Similarly there is a concept of ‘deposit works’, where one agency deposits money with another agency which then does the work of procurement and construction. Some deposit works are not voluntary—​such as when a road crosses the railway and the railway administration insists on a deposit work to be executed by the railways for safety reasons. Even in voluntary deposit works, the disadvantage is the large administrative fee typically charged by the agency taking on the work, often called a ‘centage charge’, sometimes amounting to over 15%. A rationalization of centage charges and a better system of accountability for these agencies to finish work on time and on cost is necessary, and if this is done this method holds considerable promise. It will enable procurement to be done by specialized agencies with experience and expertise. The ability to choose a specialized agency can also be very useful in disposals—​the Metal Scrap Trading Corporation is often used by departments to conduct auctions on their behalf. A similar agency (preferably

366  Possible Solutions more than one) for land disposal would be very helpful in helping governments dispose of unwanted public land. Another promising approach—​where there is a centralized transparent platform but decisions are de-​centralized—​is more suitable. The central government’s Government Electronic Marketplace (GeM) for goods and services is a huge step in this direction, which has already achieved great success. Government agencies everywhere should be mandated to shift procurement of most categories of goods and services to the GeM portal. The coverage of service contracts offered by GeM could be expanded to include all the basic types of service contracts required by public agencies. This could realize many of the benefits of SSCs without the vulnerabilities of capture.

14.2.5  To Tighten or to Loosen? A careful reading of the two preceding parts—​on integrity and efficiency, respectively—​will reveal a paradox. When it comes to increasing integrity, most commentaries tend towards asking for tighter rules, enactment of a law, reducing discretion, and so on. Yet, when it comes to improving efficiency, expert commentary is usually about loosening or simplifying the rule framework, more flexibility, increasing discretion, etc. The authors’ own belief is that, in India, the balance of advantage in this trade-​ off currently lies towards doing what increases efficiency, since the rule framework is already very tight and does not, by itself, prevent a lack of integrity.

14.3  Permits and Licences After transfers and procurements, the biggest source of corruption is permissions and clearances. Such approvals predominantly involve the environment, land, industries, and urban local governments. One category is permissions involving land transaction, use, and development. Another involves approvals to start industries and businesses. A third involves safeguards against negative environmental externalities.

Addressing execution challenges—II  367 The financial stakes associated with these approvals are often very high. Once delayed, commercial property developers and industrialists, with large investments, start to feel the pinch of rising capital costs. Further, in the case of environmental approvals, where safeguard requirements and standards have increased, applicants often try to cut corners with compliance and get their permissions. Factories awaiting consent for operations from the Factory Inspectorate place a high premium on the speed of its approval. Officials and politicians understand the stakes and may seek to maximize rents. In several cases, a full-​fledged ecosystem has emerged around permissions and clearances where significant amounts change hands. E-​Governance offers promise in addressing this problem. It is possible to increase convenience and transparency and reduce the corruption associated with these approvals by workflow automation. The entire process from application and payments to processing and approvals could be digitized and accessed through a portal. The documentation and other requirements and inspections could be standardized to limit discretion. The delays and rejections could be scrutinized closely at higher levels to identify problems and improve the process. This could be institutionalized through some form of a performance management system that holds officials at different levels accountable for the workflow approvals process. Finally, the official-​wise exception and delay reports could be publicly disclosed each year. This could turn out to be among the most powerful ease of doing business reform that is possible. All these reform suggestions demand both care in design and diligence in implementation over a reasonable period of time before they stabilize and become entrenched. The danger is one of notional adoption by way of poor implementation, a very strong likelihood. This entraps the system in a bad equilibrium, where the officials claim the reform has been implemented, whereas its lagged outcomes will invariably fall far short of expectations. However, ultimately these solutions will not work effectively unless the political executive actually wishes to see a clean and efficient process.

368  Possible Solutions

14.4  The Real Solution In the previous sections on the 3 Ps, a distinction was sometimes drawn between issues of corruption or political interference on the one hand and issues of efficiency on the other. Many of the solutions suggested, especially in the section on procurement, were on the issues of efficiency. On the issue of corruption in procurement, it was suggested that procedural or legal reform is not the solution. On permits and personnel too, it was emphasized that true improvements would not happen without a genuine commitment to clean administration. Thus, the reader may well feel dissatisfied that, for the key problems of corruption and political interference, no solutions were suggested! It is the view of the authors that when it comes to core government functions that cannot be simply deregulated, procedural reform (including e-​governance) can help but is not a true solution. The best and most lasting solution to the problems of corruption—​in personnel, procurement, and permissions—​is the most difficult: an improvement in political and civil service ethics (in that order of importance) whereby the top political executive ceases to expect bribes from these activities and the large majority among civil servants and the becomes honest. Top-​level political commitment is the sine qua non for the reduction of ‘grand corruption’—​corruption involving large or recurrent bribes on big contracts, licenses, and personnel decisions. Corruption levels are obviously not objectively measurable, but there are clear examples to illustrate that grand corruption can decline suddenly with political will. The improvement in the integrity of all three functions within the central government after 2014 is reflected in public and civil service perceptions and in various published indices. There are examples from the states too. For instance, civil service and public perceptions indicate that grand corruption declined sharply in the 1996–​2000 period in Tamil Nadu, after the 1996 election was fought on the plank of reducing corruption. In Kerala, the government in the early 1970s is reputed to have been exceptionally clean. On the other hand, ‘petty corruption’ (small amounts of money collected for routine transactions like driving licences or tax refunds) depends more on the level of ethics of the civil service. Petty corruption

Addressing execution challenges—II  369 declines when dishonesty becomes an exception and an unacceptable deviation from the norm. How to achieve such a systemic transformation is something which, to say the least, is far beyond the capability of the authors and of this book. Nevertheless, it is useful to remember that such a transformation would produce better results than the many ideas discussed here.

15 Addressing Execution Challenges—​III This chapter deals with four ideas that have become popular as possible solutions to state capability problems in execution: outsourcing and privatization, audits and certification, electronic governance and digitization, and rigorous evaluation. It concludes with some thoughts on how to actually initiate systemic change.

15.1  Outsourcing and PPPs Under the rubric of public–​private partnerships (PPPs), a wide array of government services have come to be outsourced to private providers. They range from various office management services (cleaning, catering, drivers, etc.) to utility services (maintenance of facilities, management of vehicle fleets, etc.). Outsourcing ranges from operation and maintenance contracts to outright purchase of the required service without public ownership of the assets. Technology solutions, including hardware, are now commonly procured as service contracts. Some other services have been fully privatized and are now also delivered privately. For example, telecommunications, electricity generation, mass transit, etc., are now also delivered as purely privately provided services. In some others, like water and sewerage or solid waste management, the government chips in with a subsidy provision to make access affordable. A report by the United Kingdom’s Institute for Government talks about three pre-​requisites for private contracting1

1 Tom Gash and Nehal Panchamia, ‘When to Contract’, Institute for Government (https://​ www.ins​titu​tefo​rgov​ernm​ent.org.uk/​sites/​defa​ult/​files/​publi​cati​ons/​When_​to_​c​ontr​act.pdf) (Accessed on 8 August 2021). State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0015

372  Possible Solutions a) existence of a competitive market of high-​quality suppliers, without which government opens itself to supplier opportunism and risks poor service performance; b) ease of measuring the value added by the provider, without which it is difficult to write and price contracts and monitor performance; and c) the service not being so integral to the nature of government as to make outsourcing inappropriate, for instance, when a supplier has to make key policy decisions. There have been some very good successes with public services outsourcing in India. The shift away from managing government drivers and vehicles to outsourcing the entire operations to private vehicle operators is an example. It does away with the costs associated with owning and maintaining vehicles, employing drivers, and managing their use. Instead, for a fixed fee, the required service gets delivered. And, best of all, it avoids governments making the large upfront investments required. The arguments in favour of PPP that have the most conceptual merit are: • a well-​designed outsourcing arrangement aligns the incentives of all sides to deliver better outcomes and removes incentive distortions within governments; • decision-​making flexibility and the incentives generated by the profit motive help make more efficient use of existing resources and thereby provides value for money when compared to the public sector production and provisioning. Another reason often brought up—​for example, in the case of road projects under the annuity model—​is that outsourcing helps governments overcome fiscal constraints by obviating large upfront investments. This argument has less merit—​the cost of capital for the private sector is almost always higher than for sovereign or sub-​sovereign governments. Therefore, outsourcing purely as a means of reducing the fiscal deficit is nothing but off-​budget financing of the fiscal deficit. Then there is the peculiarly Indian case for outsourcing2: 2 T.V. Somanathan and Gulzar Natarajan, ‘Partnership Lessons’, The Indian Express, 30 July 2019 (https://​indian​expr​ess.com/​arti​cle/​opin​ion/​colu​mns/​inf​rast​ruct​ure-​sec​tor-​pub​lic-​priv​

Addressing execution challenges—III  373 Political interference in recruitment, competitive trade union activity (witness the posters in every railway station), rigidities on salaries and writs in courts on service matters, reduce the efficiency of personnel management in the public sector. Activities of oversight agencies —​ Vigilance, Comptroller and Auditor General etc —​cause extreme risk aversion in decision taking, reducing efficiency of procurement and operational decisions.

This was discussed in Chapter 6. Given the problem of judicial intervention in relatively minor and routine personnel decisions and the extraordinarily strong judicial protection to government and quasi-​government servants on even minor conditions of service, outsourcing often has big efficiency advantages in labour-​intensive activities, arising purely from the absence of public ownership. In other words, private provision may be better not because the private provider is more efficient but because it is (by definition) not subject to the peculiarly onerous personnel management constraints of the Indian public sector. Therefore, quite apart from any gains in risk-​apportionment, etc., there are simple and obvious productivity gains from outsourcing—​simply in terms of enforcing normal work discipline and output norms. An example is the operations and maintenance of metro railways. If it is done through in-​house staff, the rules of the public sector and the enforcement thereof by the courts make it very difficult to maintain productivity and discipline. Personnel management—​ strict performance management, temporary or differentiated personnel hiring, redeployment of staff, periodic rationalization of work and process re-​engineering, etc.—​is difficult with in-​house staff. Attempts to do this often result in writs and litigation. The experience of the Chennai Metro Rail, for instance, shows clearly the benefits to the travelling public of outsourcing routine operations and maintenance while maintaining strategic control of routes, fares, frequency of operation, etc., in public hands. Thus given a choice between in-​house provisioning and outsourcing for certain labour-​intensive services, prudence often favours the latter. Outsourcing certainly has merits at a theoretical level and in practice too and often works effectively. However, there are also several situations ate-​partn​ersh​ips-​india-​union-​bud​get-​ppp-​model-​5861​815/​).

374  Possible Solutions where these claims do not hold. Take the example of outsourcing transportation vehicles—​water tankers and garbage transportation. Unlike with owned vehicles, there is the principal–​agent challenge of ensuring that claims are made on actual trips (and not fictitious ones), and those trips have realized the objectives (e.g. garbage transportation is full or water delivery is monitored). Municipalities have sought to address them through various means—​installing location tracking devices on the vehicles and monitoring them, using weighbridges, RFID tags, smart meters, and digital solutions to audit water supply, etc. But despite the best efforts, only a few municipalities have succeeded in stabilizing these systems enough to claim conclusively that outsourcing has delivered value for money compared to in-​house operations. Another Institute for Government report, which examined the experience of private contracting across 11 service categories over nearly four decades in the United Kingdom, also found four themes that were common across contracts that did not succeed3 a) a failure on the part of government to engage with the market early and be clear about what it is buying; b) an excessive focus on price and an insufficient assessment of quality in selecting bids; c) the transfer of risks to suppliers that they are unable to price or manage well; and d) contract managers lacking the skills to administer the contracts. In general, the challenges with outsourcing in the Indian context echo many of those encountered in the United Kingdom. For a start, credible and simple measurement of outcomes is a big challenge, especially in activities where quality is of importance. Second, technology solutions, for all their wonders, are still not reliable enough in such demanding conditions and either fail technically or are subverted. Third, even with technologies, reliable assessment of the outputs—​quality of service delivery and realization of objectives—​is difficult. Fourth, contractors often 3 Tom Sasse et al., Government Outsourcing: What Has Worked and What Needs Reform?, Institute for Government, September 2019 (https://​www.ins​titu​tefo​rgov​ernm​ent.org.uk/​sites/​ defa​ult/​files/​publi​cati​ons/​gov​ernm​ent-​outs​ourc​ing-​ref​orm-​WEB.pdf).

Addressing execution challenges—III  375 find weak links in the workflow and game the processes. In this, they find enough help from municipal staff. Fifth, given all the earlier three challenges, contract management and monitoring of implementation becomes critical. But here state capability weaknesses make it difficult. Sixth, there is the under-​appreciated problem of low-​ball bids, whereby less-​than-​competent contractors bid unviable rates and crowd out the better contractors to get the contract in the belief that they can game the system to make a good profit. Seventh, the displacement of the status quo does not come without resistance as entrenched vested interests fight back to sabotage the reform. Finally, like with all markets in their initial stages, and outsourcing in most sectors is fledgling, the supply-​side is constrained in its ability to deliver with high quality, at scale, and at reasonable rates. In cases where the outcomes cannot be reliably measured, outsourcing can pose serious principal–​agent problems. Services in which quality is paramount and is not observable are likely to struggle with effective outsourcing. In any case, where outsourcing is proposed, government, at an appropriate level (typically state government), should make available model tender and contract documents, which outline the eligibility requirements, scope of work and its terms of reference, process standardization, outcome metrics and measurement protocols, and other contract management elements. This should form the basis for outsourcing contracts. Most critically, even when outcomes are credibly contracted, monitoring and management of the contract is critical. The capability to manage contracts is perhaps one of the most important priorities that governments should strive to equip their personnel with in the coming years. The success of outsourcing is critically dependent on the existence of a competitive enough private market. If there are only one or two providers, they may form a cartel and exploit the government. In many cases, the outsourcing market itself may be nascent, and there is little that could be done to hasten its maturity beyond a point. However, one or two successful contracts may create new vendors. As a last point, and without getting into its details, the intense debate on the assessment of PPPs and outsourcing in developed countries4 is 4 Gulzar Natarajan, ‘PPPs and Value for Money?’, Urbanomics Blog, 6 February 2018 (https://​gulza​r05.blogs​pot.com/​2018/​02/​ppps-​and-​value-​for-​money.html) (Accessed on 24 December 2018).

376  Possible Solutions relevant. The United Kingdom, one of the leaders of the movement towards PPP, has, in recent years, witnessed a sharp turn of the tide.5 This has been sparked by a series of high-​profile exposures of poor quality and inefficient service delivery, violations of contractual terms, asset-​stripping, overcharging, and outright fraud and corruption among such private providers.6 A comprehensive National Audit Office assessment of the United Kingdom’s pioneering Private Finance Initiative (PFI) program has found that schools and hospitals built and managed as PPPs cost 40% and 60%, respectively, more than their public sector comparators.7 Accordingly, many of the outsourced public assets and services are now being restored into public ownership, management, and provisioning. In early 2018, one of the world’s largest outsourcing service providers, Carillion, filed for bankruptcy, forcing governments across the UK to take over all its activities.8 In his 2018 Budget speech, the UK Chancellor of Exchequer effectively declared the obituary of PFI, stating that9: I remain committed to the use of public-​private partnership where it delivers value for the taxpayer and genuinely transfers risk to the private sector. But there is compelling evidence that the Private Finance Initiative does neither . . . I have never signed off a PFI contract as Chancellor . . . and I can confirm today that I never will. I can announce that the Government will abolish the use of PFI and PF2 for future projects.

Much the same has been happening across continental Europe. In the United States too the costs of the inefficiencies and excesses from 5 Gulzar Natarajan, ‘Is the Tide Turning on PPPs?’, Urbanomics Blog, 15 August 2017 (https://​gulza​r05.blogs​pot.com/​2017/​08/​is-​tide-​turn​ing-​on-​ppps.html) (Accessed on 24 December 2018). 6 National Audit Office, UK, Memorandum for Parliament, ‘Role of Major Contractors in the Delivery of Public Services’, 12 November 2013 (https://​www.nao.org.uk/​wp-​cont​ent/​uplo​ads/​ 2013/​11/​10296-​001-​BOOK-​ES.pdf) (Accessed on 24 December 2018). 7 National Audit Office, UK, Report by the Comptroller and Auditor General, PFI and PFI2, 18 January 2018 (https://​www.nao.org.uk/​wp-​cont​ent/​uplo​ads/​2018/​01/​PFI-​and-​PF2.pdf) (Accessed on 24 December 2018). 8 Gulzar Natarajan, ‘Lessons from the Carillion Collapse’, Urbanomics Blog, 22 January 2018 (https://​gulza​r05.blogs​pot.com/​2018/​01/​less​ons-​from-​carill​ion-​colla​pse.html) (Accessed on 24 December 2018). 9 Philip Hammond, Budget 2018 Speech, 29 October 2018 (https://​www.gov.uk/​gov​ernm​ent/​ speec​hes/​bud​get-​2018-​phi​lip-​hammo​nds-​spe​ech) (Accessed on 24 December 2018).

Addressing execution challenges—III  377 high-​profile outsourcing and privatizations have repeatedly surfaced in recent times.10 The Institute for Government report found that outsourcing can succeed for waste collection, cleaning, catering, and maintenance of assets.11 It also found that outsourcing does not work for construction, clinical health care services, employment services, and adult social care. Its recommendation to governments was: They must outsource for the right reasons –​where private providers benefit from expertise, economies of scale or new technologies that enable them to deliver services better or more cheaply, or where competition can improve performance. [To this, in India, must be added the (peculiarly Indian) dimension already described, of advantages in personnel management and operational freedom.] They should not outsource where there is no reason to think it will work, as with probation, or in pursuit of unrealistic cost savings. They must develop better evidence to inform these decisions. When government chooses to outsource, it must do it better: understanding what it is buying, choosing bids that deliver sufficient quality and value for money, allocating risks intelligently and managing contracts effectively.

The clear lesson for India is that it does not make sense to outsource unless there are clear efficiency gains to be had.12 This may sound obvious, but it is important because governments often resort to outsourcing and PPPs to overcome the fiscal constraint of having to make up-​front investments, even though there are limited or no efficiency gains. Before tightly and unambiguously embracing PPPs across sectors, India needs to reflect on this history of outsourcing and privatizations in countries with some of the most mature markets and capable states. Governments should decide on the extent of outsourcing depending on the context and activity. It is possible that certain activities, while typically outsourced elsewhere, may

10 ‘Bottom Line Nation’, The New York Times (https://​www.nyti​mes.com/​ser​ies/​priv​ate-​equ​ ity-​bot​tom-​line-​nat​ion) (Accessed on 24 December 2018). 11 Sasse et al., Government Outsourcing: What Has Worked and What Needs Reform?. 12 T.V. Somanathan and Gulzar Natarajan, ‘Right Reasons to Get Hitched’, Indian Express, 2 August 2014 (https://​indian​expr​ess.com/​arti​cle/​opin​ion/​colu​mns/​right-​reas​ons-​to-​get-​hitc​ hed/​).

378  Possible Solutions have to largely continue in-​house because the market would not be able to support wholesale outsourcing. There may be others where for India-​ specific reasons, outsourcing may be desirable even if it is not common elsewhere. A selective and phased roll-​out of outsourcing is the most appropriate strategy given the need to develop both internal expertise and capability as well as let the market develop.

15.2  Audits, Ratings, and Certification Another trend that is gaining ground is the use of agencies to audit compliance to some specified standard. These objectives can include adherence to a set of processes or practices (say, a process certification), or acquisition of certain skills (say, an occupational licence), or compliance of physical facilities to certain standards (say, laboratory or college), or realization of certain outcomes (say, open defecation–​free (ODF)). Such validations can be done as a one-​off exercise or on an ongoing basis, depending on the thing—​process, service, product, and outcome—​ being certified. When done on a recurring basis, the certification is also referred to as an audit. When a government or semi-​government agency certifies an entity, it is also called an accreditation or an authorization to deliver specified services. Certification has been used in addressing development problems for a long time.13 In the early 2000s, there was an ISO certification boom, with government agencies and departments queuing up to have their offices certified. The process triggered efforts on documentation, standardization, service delivery levels, and so on, all important requirements to improve public service delivery. Third-​ party agencies have been certifying diagnostic laboratories and their equipment and technical college facilities. The National Accreditation Board for Hospitals and Healthcare Providers (NABH) and National Accreditation Board for Testing and Calibration of Laboratories (NABL) are two examples. Similarly, industry-​managed organizations 13 Jose Luis Guasch et al., Quality Systems and Standards for a Competitive Edge, World Bank, Washington, 2007 (https://​sitere​sour​ces.worldb​ank.org/​EXTEX​PCOM​NET/​Resour​ces/​2463​ 593-​121388​7855​468/​69_​LAC_​Quality​_​and​_​Sta​ndar​ds_​P​ub_​N​ov_​2​007.pdf).

Addressing execution challenges—III  379 have been responsible for occupational licensing, like those of town planners and building inspectors, auditors and chartered accountants, and so on. Another early use was in third-​party quality control checks of engineering works. The objective was to get an idea of the quality of construction, independent of that of the departmental engineers. In terms of public policy innovations, it should count as a success. There are numerous third-​party quality audit providers in each state. In many states, such agencies today audit engineering works done at all levels of government. Third-​party audits have since been replicated elsewhere. For example, pollution control boards have come to use it to verify effluent emissions by factories. In recent times, certification agencies have come to be used for a broader set of activities. Accordingly, quality of facilities in teacher training institutes, open defecation, cleanliness and hygiene, adherence of Industrial Training Institutes (ITIs) and Vocational Training Institutes (VTIs) to standards, etc., are all now certified by third-​party entities. The Quality Council of India (QCI), an entity under the Department of Industrial Policy and Promotion (DIPP), has been at the forefront of this certification push. In recent years, governments have initiated challenge competitions on various issues, where the agencies—​local governments, districts, public sector agencies, etc.—​declare realization of some defined objective and the same is certified independently. The Swachh Bharat (or Clean India) Mission has been a pioneer in this regard with its ODF certification process. For governments struggling with the delivery of services to specified standards or realizing various development outcomes, such certifications can be very appealing. Supporters point to the compelling argument that certification provides an independent assessment of the activity. It, therefore, aligns with the incentives of all concerned. While theoretically unexceptionable, there are challenges. The certification is credible only when the assessment metrics are accurate and reliable indicators of the success of the activity being assessed. Secondly, the government agency concerned must have the capacity and resources to meet the certification standards. The corollary is that the standards have

380  Possible Solutions to be realistic for the context. Thirdly, the certification agency must be honest with the capture and reporting of the information and also have the expertise and capability to deliver on the contract. These assumptions are often questionable. It is often virtually impossible to get accurate and reliable metrics that truly measure the attainment of the intended outcome. The standards may be (or felt to be) unreasonable, as is argued to be the case with benchmark learning levels for each grade in school education.14 Further, certification agencies are susceptible to being captured by vested interests and the. Its likelihood increases as the stakes associated with the certification increase—​for instance, certification of a laboratory in a private college which could make or break the college’s ability to take students. There is the issue of the supply side. Given the nascent nature of the market, it is most likely that there are only a few providers, and even among them just one or two have sufficient capability to deliver effectively. It is not as simple as ‘create the demand and supply will take care of itself ’. Despite all these limitations, when corruption in existing inspection arrangements is common, certification by private agencies offers considerable advantages. Certification, even if imperfect, may not be worse than a corrupt official inspection system and may be considerably easier to do business. It is important that the service being identified is amenable to credible certification. It should be possible to accurately and reliably measure and capture the realization of the true objectives. Typically, services where quality is the paramount consideration struggle to meet this requirement. Physical facilities, logistics activities, etc., are credibly certifiable. There is no alternative to rigorous monitoring of the work of certification service providers. Governments need to develop the capability for this.

15.3  Digitization/​E-​Governance The most promising and attractive idea of our times is arguably digital technology. In the last decade or so, a whole new world revolving around 14 Lant Pritchett, The Rebirth of Education, Center for Global Development, Washington, 2013 (https://​www.cgdev.org/​sites/​defa​ult/​files/​rebi​rth-​educat​ion-​int​rodu​ctio​n_​0.pdf).

Addressing execution challenges—III  381 information technology solutions and platforms has opened up, which offers exciting opportunities. Digital technology can help teach to the child’s level in primary schools through adaptive learning, complement traditional skill acquisition with blended learning models, and expand access to high-​quality healthcare through telemedicine. Precision agriculture can overcome the constraints of capital and irrigation by enhancing productivity manifold. The Internet of Things (IoT) and technology solutions can improve everything from optimizing water and energy use, utility service quality, transportation management, etc. E-​governance solutions have become popular remedies to addressing government inefficiencies and transforming the quality of public service delivery. A powerful strategy to address informality and the inefficiencies engendered is to harness the power of information technology, workflow integration, and data management. It can help capture financial transactions (formalize financial flows), process flows (avoid duplication of data entry), and re-​engineer processes. These techniques have a promising role to play in areas like the application and processing of statutory government services that can simplify and ease the interface of citizens and businesses with government agencies. Even in policy advice, there is scope for quick and major improvement: • Conversion of a plethora of orders, precedents, and rules into comprehensive and self-​contained ‘master circulars’ available online to all staff and/​or to the public and updating thereof whenever a change is made can reduce the ‘power of the clerk’. It can improve consistency in the application of policy. A systematic effort to consolidate all internal instructions and guidelines in each central and state department and make them available on an intranet would be a major qualitative improvement to the functioning of the civil service. The key is that the compilation should be definitive and comprehensive and any instruction not there should be deemed to be withdrawn. • Electronic workflow for file decisions is another promising area, provided it is designed in a user-​friendly way. It is crucial that electronic workflow preserves the record of advice tendered at different levels without allowing any level to obliterate previous recommendations.

382  Possible Solutions There is no doubt that e-​governance has great potential, but is not a panacea. The repeated references to railway reservations (and recently, passport issuance) in most Indian discussions of e-​governance paradoxically illustrates how few the known success stories are. E-​governance in the policy and regulatory sphere (as distinct from service delivery, where it has more potential) presents a greater degree of difficulty. One can sense almost an impatience within the society and among opinion makers at the slow pace of the adoption of digital solutions. Governments have come to realize their adoption as signalling aspirational motives and being electorally popular. Officials see these projects as glamorous and worthy of pursuit and flaunt them in their resumes. Citizens suspend disbelief and summon faith to believe that these shiny new things may perhaps transform their lives. The private providers are naturally interested and willing to do anything to boost the hype and expand the market as quickly as possible. Finally, a whole army of consultants, advisors, project managers, third-​party auditors, and so on see the market catalysis opportunity and have jumped into the market.

The practical problem is that suppliers of such solutions (including the consultants advising on them) often view these as opportunities to fit the available digital solutions to the prevailing problems. Understanding the problem comprehensively, figuring out the ‘theory of change’ that would fix it, identifying constraints, developing solution choices, prioritizing them, building execution capability, and then executing them with high fidelity are very important. The history of digital technology interventions

Addressing execution challenges—III  383 from across the world in all these sectors shows that, in most cases, these are more important than the technology solution itself.15,16 The latter is easier said than done. For a start, stakeholders, including governments, are often not initially aware of these challenges and concerns lying beneath the surface. Hurried implementation and short-​time horizons may exacerbate the problem. For example, if the issue is delay and corruption in giving new electricity connections, workflow automation can help only with the application and approval of the connection. The actual delivery of the connection, which involves fieldwork with all its practical challenges, including the workload of the local officials, still remains to be addressed. A property tax assessment runs into problems with multi-​layered tax slabs, which leave ample room for discretion and associated harassment and local bargaining. In the issuance of statutory certificates, the threat of rejections on several possible grounds is often big leverage with officials which digitized workflow can do little to overcome (except in providing a track record of delay—​but this can be circumvented by quickly rejecting applications without delay, whenever the bribe is not paid). The potential of Aadhaar and digital technologies is circumscribed by the fact that they can only identify the person but cannot say whether the identified person is eligible for accessing a certain benefit or not. Similarly, Aadhaar-​enabled bank account and mobile phone–​based digital transfers can enable access to formal banking but cannot be a substitute for bankability and creditworthiness. The experience with the use of the vaunted ‘big data’ in creditworthiness assessment has so far not been very promising.17 Transformations from artificial intelligence (AI) remain elusive and there may be fundamental limits to what it can do, including in e-​governance.18

15 Iqbal Dhaliwal and Rema Hanna, ‘The Devil Is in the Details: The Successes and Limitations of Bureaucratic Reform in India’, Journal of Development Economics, 124 (January 2017): 1–​21 (https://​www.scienc​edir​ect.com/​scie​nce/​arti​cle/​pii/​S03043​8781​6300​669) (Accessed on 29 December 2018). 16 Esther Duflo, Rema Hanna, and Stephen Ryan, ‘Monitoring Works: Getting Teachers to Come to School’ (https://​pdfs.sema​ntic​scho​lar.org/​c39d/​1baf842d1​bf2e​835d​85af​4349​a763​ bcdb​2da.pdf) (Accessed on 29 December 2018). 17 Yuan Yang, ‘Does China’s Bet on Use of Big Data for Credit Scoring Work?’, Financial Times, 20 December 2018 (https://​www.ft.com/​cont​ent/​ba163​b00-​fd4d-​11e8-​ac00-​57a2a​8264​23e). 18 Tim Cross, ‘An Understanding of AI’s Limitations Is Beginning to Sink in’, The Economist, 11 June 2020 (https://​www.econom​ist.com/​tec​hnol​ogy-​quarte​rly/​2020/​06/​11/​an-​unders​tand​ing-​ of-​ais-​limi​tati​ons-​is-​start​ing-​to-​sink-​in).

384  Possible Solutions The availability of a technology does not automatically enable access and utilization. For example, adaptive learning techniques require investments in IT hardware, including personal equipment, which runs into resource constraints. Precision agriculture techniques require large-​scale investments that are beyond the capability of the overwhelming majority of farmers. Fundamentally, one needs to be at the starting line to be able to make use of the wonderful world of exciting opportunities that become available due to digital technologies. Unfortunately, many Indians are far away from this starting line to be able to capitalize on these great technologies. Entrenched human behavioural constraints often cannot be overcome by technology. Most critical public policy and social issues remain massive challenges because they involve complex human engagement. This introduces behavioural challenges that pose formidable last-​mile gaps. Enabling access to hospitals, nurses, and even vaccines does not automatically translate into universal immunization. Opening a bank account, even easy access, does not automatically lead to its use and increase financial security. Most importantly, there is the weak ability to enforce rules and regulations, which applies even in the implementation of digital technologies. Consider the example of state-​of-​the-​art crowdsourced complaints or grievance redressal systems (apps, websites, toll-​free numbers, etc.) now commonplace across the largest cities and districts. While this has dramatically improved the process of consolidating and rendering complaints about utility services to the officials concerned, the municipal body’s capability to effectively respond to these complaints has scarcely improved. Similarly, a biometric attendance system without the information being used as decision-​support in performance management and follow-​up action is meaningless. In general, any digital workflow and information management system that provides information that the system is incapable of responding to is of limited utility unless accompanied by the necessary non-​digital managerial changes. Last but also least, there is the issue of resources—​the funds needed for, say, technology-​driven parking meters, Geographic Information System (GIS)-​based property tax management, Supervisory Control and Data Acquisition (SCADA)-​based water distribution, and the vast array of opportunities with IoT sensors.

Addressing execution challenges—III  385 In other words, digital technologies undoubtedly have the potential to assist in all these endeavours, improve the quality of service delivery, and enhance the efficiency of public spending. But daunting challenges remain before their potential can be realized.

15.4  Evaluating Programs One of the common complaints about government programmes is that they are not subject to outcome assessments. In international development, evaluations have come to become a standard requirement with aid programs. Evaluation and the wider term ‘monitoring and evaluation’ have been around for decades. The quality of evaluations varies widely in rigour, ranging from pure qualitative studies to rigorous field experiments and econometrics. Further, they are also done internally or by independent third parties. Naturally, different issues and contexts demand different types of evaluations.

15.4.1  The Case for Evaluations As governments grapple with ever-​increasing demands, the meeting of which far exceeds the fiscal capacity, the need to generate value for money or maximize social value from public spending has become all the more important. In this regard programme evaluation and evidence-​based policymaking assume relevance. Given that the proof of policy success is in its implementation, good policies should incorporate a credible and practical feedback mechanism that can help address emergent failings in the policy during its implementation, especially in the initial stages. Further, for policies that have broader implications, impact, or financing, there should also be an independent evaluation mechanism to facilitate course corrections and assess performance. They are important also because very few policies are actually made afresh. Most often, a policy or program is built on an extant one, either by expanding its scope or by improving on it. For example, successive

386  Possible Solutions governments have had flagship programs on self-​employment, rural drinking water supply, weaker section housing, rural roads, skill development, maternal and child health, and so on. In fact, every major sectoral policy and the programs thereon initiated by successive governments over the years are essentially enhanced versions each time. In the circumstances, a credible evaluation of the preceding policy or program and its findings can be useful while designing the new version. While the essential features are dictated by political economy, there is considerable space for improvements within the operational design. In the absence of a credible evaluation, there is a strong risk of regurgitating the existing version with a few tweaks to accommodate some new priorities and subjective preferences of individual policymakers. The nature of bureaucratic decision-​making, with its preference for the status quo, unless otherwise required, increases the likelihood of this approach. For sure, as mentioned, evidence and value for money is hardly the basis for headline decisions on development program choices. For example, farm loan waivers and free farm power are essentially political choices exercised by elected governments. There is no dearth of studies showing how sub-​optimal and even counter-​productive such schemes are yet, and most governments give assurances to the beneficiaries that they will not be withdrawn. The ability to manoeuvre with iterative adaptation in policy design in general is limited in most cases. For sure, it is possible to tinker with the design at margins, though these are most likely to yield only marginal benefits. But the implementation of other programs like the Samagra Siksha Abhiyan or National Health Mission or Skill India Mission or Smart Cities Mission or Housing for All or Employment Guarantee offers significant scope for refinement with evidence. Further, all of them are long-​running and likely to remain so and therefore have the potential of opportunities for assessment and improvement. Also, in the case of policies and programs which are considered politically sensitive, an objective performance assessment against the mandated outcomes can help contribute to the gradual emergence of the political support for reforms. In the case of programs that offer the luxury of the phased roll-​out, as may be the case with large new programs, there is a case for making rigorous program evaluations mandatory. The information generated could be used to improve the program design for subsequent phases. The

Addressing execution challenges—III  387 minimal cost required for such evaluations, as a share of the total program cost, makes evaluation a high value for money intervention. In fact, the mere availability of credible evaluations creates institutional incentives and increases the likelihood of more effective policy and program design. For it could then become a bureaucratic imperative to incorporate the evidence and improve the policy or program’s design. After all, if such evidence is not considered and the failings repeat, as would be likely, with attendant wasteful public expenditure, a post facto audit can fix accountability and hold the policymakers concerned for their failure to heed evidence while designing the policy. This would require institutionalizing and tightly coupling policy and its evaluation. Such assessments can be credible only if they are independent and sufficiently rigorous. This is important since it has become a routine practice to have program evaluations involving anecdotal evidence—​case studies, descriptive assessments, unscientific surveys, etc.—​done by hiring consultants. Such evaluations are generally done by the government agency as part of program-​related public relations exercises rather than as a genuine effort to uncover deficiencies and feed them back to improve the program’s implementation effectiveness. Even when they use some quantitative metrics they do little to rigorously examine the program’s impact, much less establish causality between the program and certain outcomes. These so-​called evaluations usher in the sense of complacency and perpetuate inefficiency, leakage, and corruption. Besides, they can rarely ever be used to meaningfully influence decision-​making. The conflict of interest in evaluations done by the same implementing agency is also another important problem. It is therefore important for government agencies to hire evaluation agencies with the specific mandate of providing independent and rigorous enough assessment of the program. We examine these in the next two sub-​sections.

15.4.2  Limitations of Evaluation However, the case for evaluations, especially outcome evaluations, is not without its sceptics. The scepticism is not so much about the theoretical merits of outcome evaluations as their practical value.

388  Possible Solutions Firstly, the theory of how evaluations are useful does not reckon with the political nature of many government programmes. If programmes were launched based on a careful assessment of alternatives and then subject to apolitical decisions on continuation based on evaluation, that would be one thing. However, if programmes are launched based on loosely expressed manifesto promises, then evaluation is often incapable of changing design, however good the quality of the evaluation itself. Moreover, frank and negative evaluations that become available in public may get used as a political stick to beat the ruling party with. When a tool meant for corrective action becomes a tool of ‘fixing blame’, it becomes much less useful and administrators become reluctant to use it actively. Secondly, in practice, programs and interventions are rarely ever complete successes or failures. The vast majority of programs realize at least some of their objectives. So, any outcome evaluation runs the risk of not being able to offer actionable conclusions. Thirdly, all programs are dependent on a variety of contextual factors, at least some of which are beyond the control of the implementing agency. It is therefore easy to blame shortfalls on these factors. In fact, it would be practically difficult to refute post facto explanations adduced by the departments for such shortfalls. Also, outcome evaluations struggle to capture the causes for their shortfalls or their successes. Fourthly, in the case of the vast majority of development interventions—​schooling, primary health care, agriculture extension services, livelihood skills development, nutrition, cash transfers, etc.—​it may be difficult to dispute the premise of the programmes themselves. The design of the programs too is unlikely to have egregious failings. Even with program elements, very few would be matters of dispute. Therefore, there is little need for evidence generation on them. Instead, the failures are mostly in implementation, and that too, mostly in plain, simple execution. Weak state capability constrains high fidelity execution, but outcome evaluations hardly offer anything of relevance in addressing this problem. Fifthly, outcome evaluations are post-​mortems of programs and therefore offer little by way of suggestions to improve ongoing programs. Finally, the global evidence on the effectiveness of evaluations in improving program design, implementation, and outcomes is very thin. For example, though the World Bank has done rigorous evaluations of most

Addressing execution challenges—III  389 of its programs for over 60 years, it has rarely changed the reality surrounding the program by much in the host country. It is perhaps a moot point whether the World Bank’s elaborate institutionalized mechanisms of feeding back the output of evaluations of its programs have improved the outcomes from those programs elsewhere.

15.4.3  Institutional Architecture for Evaluation The central government in India has a long history of evaluation. The Planning Commission, responsible for preparing the Five Year Plans and monitoring their implementation, has had a Program Evaluation Office (PEO) since 1952. It even had regional evaluation offices. It was tasked with the systematic evaluations of community development programs and certain selected centrally sponsored schemes. The objective was to ‘assess the performance, process of implementation, effectiveness of the delivery systems and impact of programs’.19 The evaluations were to be diagnostic and generate ‘lessons for improving the performance of existing schemes through mid-​course corrections and better design of future programs’. The PEO undertook ex post evaluations of all flagship programs, primarily by conducting field surveys and analysing administrative data.20 While some of them have revealed important insights, by and large, these evaluations ended up as routinized administrative exercises done by a group of statisticians. They suffered from inadequate rigour and rarely generated valuable and actionable information on either course corrections for implementers or insights for policymakers. The biggest problem was the lack of ownership by the individual line departments who were ultimately responsible for those programs. Neither did they have much engagement in the evaluation process itself, nor was the feedback integrated into the departmental decision-​making process. Not unexpectedly, the evaluations became another one of the silos that characterize the functioning of government bureaucracies. 19 Organisational Structure and Functions, Program Evaluation Office, Planning Commission of India (https://​niti.gov.in/​pla​nnin​gcom​miss​ion.gov.in/​docs/​rti/​doc_​rti/​divi_​deta​ils/​peo.pdf). 20 Program Evaluation Studies, Planning Commission of India (https://​niti.gov.in/​pla​nnin​ gcom​miss​ion.gov.in/​docs/​repo​rts/​peorep​ort/​index.php?repts=​peob​ody.htm).

390  Possible Solutions In recent times, there have been efforts to make evaluation more rigorous. In 2014, the Government of India established the Independent Evaluation Office (IEO) primarily to conduct evaluations of the largest flagship programs for their effectiveness, relevance, and impact.21 In fact, one of its three focus areas goes to the heart of this book itself, ‘The inability to implement and deliver public services is well known but as to why India finds it so difficult to deliver public services is a puzzle.’ The IEO has since been replaced recently by the Development Monitoring and Evaluation Office (DMEO), located within the NITI Aayog. In light of this experience with evaluations, it may be useful to examine what is the best strategy to institutionalize independent evaluations into departmental programs. It is clear that in order to be perceived as credible and valuable by the line departments, such evaluations need to be rigorous, concurrent with the implementation, and provide actionable insights. The presence of a nodal evaluation agency is necessary to standardize processes and protocols and thereby ensure the rigour of the evaluations. They could also certify evaluations done by others. It is also necessary to maintain the independence of the evaluations by mitigating perverse incentives of the departments whose programs are being evaluated. This does not mean that all evaluations should be outsourced to DMEO. There is no substitute for departmental ownership. The nodal agency for evaluation has to work closely with the department in both the design and execution of the evaluation. The evaluation has to both respond to the need to have an independent assessment of the program and inform the department about course corrections and revisions. In fact, departments should also be encouraged to hire independent agencies and do program evaluations in accordance with the principles laid down by the nodal evaluation agency. The nodal agency could prepare and make available model contracting documents and evaluation templates for different programs and sectoral activities. District Collectors and state government departments, for example, could take

21 Independent Evaluation Office launched, Business Standard, 27 February 2014 (https://​ www.busin​ess-​stand​ard.com/​arti​cle/​econ​omy-​pol​icy/​inde​pend​ent-​eva​luat​ion-​off​i ce-​launc​ hed-​114​0226​0082​9_​1.html) (Accessed on 11 November 2021).

Addressing execution challenges—III  391 these documents and procure their own third-​party assessment providers. This could help create a culture of evaluations and associated accountability. The DMEO could also forge partnerships with universities, colleges, state administrative training institutes, and think tanks to support with independent evaluations. The DMEO could work as an umbrella organization to catalyze a culture of independent evaluation and an ecosystem to support it. Doing evaluations is one thing, making use of them is another. It is therefore necessary to have appropriate institutional requirements to ensure that the evaluations are taken seriously and their findings used as decision-​support by departments, either in improving program design or as an input while formulating the new policy or program. In this regard, the output-​outcome measurement framework (OOMF) instituted in the union budgeting process since 2017–​2018 is a step in the right direction. It mandates that in addition to the financial outlays, Ministries also present their targeted outputs and outcomes with associated indicators in their budget proposals. In 2017, the Department of Expenditure made evaluations compulsory for the appraisal of ongoing schemes at the end of each five-​year cycle and this was implemented in the appraisal cycle for 2021–​2026.

15.4.4  Ensuring Rigour in Evaluations Many program evaluations are insufficiently rigorous. The evaluations of flagship programs by government’s own evaluation agencies generally confine themselves to measuring the change in specific outputs before and after the program. This is done mostly using administrative data and some sample surveys. It is supplemented with qualitative assessments involving case studies and anecdotal records. Often, such assessments often do not capture the real program impact because: • they may not control for other factors that either amplify or attenuate the program’s impact;

392  Possible Solutions • even where there is data analysis, it is mostly confined to measuring basic changes; • the quality of administrative data used to measure outputs is often questionable; • they generally do not measure outcomes; • assessments have an inherent optimism bias, most often looking for positive case studies and quotes; • steered as they are by the department concerned, they often lack sufficient independence; and • they are generally one-​off and do not involve longitudinal assessments of program impact.

Addressing execution challenges—III  393 However, this critique should not be taken to mean advocacy for rigorous Randomized Control Trials (RCTs), which have become a fad in international development circles (see later in this chapter). In fact, given their limitations, cost, and time requirements, RCTs should be used sparingly. Instead, there is much scope for better evaluation using quantitative and qualitative assessments. Examples of possible techniques are high-​ quality sample surveys to measure impact and process evaluations. Any program or intervention is defined by a series of processes, the execution of which with high fidelity is the biggest challenge. It is a reasonable assumption that if the processes are complied with, then the program would deliver its objectives. Process evaluations inform compliance with the program processes during implementation. For example, a mid-​day meals program may have four major processes—​procurement of ingredients, cooking of meals, consumption by children, and program management. There is a protocol for each process. The protocols could be reduced to a checklist of requirements for each process. Their adherence could become a focus of evaluations. With the support of DMEO, the departments could formulate process evaluation checklists for different programmes. These evaluations have an advantage over outcome evaluations in that they can also be done concurrently with implementation. Further, process evaluations could shine a light on critical operational weaknesses. This makes it possible to use the evaluation results to improve the program implementation. Periodic process evaluations, done by an independent entity, can provide useful insights about whether the intervention will deliver the desired results. Regular data analytics is the simplest and cheapest, and perhaps the most important, means to gather insights about program performance.

15.4.5  RCTs and Their Limitations Evidence-​based policymaking is currently the buzzword in development, and evidence emerging from RCTs is considered more relevant to policymaking. Some of the most reputed researchers even consider other forms of evidence not of an acceptable standard. Given the salience of RCTs in international development (funding decisions by multilaterals and other

394  Possible Solutions donors are swayed by evidence from RCTs), though not so much in policymaking in India, it is important to place it in perspective. Essentially, evaluating an intervention faces the problem of confounding factors—​several unobservable things may be going on which are contributing to the intervention effects. For example, economic growth and time itself contribute to some trends which get conflated with the intervention’s impact. The cleanest way to isolate causality is to randomly generate two samples and treat (with the intervention) one sample and use the other as a control. The measured differential in outputs or outcomes is generally a credible measure of causality. Thus the value of RCTs. However, it is important that practitioners avoid an obsession with RCT evaluations, to the exclusion of all methods. The major shortcomings of RCTs are well known.22 They are expensive and time consuming and suffer from well-​known generalizability or external validity problems. Further, very few development issues and policies are amenable to an RCT evaluation.23 In fact, RCTs, by their very nature, favour private goods over public goods.24 There are also questions about interpreting the statistical significance of the results of RCTs.25 Another less discussed problem with RCTs is their ‘implementation validity problem’. A typical RCT is a small experimental pilot with the smallest sample size consistent with statistical requirements done under the supervision of some principal investigators and with the field support of smart and committed research assistants (RAs) and field managers. This poses a problem. RCTs draw on the immense energies of reputed researchers and their committed and passionate young RAs to protect the integrity and fidelity of the experiment. This effect, which would be absent in business-​as-​usual implementation, in turn, contributes to the experiment’s effective implementation. This raises the question, how do 22 Angus Deaton and Nancy Cartwright, ‘Understanding and Mis-​ Understanding Randomised Control Trials’, Social Science and Medicine, 210 (August 2018): 2–​21 (https://​www. scienc​edir​ect.com/​scie​nce/​arti​cle/​pii/​S02779​5361​7307​359). 23 Florent Bedacarrats et al., Randomised Control Trials in the Field of Development: A Critical Perspective, Oxford University Press, USA, 2020. 24 Jeffrey Hammer, ‘Randomised Control Trials for Development? Three Problems’, Brookings, 11 May 2017 (https://​www.brooki​ngs.edu/​blog/​fut​ure-​deve​lopm​ent/​2017/​05/​11/​ran​domi​zed-​ cont​rol-​tri​als-​for-​deve​lopm​ent-​three-​probl​ems/​). 25 Deaton and Cartwright, ‘Understanding and Misunderstanding Randomised Control Trials’.

Addressing execution challenges—III  395 we evaluate the treatment (or the policy or intervention being proposed) in the business-​as-​usual environment? An RCT typically establishes the efficacy of the treatment. But it does not say much about its effectiveness, a measure of both the efficacy and implementation fidelity. This assumes great significance since the same intervention or treatment is generally implemented through government systems, which are notoriously enfeebled. In fact, given the weak state capability, trying out interventions whose efficacy has been established through RCTs is likely a recipe for failure. We describe this as the implementation validity problem. Then there are practical operational issues. For a start, in most cases, the bureaucrat does not have the flexibility to create treatment and controls. Second, isolating a problem and its potential solutions neatly, so as to explore attribution, is most often impossible. Third, being embedded in large and complex systems, where the influencing factors are not easily identified, contributions rather than attributions are easier to explore. Fourth, officers take decisions in real time and therefore do not have the luxury of long-​drawn experiments. Fifth, most often the immediate results of these interventions merely reflect partial equilibriums, and steady-​state results take a long time to become evident. So, headline evaluations rarely serve the purpose. Related to the previous point, there is no one good solution to a problem that can be picked up and implemented. Instead, problems (including internal sabotage) start to surface when the solution hits the road, necessitating iterative adaptation, especially in the initial period, before it stabilizes. At a more fundamental level, the international development community’s obsession with RCTs has crowded out valuable mixed-​ method studies. Examples of such valuable studies are those by Dasgupta and Kapur on the challenges faced by the frontline development bureaucracy in India26and by Mangla on the relatively successful implementation of universal primary education programs in Himachal Pradesh.27 26 Aditya Dasgupta and Devesh Kapur, ‘The Political Economy of Bureaucratic Overload: Evidence from Rural Development Officials in India’, American Political Science Review, 114 no. 4 (2020): 1316–​1334 . 27 Akshay Mangla, ‘Bureaucratic Norms and State Capability in India’, Asian Survey, 55, no. 5: 882–​908 (https://​www.hbs.edu/​facu​lty/​Publ​icat​ion%20Fi​les/​Asian_​Sur​vey_​5505​_​03_​Mang​ la_​2​19b4​d54-​2a4f-​4a43-​a0ab-​d6326​28f8​fae.pdf) (Accessed on 20 December 2018).

396  Possible Solutions Both are useful documentations of field realities that are deeply internalized by those working within these systems but not amenable to any randomizable study and therefore less known to researchers and other opinion makers. Finally, we need an evaluation only where there is a dispute or uncertainty about a hypothesis. Expensive and time-​consuming RCTs have been done to show that telephone-​based verification improves the quality of public services delivery; third-​party audits lower industrial emissions; random inspections reduce drunken driving; diabetics could do with walking and should be incentivized to walk; and sleep deprivation reduces productivity; the endless series of research on nudges of various kinds and so on.28 None of these demanded any RCT evaluation. In all these examples, and also the vast majority of RCTs in development, any reasonable bureaucrat would claim that little actionable new information has emerged from these expensive and time-​consuming RCTs. The following candid assessment of RCT research by Devesh Kapur is insightful29: Suffice it to say the effects have been much more positive for the careers of US-​based researchers than for India. Ironically, some of the very strengths, such as the stress on identification and causal inference, have been a source of weakness. The stress on these methods as “the gold standard” comes at the cost of relevance and timeliness. Only certain types of questions can be addressed by these methodologies. This is not to say there aren’t excellent studies that address important policy questions. But more often than not, even if they can address them, the costs and duration of these studies means they are more useful as citations than policy. When asked how many of these expensive RCTs had moved the policy needle in India, Arvind Subramanian, Chief Economic Advisor,

28 Interview with Gulzar Natarajan, in Florent Bedacarrats et al., Randomised Control Trials in the Field of Development: A Critical Perspective. 29 Devesh Kapur, ‘The Study of India in the United States’, Centre for Advanced Study of India, University of Pennsylvania, 29 June 2018 (https://​casi.sas.upenn.edu/​iit/​deve​shka​pur2​018) (Accessed on 20 December 2018).

Addressing execution challenges—III  397 GOI, was hard pressed to find a single one that had been helpful to him in addressing the dozens of pressing policy questions that came across his table.

Thus, RCTs need to be used where appropriate.

16 A Practical Approach to Better Capability This book has so far looked at the problems of state capacity and at possible remedies. Considering the magnitude of the problems and the uncertainties of the remedies, it is easy to be overwhelmed. There is a massive and complex system which needs to improve. What can one actually try to do to put in practice the various ideas and suggestions discussed? This concluding chapter starts by discussing some ways to successfully bring about systems change in developing countries1 like India.

16.1  Encouraging Positive Deviances ‘Positive deviances’ are examples of distinct improvements from prevailing normal practices. Encouraging positive deviances is something that is practical, can be initiated anywhere, and can (if successful) trigger system-​wide effects. This involves spotting champions of change and supporting them to emerge as ‘positive deviances’. It requires identification of the appropriate level and spotting the champions, which in turn require deep contextual and systemic understanding. In the words of Mark Moore, this approach is about creating ‘restless armies of value creating agents’.2 Once there is a small sample of such champions, one can hope that some of them are replaced by equally good successors and the reforms have enough time to stick. A coalition of interested stakeholders may eventually coalesce to form a critical mass for systemic change. 1 While only four factors are mentioned here, there may be other pathways to change. 2 Mark H. Moore, Creating Public Value: Strategic Management in Government, Harvard University Press, USA, 1997. State Capability in India. T. V. Somanathan and Gulzar Natarajan, Oxford University Press. © Oxford University Press 2022. DOI: 10.1093/​oso/​9780192856616.003.0016

400  Possible Solutions In the Indian context, the district is often a level where positive deviances emerge, though they can emerge elsewhere too (for instance, in one Secretariat department). In terms of champions, leaders at the district and upward levels are sufficiently empowered to undertake systemic transformations. They meet the appropriate delegation, information, finance, and motivation requirements in terms of design elements for different relationships. The whole e-​governance ecosystem in India can trace its origins to several positive deviances, both on use cases (functions covered) as well as workflow innovations (methods of carrying out the functions), that emerged over the early 2000s across districts of India. Over time the best features of different positive deviances got consolidated to form robust e-​ governance solutions across the spectrum of development activities. Several national programs of today have emerged from pilots in different parts of the country. District Collectors and state governments have experimented with innovative ideas and some have emerged successful, leading to their adoption as a best practice. As mentioned earlier, the entire landscape of e-​governance interventions can be traced back to some successful experiment in a district or city or state. Globally too, the use of public–​private partnerships (PPPs) in infrastructure contracting has evolved gradually through the emulation of positive deviances in different types of contracting structures and across different sectors, especially in the United Kingdom, Canada, Australia, and Latin America. Resources like the World Bank’s Public Private Partnership Infrastructure Advisory Facility (PPIAF)3 have helped in consolidating the learnings and providing resources that interested parties could take off the shelf and use with minimal customization.

16.2  Promoting Anchor Interventions Given the amorphous nature of what is required to realize system transformation, it is useful to have a clear and actionable starting point. ‘Anchor interventions’ are interventions designed around an aspirational

3 Public Private Partnership Infrastructure Advisory Facility, World Bank (https://​ppiaf.org).

A practical approach to better capability  401 or politically attractive agenda, which can be influential among important stakeholders. India’s Smart Cities programme is a good example of an anchor intervention for systemic changes in urbanization. The prevailing policy on urbanization emphasized the provision of infrastructure. A pivot to the quality of utility service delivery and standard of life in cities demanded a changed objective. The Smart City initiative with its focus on aspirational elements—​technology, liveability, transit-​oriented development, etc.—​ has all the elements that respond to the new objectives. Technology platforms—​by consolidating data and making available actionable information as decision-​support and monitoring systems and which can be used to gradually build out modules covering other use-​ cases—​are a good example of likely anchor interventions. Similarly, the ease of doing business (EoDB) rankings reflect improvements in the business environment and consequently signal investment attractiveness to investors resonates with a political imperative. The output of the National Assessment Survey (NAS) to measure student learning outcomes can form the basis for the construction of the School Education Quality Index (SEQI), which in turn could be used to generate competition among states on the issue of learning outcomes. However, anchors are only a means to an end. This is important to note since there is the possibility of the anchor becoming an end in itself.

16.3  Making Disruptive Changes Sometimes, systems are stuck in a really bad equilibrium.4 In such circumstances, incremental approaches are unlikely to be effective. This demands a big shock to shake up and push the system into a completely new trajectory. Such sudden regime shifts replace legacies—​new rules of the game, new personnel, new entities, new political regimes, etc. They are essentially

4 Matt Andrews, Lant Pritchett, and Michael Woolcock, ‘The Big Stuck in State Capability for Policy Implementation’, CID Working Paper 318, 10 May 2016 (https://​bsc.cid.harv​ard.edu/​ publi​cati​ons/​big-​stuck-​state-​cap​abil​ity-​pol​icy-​imp​leme​ntat​ion) (Accessed on 3 January 2019).

402  Possible Solutions mutations that cannot be planned for, though one could work actively to create the conditions for them. Legislative and regulatory changes are perhaps the most prominent examples of consciously planned regime shifts. They include India’s bank nationalization of 1969 and economic liberalization of 1991, the de-​collectivization of agriculture and promotion of private town and village enterprises in China, the collapse of the Communist regimes, etc. The Indian Insolvency and Bankruptcy Code (IBC) is a recent example of how a mutation can trigger the emergence of conditions that aligns incentives of lenders and borrowers to create a strong bankruptcy and resolution environment. It immediately disrupted the entrenched moral hazard that large borrowers had with their creditors—​given the large exposure and the limited means to recover them, creditors had no option but to reschedule and roll over loans. Demonetization in 2016 was a disruptive change that has altered tax compliance behaviour significantly. Under the then Chief Minister of Tamil Nadu, Mr. M.G. Ramachandran, the centuries-​old traditional institution of village officers was abolished overnight by promulgating an Ordinance on 13 November 1980. The rediscovery of the Election Commission of India’s (ECI) constitutional powers in the early 1990s can be traced to the activism of the then Chief Election Commissioner of India, Mr. T.N. Seshan. His willingness and firmness to exercise all its powers transformed the institution itself and reshaped the expectations among political parties and electors. It dramatically transformed the electoral process and eliminated egregious electoral malpractices, which had become pervasive. However, similar examples of sustained individual-​based transformations are rare, partly because few institutions have such power concentrated in one individual with such high job security as India’s then Election Commission. (The Commission has since become a multi-​ member collective decision-​making body, partly in reaction to Mr. Seshan.)

A practical approach to better capability  403

Disruptive change is a high-​risk-​high-​reward technique. History is replete with instances of its backfiring. And history also records several successes. Thus, it needs to be deployed with care and after a careful assessment of the risks. Some disruptive changes may require secrecy and suddenness for legal or practical reasons; a special difficulty in such cases is that success may require secrecy and diminish the potential for consultation—​thus increasing the risk.

404  Possible Solutions

16.4  Identifying and Using the Cumulative Effect of Multiple Factors Experience shows that things that seemed impossible can sometimes appear to just happen. The fall of the Soviet Union and the Berlin Wall are examples. In such cases it is difficult to attribute the change to any proximate causes, but more likely is the cumulative effect of several diffuse factors over a long period of time. The awareness of this fact—​that change that appears difficult can happen when conditions are ripe—​leads to a method for policymakers and administrators to induce change. The approach is to consciously try to create, and then build on, the cumulative effect of such factors. Many changes require traversing the path of exhausting alternative options and demonstrating their inadequacy or failure, thereby creating the conditions for acceptability of the desired change element. This is often the political reality. In fact, in many cases the desired change elements themselves emerge only as implementation progresses, for example, the shift in the consensus on education in India. In the absence of buildings and teachers, it is understandable that focussing on learning outcomes could not have been a bureaucratic or political imperative. The shift in focus towards learning outcomes over the past few years does not have any immediate or sudden cause. It has been the cumulative result of several factors, the most prominent being the receding importance of inputs and the growing salience of poor learning outcomes and its increasingly evident adverse impact on productivity and economic growth. However, the cumulative nature of the change has also meant that the agenda on what and how to move to address deficient learning outcomes has remained still-​born and without any champions to lead it. Without the difficult experience and struggles of having done engineering works through government agencies (and memory thereof), it would be difficult for systems to embrace PPP contracts and concessions. Without undergoing the bitter suffering and damage associated with fiscal austerity and the unhinging of attendant moral hazard, it would have been politically difficult for governments to embrace expansionary policies in Europe. The political will to embrace delegation of power requires a history of failure with centralized administration.

A practical approach to better capability  405 Triggering system-​wide transformations may require a deliberate and concerted combination of several measures with flexibility and corrective feedback loops. In a sense, this is the polar opposite of disruptive change. Several of the ideas discussed in this book (and others not discussed here) may need to be deployed in an orchestrated combination. Examples of possible measures—​by no means exhaustive—​include promoting pilots of innovative ideas; cultivating champions at all levels; empowering lower levels extensively (even at the risk of short-​term failures); nurturing positive deviances (and encouraging their diffusion); making procedural, personnel, and budgetary changes; creating the conditions for internal pressure for change, etc. Once initiated, the challenge is to observe closely and engage opportunistically based on emergent dynamics of the system and steer the course without letting things get off control. This approach is not easy to put into practice. It is a long-​drawn effort with uncertain pathways of change. The most difficult part of this approach is that it cannot be planned in any detail. It can also be uncertain in the time taken to achieve results—​and it can fail. Nevertheless, there are situations where it may be worth trying. For those leading the change, it needs patience, tact, and an ability to continuously step back and in as required. It is difficult to reduce this approach to even a plan, much less a comforting replication toolkit. The box illustrates a possible methodology of triggering multiple pathways to change.

Education System Reform: An Illustrative Proposal

School education reform to realize student learning outcomes presents a very complex challenge. It defies the conventional top-​down, one-​size-​fits-​all, and uniformity associated with most development interventions. Crouch and DeStefano* and Burns** argue that getting decentralized schools and teachers to adopt new behaviours depends on: • Setting and communicating expectations for education outcomes • Monitoring and holding schools accountable for meeting expectations • Intervening to support students and schools when needed.

406  Possible Solutions In the context of preparing a plan for India with its continental size, diversity, and complexity, two more points need to be added to complete the framework. • Setting and communicating expectations for education outcomes • Informing about the possible menu of intervention options to improve learning outcomes • Supporting states, districts, and schools with adoption and implementation • Monitoring and holding schools accountable for meeting expectations • Appropriately incentivizing the school system. The first step should involve mobilizing a coalition and solidifying a strong national commitment to realize transformative change. The plan should lay down the broad contours and decentralize implementation. It should guide with the possible set of learning outcomes-​focussed interventions, define very simple and easy-​ to-​ understand metrics of success, list out a set of process indicators that are proximate determinants to success, and some broad enough milestones and timelines. But the strategy should resist the temptation of prescribing specific interventions, much less their implementation strategy, and instead only create the enabling and incentivizing conditions for realizing the outcomes using the possible set of interventions. It should eschew hard timelines and allow school systems sufficient space and time to innovate to achieve their goals. Central and state governments should help school systems implement their chosen set of interventions by supporting them with all technical assistance support. For example, if a school or a district wants to do third-​party assessments or a type of remedial instruction, the state or central government should support with model guidelines and implementation toolkits (say, procurement and contracting document, if it requires hiring a private provider). The program will naturally get implemented at varying paces and intensities. Some positive deviances will quickly emerge, often from

A practical approach to better capability  407 unexpected places, and there will be similar consistent laggards. The rest will exhibit ebb and flow, like when officers arrive or depart. The challenge is to learn from the successes and failures, make the positive deviances aspirational for those in the middle, and get everyone to comply with some very basic process requirements. The success of any such plan would depend on a robust monitoring system, which tracks progress on certain basic process indicators as well as learning outcomes and holds schools accountable. School systems that aspire higher should embrace more rigorous monitoring of student learning outcomes. Finally, there should be a system of competition that makes schools compete with each other at a local level in a low-​stakes manner but one which incentivizes enough to generate effort. Such competition should be a celebration of the success of the positive deviances and seek to motivate others to emulate. The trajectory of implementation will be decidedly second-​best in terms of efficiency and speed, not to speak of outcomes. But it will allow schools the freedom to experiment and iterate to their sweet spots of implementation. Some variation of this may perhaps be the only practical ‘strategy’ to realize sustainable and transformative change with learning outcomes. *

Luis Crouch and Joseph DeStefano, ‘Doing Reform Differently: Combining Rigor and Practicality in Implementation and Evaluation of System Reforms’, RTI Working Paper No 2017-​01, February 2017 (https://​www.rti.org/​sites/​defa​ult/​files/​resour​ces/​rti-​publ​icat​ion-​ file-​0ed2b​b7f-​7248-​442d-​9c0d-​658af​d2e0​922.pdf). ** Barbara Burns, ‘In Bringing a Literacy Project to Scale, Has Kenya Found a Holy Grail?’, Centre for Global Development, 26 October 2018 (https://​www.cgdev.org/​blog/​bring​ing-​liter​ acy-​proj​ect-​scale-​has-​kenya-​found-​holy-​grail).

16.5  Keeping Expectations Realistic In recent years, driven by an increasingly youthful population, rapid economic growth, international travel, and media exposure to the living standards of other societies, people’s expectations from the government have risen dramatically. Going beyond their traditional roles, governments have been entrusted the responsibility of delivering certain fundamental rights and a panoply of basic services to citizens while also

408  Possible Solutions catalyzing market development. India has in fact made tremendous progress since 1947. Governments and officials naturally like to project their actual performance, especially when it is reasonably good. But governments are often left surprised with electoral setbacks, and civil servants with public criticism, despite what may be reasonably good actual performance. Herman ‘Dutch’ Leonard of the Harvard University’s Kennedy School of Government had a brilliant conceptual characterization of the political economy challenge that governments face. He used a theory that has three kinds of performance: • Expected—​this is the performance level the electorate expects; • Actual—​the actual level of performance; and • Perceived—​the perception of the electorate of the government’s performance —​ which may be better or worse than the actual performance. According to this theory, a government is successful and hence re-​elected when its perceived performance is better than expected performance. It is judged to be unsuccessful and loses when its perceived performance falls below expectation. The keyword here is ‘perceived’. Note that in this framework ‘actual performance’ does not appear at all as a determinant of success. The point is that in a modern media-​influenced representative democracy, actual performance is only relevant to the extent it influences perceived performance. When perceived performance falls below actual performance, good communications and public relations can correct the gap. On the other hand, with good public relations, even not-​so-​good actual performance may be perceived as good. If expectations keep rising, even an improved performance may lead to electoral defeat if perceived performance does not keep up. According to this theory, the answer lies in either raising perceived performance (through better performance and/​or better public relations) or lowering expectations. Much of this book has been devoted to improving actual performance, but practical administrators need to be conscious of the role of expectations. Applying Leonard’s theory in practical terms, lowering (or at least

A practical approach to better capability  409 not raising) expectations is often prudent. On the other hand, competitive politics often requires the deliberate raising of expectations, so ‘not raising expectations’ is often not a feasible political proposition. Unlike politicians who face the same electorate on a continuing basis, transferable bureaucrats with 2–​3 year tenures have the luxury of treating their engagements as one-​off ‘games’. This can distort their incentives and encourage them to prioritize short-​term goals and manipulate perceptions, which has been one of the weaknesses of the bureaucracy. However, for apolitical civil servants seeking genuine systemic improvement, keeping expectations realistic is important. Given their poor reputation (deserved or otherwise), expectations of government agencies are generally likely to be low and it often requires effort to raise them. For an administrator trying to improve performance, restraint in raising expectations is a highly effective approach. It is usually a good idea not to seek excessive publicity for plans and wait until actual achievements can be shown. Showcasing achievements, rather than mere plans, strengthen public confidence and also lead to better public satisfaction. For the practical civil servant, academic, or consultant genuinely trying to improve governance, the message is clear: under-​promise and over-​deliver.

16.6 Epilogue This work is a modest attempt at presenting the institutional structure of the state, with particular reference to India, identifying the important policymaking and implementation issues of our times, and offering suggestions to improve state capability. The endeavour has been to combine conceptual rigour with practical realism. Good public policy choices and the political acceptability of those choices are important in development. Yet, the outcome of even technically correct and politically acceptable ideas is dependent on the effectiveness of their design and implementation. Therefore, building state capability is arguably the most important development challenge. However, it is a multi-​dimensional and long-​drawn endeavour, demanding creativity, tenacity, and patience. Thus, it may also be the most difficult development challenge.

About the Authors T.V. Somanathan T.V. Somanathan holds a Ph.D. in Economics (Calcutta University) as well as M.A. (Econ.) and Bachelor of Commerce (B.Com.) Honours degrees from Panjab University. He has completed the Executive Development Program of Harvard Business School and is a fully qualified accountant and company secretary: ACA (England and Wales) FCCA London, FCMA London, Fellow of the Chartered Governance Institute, London, ACMA, and ACS (India). He was the recipient of several academic awards and distinctions including the Ralph Bell Prize for Best Finalist of the Chartered Governance Institute. He joined the Indian Administrative Service in 1987 (ranked 2nd All India) and was awarded the Gold Medal for the Best IAS Probationer of his batch. He is currently Finance Secretary of the Government of India. He was previously Joint Secretary, Ministry of Corporate Affairs (2010–​ 2011), and Joint Secretary/​Additional Secretary to the Prime Minister from 2015 to 2017. He has worked in several senior positions in the Government of Tamil Nadu state, including Deputy Secretary (Budget), Joint Vigilance Commissioner, Executive Director, Metropolitan Water and Sewerage Board, Secretary to Chief Minister, Commissioner of Disciplinary Proceedings, Commissioner of Commercial Taxes, Secretary, Planning, Development and Special Initiatives, and Project Director, Road Sector Project. As founder Managing Director of Chennai Metro Rail Ltd, he was responsible for achieving financial closure and awarding the initial tenders for the Chennai Metro Rail Project. In 1996 he joined the World Bank, Washington, D.C., through the Young Professionals Program as Financial Economist in the East Asia and Pacific Regional Vice Presidency. In 2000, he became one of the Bank’s youngest Sector Managers when he was appointed Manager of the

412  About the Authors Budget Policy Group. In 2011 his services were sought by the World Bank and he served as Director from 2011 to 2015. He has published over 80 papers and articles in journals and newspapers on economics, finance, and public policy and is the author of two books, viz., Derivatives (McGraw Hill Education—​1998, 2nd ed. CUP 2018) and The Economics of Derivatives, (Cambridge University Press—​ 2015, co-​authored), and of chapters in The Oxford Handbook of the Indian Constitution (Oxford University Press—​2016), Land Reforms in India (Sage—​2003), Public Institutions in India (Oxford—​2005), Rethinking Public Institutions in India (Oxford—​2017), and On the Trail of the Black (Rupa—​2017).

Gulzar Natarajan Gulzar Natarajan holds a bachelor’s degree in Chemical Engineering from the Indian Institute of Technology, Chennai, and a master’s degree in International Development (MPA-​ID) from Harvard Kennedy School. He joined the Indian Administrative Service in 1999 and belongs to the Andhra Pradesh cadre. He is currently working as Secretary, Finance Department in the Government of Andhra Pradesh. He has previously worked as Senior Managing Director at the Global Innovation Fund, based in London, co-​leading its global portfolio of investments. Before that he has worked as a Director in the Office of the Prime Minister of India, Vice Chairman and Managing Director of the Infrastructure Corporation of Andhra Pradesh, District Collector of Hyderabad, Municipal Commissioner of Vijayawada, Chairman and Managing Director of a power distribution company based at Visakhapatnam, Project Officer of two Tribal Development Agencies, and Executive Assistant to the Chief Secretary to Government of Andhra Pradesh. He has rich experience in the design, procurement, and implementation of infrastructure projects on both public procurement and public–​ private partnership bases across several sectors. He has also led the implementation of large-​scale projects in health, education, skills and livelihoods, poverty reduction, etc. Besides, he has extensive experience in global impact investing.

About the Authors  413 He has co-​authored two books, namely Can India Grow? Challenges, Opportunities and the Way Forward (Carnegie India—​2016) and The Rise of Finance: Causes, Consequences, and Cures (Cambridge University Press—​2019). He has chapters in Randomized Control Trials in the Field of Development: A Critical Perspective (Oxford University Press—​2020) and Oxford Global Handbook of Social Program Design and Implementation Evaluation (mid-​2022). He also has written several papers and over a hundred op-​eds in leading national dailies on economics, finance, and public policy. Besides, he maintains a top-​rated Economics and Public Policy blog, Urbanomics.

Index accreditation, 51, 378 Additional Secretary & Financial Advisor (AS&FA), 174 Administrative Reforms Commission (ARC), 2nd, 220 advisors, 336, 337 appointment benefits, 338 balanced approach, 341–​342 consequences of appointing, 340–​341 execution challenges, addressing, 336 All India Services (AIS), design, 53, 58–​60 All-​India character, 58 Article 311, 58 cadre posts, 55 dual control, 58 independent selection, 58 initial design, 59–​60 intermediate and lower services of, 60–​64 members, 53 merit-​based selection, 58 promotions into, 229 protection against arbitrary punishment, 58 services, 63 special considerations, 59 anchor interventions, 400 EoDB, 401 India’s Smart Cities programme, 401 NAS, 401 SEQI, 401 Annual Confidential Reports (ACRs), 230 Annual Direct Recruitment Plan (ADRP), 104 Appointments Committee of the Cabinet (ACC), 227 asset reconstruction companies (ARCs), 139

audits, 169, 250, 251, 378 arrangements, 169 Article 149, 169 CAG, 170, 171, 249 corruption, 171 expanded role, 170 ex post facto**, 290 performance, 246 post facto**, 387 third-​party, 379 Banking Regulation Act 1949 Section 35A, 140 Section 35AB(2), 140 Bayesian updation, 298–​300 behavioural change, better policymakers, 178, 273 achievements, detachment from, 287 alternative worldview, 279 ambiguity, comfort with, 284–​286 anchoring biases free, 288–​289 bad ideas, no acceptance, 305 basic administration improvement, 305–​307 Bayesian updations, willingness to make, 298–​300 best-​practice institutions, 281 development and humility, 307–​308 draft solution, iterating with, 294–​295 features of IAS, 308–​309 follow rational scepticism, 300–​301 GIS-​based mapping of property, 281 holding multiple hypotheses, 283–​284 legitimate private benefit, 277–​280 logical solutions, 280–​281 negotiating policy choices, 296–​297 objects and agents, 274 partial and general equilibrium, 290–​292

416 Index behavioural change, better policymakers (cont.) personal effort/​systemic improvement, 304 predecessor-​successor syndrome, avoid, 303–​304 probabilistic outcomes, 297–​298 program implementation vs. policy making skills, 274 public policy space, perception of, 278, 279 tolerate failures, 301–​303 top-​down vs. bottom-​up perspectives, 289–​290 trust deficit, connecting, 275–​277 unintended consequences, law of, 292–​294 voluntary and constructive efforts, 310 big-​bang reforms, 3 bloated government, 89, 91 administrative units, 89, 101 diplomatic cores, strength of, 93 IAS officers, authorized vs. actual working, 92 inadequate safety standards, 96 investigating more, 91 law and order management, personnel strength in, 94 manpower deficiency, 98, 99 Ministries and Departments, expansion of, 89 origins, 89 pending cases in courts, 95 personnel, expansion of, 91 reality is different, 11–​117 urban planners shortage, 95 welfare role of government, expansion of, 90 British National Audit Office, 246 bureaucrats career progression and incentives, 8, 102 Chinese, 142 frontline, 183, 331 permanent, 339

street-​level, 26, 183, 214, 331, 332 transferable, 409 bus rapid transit (BRT), 257–​258 Business Rules/​Rules of Business, 33 capable state, 6 capacity utilization, 4 Central Civil Services, 55, 56–​57 Central Drugs Standard Control Organisation (CDSCO), 96 Central Public Sector Units (CPSUs), 352 Central Services, 53. See also civil service employees categories, 60 incentive structure, 78–​80 intermediate and lower services of, 60–​64 list, 56 modes of recruitment, 61–​63 recent trends affecting, 80–​85 recruitment examination, 55 specialist vs. generalist, 132 Central Staffing Scheme (CSS), 226 ACC, 227 objectives, 226–​227 postings suggested process, 227–​228 certification, 378, 380 agencies, 379 challenges, 379–​380 ISO boom, 378 ODF process, 379 civil service. See also Indian Administrative Service (IAS) assessing effectiveness, 67–​71 decision paralysis, 137 effectiveness, criteria of, 65–​67 higher, 53, 55 incentive structure, 78–​80 political executive, 36 poor performance, causes of (see political interference) recent trends affecting, 80–​85 structures, 135 weaknesses in policymaking, 126

Index  417 Civil Services Board (CSB), 221, 232 empanelment, problems with, 232–​233 lateral entry, 234 recommendation, 232 classic bureaucracy, 339 compassionate appointments, 63 Competition Commission of India (CCI), 150 Comptroller and Auditor General (CAG), 169–​171 banking, 170–​171 2G case, 170 Indian Audit and Accounts Service, 55 Indian Audit Department, 171 constitutional provisions, 33 Articles 77/​166, 33 Articles 53 and 154, 33 Articles 73 and 162, 34 executive on, 33–​35 ministerial responsibility, 35–​36 organizational structure, 36–​40 regulatory bodies, 41 consultants balanced approach, 341–​342 classic bureaucracy, 339 consequences of appointing, 338–​340 execution challenges, addressing, 336 government hiring, 336–​338 infrastructure sectors, 339 new ideologies, 339 role in government, 338 corruption legal/​procedural reform, 353–​358 meaning, 343 recruitment in, 349 decision, 163 delayed projects, 163 diligent, 167–​168 mala fide, 168 negligent, 168 poor investigative training, skills, and resources, 63–​64 decision paralysis, 9, 12, 16, 137, 199 apex regulatory authorities, 140

China’s bureaucratic system, 141–​142 comptroller and auditor general, 168–​171 constant flux in administrative law, 143 decisions of lower courts, 248 finance function, 172–​175 India policymaking, 137, 141 Indian judiciary, 142 overcoming, 245 Prevention of Corruption Act 1988, 246 problem, 137–​13 public sector banks, 138–​140 RTI, 2005, 137 stalled projects, 141 vigilance and investigative agencies (see Prevention of Corruption Act) delegation of authority, enhanced, 202 central/​state secretariats and departments, 203–​205 control, 206 district and local administration, 205–​207 Delivery Monitoring Unit (DMU), 327 delivery units (DUs), 324–​329 DMU, 327–​328 establishment, 32 features, 327 logical appeal, 325 Malaysia and Tanzania, 326 PEMANDU, 325 pioneer with, 325 PMDU, 235 PRAGATI, 328 Purpose, 329 Department of Personnel and Training (DoPT), 55 Development Monitoring and Evaluation Office (DMEO), 390 development outcomes determinants history of colonization, 5 institutions, 4 knowledge and expertise, 5 resource endowments, 5

418 Index digitization/​e-​governance, 112, 380–​385 limitations, 382–​384 solutions, 381 techniques, 381 dispute resolution, 51 disruptive change, 403 District Education Officer, 193 earnest money deposit (EMD), 363 ease of doing business (EoDB), 265 effectiveness, criteria of, 65–​67 effective service delivery, 182 Election Commission of India (ECI), 187 elections and census paradox bureaucracy, 190 ECI, 187 intensive field inspections, 189 macroeconomic and microeconomic outcomes, 186 mundane activities, 188 electronic monitoring systems, 313 conceptual challenges, 314–​318 design challenges, 318–​319 incentive distortions, 319–​321 MISs, 314 possible solutions, 321–​324 RTGS, 314 technology solutions, 314 empanelment, 76 enforcement-​harassment trade-​off, 342 data analytics, 346–​347 GST, 344 honest and corrupt officers, 343–​344 lessons, 348 matrix, 345 public agencies, 344 tightening, 346 evaluations case for, 385–​387 institutional architecture, 389–​391 limitations, 387–​389 RCTs and limitations, 393–​397 thorough and careful, 391–​393 execution challenges, 177 coordination vs. empowerment, 190–​195

elections and census, 186–​190 government’s routine activities, 180 implementation deficit, 177, 178 irrigation, 179–​180 systems change, 180–​182 (see also public service delivery) execution challenges, addressing account-​based accountability, 329–​330 certification, 378 certification agencies, 379 consultants and advisors, use of, 336 digitization/​e-​governance, 380–​385 DUs establishment, 324–​329 electronic monitoring systems, 313 enforcement-​harassment trade-​off, 342 enhanced delegation, 330–​332 evaluating programs (see evaluations) outsourcing and PPPs, 371 permits and licences, 366–​367 procurement, 351 social media, use of, 333 solution, 368–​369 supervision strengthening, 311–​313 third-​party audits, 379 transfers and postings, 349–​351 executive power, 34 Federal Aviation Authority (FAA), 93 Federal Reserve Open Market Committee (FOMC), 161 Fixed Cost Selection (FCS) method, 362 Food and Drug Administration (FDA), 96 framework agreement, 362 frontline problems 183 client power relationships, dysfunctional, 184 dysfunctional management, 184 empowerment and motivation, 186 public service delivery, 182 sustainable transformations, 185 technical side, 27 top-​down accountability, 184–​185 gazetted officers, 60 General Financial Rules (GFRs), 174, 352 gaps in, 360

Index  419 Geographic Information System (GIS), 282 good development outcomes, 6 good policy, 5, 6 good policymaking process, 121 characteristics, 121–​124 criteria, 124 trade-​ offs, 124–​126 Goods and Services Tax (GST), 259, 344 Government of India (GoI) DMEO, 390, 391 IEO, 390 organizational structure, 36–​40 policy making structure, 128 harassment, 345, 383 corruption, 345, 346 enforcement of laws, 342–​345 implementation deficit, 178 meaning, 343 trust deficit, 277 Independent Evaluation Office (IEO), 390 independent regulators, 41, 45–​46, 47 advantages and disadvantages, 48 primary rationale, 43 regulatory bodies, 41 India, implementation deficit, 177 diagnosis, 177 irrigation, 179–​180 Indian Administrative Service (IAS), 36, 38, 53, 54 all-​India character, 58 Article 311, 58 cadre posts, 55 design, 58–​60 distribution by service of senior officers, 56 dual control, 58 effectiveness, 65 higher civil services, 53 independent selection, 58 initial design, 59 mandarin-​type career civil service, 54 merit-​based selection, 58 Municipal Commissioner, 39

officers, 55 origin, 55 policymaking, role in, 63 protection against arbitrary punishment, 58 recruitment, 54, 55 special considerations, 59 Indian Audit and Accounts Service, 55 Indian banking system ARCs, 139 NPAs, 139 Indian Constitution, 55, 354 independent vs. administrative regulators, 49 Indian Foreign Service (IFS), 53 Indian Forest Service (IFS), 53 Indian Insolvency and Bankruptcy Code (IBC), 402 Indian Police Service (IPS), 53 Indian Revenue Service (IRS) generalist/​specialist, 133 Indian state, personnel of AIS, design of, 58–​60 higher civil services, 53–​55 intermediate and lower services, 60–​64 India’s policymaking, weaknesses in, 126 consultation inadequacy, in-​house specialist, 131 excessive fragmentation, 130 excessive overlap, 127–​128 extreme fragmentation in structure, 127 generalist/​specialist, 132–​135 intelligent generalist, 134 mediocrity of in-​house specialists, 131 non-​governmental inputs and informed debate, inadequacy, 129–​130 professionalism, inadequate, 130–​131 time management, lack of, 130 India’s Smart Cities programme, 401 informed generalist career progression, 223 need of, 130

420 Index Insolvency and Bankruptcy Code (IBC), 259 institutions, 4 Integrated Financial Advisor system, 174, 175 intermediate and lower services of, 60–​64 officers categories, 60–​64 judicial activism, 147 Harish Salve, blaming Supreme Court, 148–​149 judgements legislated, 147–​148 Sudeep Mahajan’s observations, 152–​154 Judicial power, 155 judiciary administrative decisions, 142 Article 136, 150 CCI, 150 doing business, 154–​155 features, 143 in-​built limitations, 155 judicial activism, 147 multiple benches, regulatory decisions, 146 routine administration, 149 single bench, regulatory decisions, 145 Suo Motu writ petitions, emergence of, 151–​152 knowledge and expertise, 5 Lal Bahadur Shastri National Academy of Administration (LBSNAA), 206 lateral entry case against, 234–​235 case for, 235–​236 IAS, into, 238–​239 objectives, 239–​243 practical solution, 236–​238 legislative and regulatory changes, 402 mala fide decision, 168 management information systems (MISs), 314

Million Marginal Revolutions of China, 267–​268 minimum viable product (MVP) approach, 254 government, 256–​259 GST, 259 IBC, 259 iteration, 14, 96, 225, 253, 261, 264 program design, 260 reasons to choose, 259–​261 ministerial responsibility Articles 75 (3)/​164 (2), 35–​36 National Accreditation Board for Hospitals and Healthcare Providers (NABH), 378 National Accreditation Board for Testing and Calibration of Laboratories (NABL), 378 National Assessment Survey (NAS), 401 National Audit Office (NAO), 364 National Company Law Appellate Tribunal (NCLAT), 146 National Company Law Tribunal (NCLT), 146 negligent decision, 168 new public management (NPM) approaches, 207–​210 features, 207–​208 OECDs, 209 OPD, 209 non-​performing assets (NPAs), 139 Office of Management and Budget (OMB), 211 characteristics of good performance goals, 211 Old Public Discipline (OPD), 209 Organisation for Economic Cooperation and Development’s (OECD’s), 209 outsourcing, 116, 306, 364, 371. See also public-private partnerships (PPPs) advantage, 115 challenges, 374–​375 market, 115 political interference, 373 pre-​requisites, 371–​372

Index  421 privatization, 116 range, 371 recommendation to governments, 377 success, 372, 375 overcoming decision paralysis, 245 CAG guidelines, 246 complex public policies, 263 ex-​post policy assessments, 246 hindsight, application of, 245 judicial action, 247–​248 MVP approach, 253 oversight agencies, 248–​252 policy decisions, 245 Prevention of Corruption Act 1988, 2018 amendment, 246 regulatory institutions, 252–​253 simplifying policy design, 268–​272 Performance Appraisal Reports (PARs), 230 grade inflation, 230 problem with, 230 Performance Assessment Rating Tool (PART), 211 Performance Management and Delivery Unit (PEMANDU), 325 performance-​based incentives, 210 administrative uncertainty principle, 213–​214 Bourgon’s finding, 210 credibility, 218–​219 cultural internalization, 216 cultural/​social acceptance, 215–​216 incentive distortions, 217 OMB, 211 public service delivery, moral issues with, 214–​215 quantification and measurement problems, 212–​213 scale dynamics, 219 permits and licences, 366–​367 personnel management all India services, promotions into, 229 appointments reforming, 226–​ 229 (see also Central Staffing Scheme [CSS])

career progression, 223–​226 lateral entry, 234 reforming empanelment, 229–​233 Policy Advisory Group, 207 policymaking and program design, 6. See also behavioural change, better policymakers; program implementation acceptance of failure, 301–​303 achievements and neutrality, 287 ambiguity, comfort with, 284–​286 baby ontology, 274 Bayesian updation, 298–​300 behavioural change, 273 best-​practices, 281–​283 challenges addressing (see structural reforms) challenges in, 119 choices negotiation, 296–​297 consultants, role of, 338, 340 debate, 120 disagreements, 120 evidence-​based, 393 experiences/​data points, judging correctly, 288–​289 flexibility, 25, 120 India underperform, causes, 119–​121 iterative approach, 294–​295 legalized private benefit, 277–​280 logical solutions not always, 280–​281 multiple hypotheses approach, 283–​284 outcomes, 297–​298 overlapping, 25 political considerations, 120 politics, 5 poor state capability, 8 program formulation, difference with, 24 proposed structure, 200 public policy design, 12–​14 scientific skepticism, 300–​301 trust deficit connections, 275–​277 vs. execution, 21 weakness in, 8–​9, 9–​12

422 Index political interference arbitrary transfers, 76 Article 311 of Constitution, 74 civil service, poor performance, causes of, 72–​73 design principle, 74 drastic pay compression, 77 dual control, 76 judicial interpretation, 77 non-​transparency, 76 politicization, 76 promotion, 74–​75 socio-​economic trends, 77 special reward, 75 sticks, 75 successful prosecutions, dearth of, 77 positive deviances, 399 encouraging, 399–​400 postings, 226 challenge, 227 CSB, 227 officers, 226 process, 227, 228 reform process, 226–​228 state and central government levels, 226 suggested process for CSS, 227–​228 Prevention of Corruption Act mala fide decision, 168 negligent decision, 168 Section 13(1)(d) (iii), 163–​168 price setting, 51 Prime Minister’s Delivery Unit (PMDU), 325 Private Finance Initiative (PFI) program National Audit Office assessment, 376 Pro-​Active Governance and Timely Implementation (PRAGATI), 328 Problem Driven Iterative Adaptation (PDIA) strategy, 263 procedural legitimacy, 122 procurement, 351 centralized earnest money, 363 deterrent penalties, 354 dispute control, 361–​362 efficiency issues, 359–​360 FCS method, 362

gaps in GRF, 360 information and RTI, 355 L 1 culture, 359 legal/​procedural reform and corruption, 353–​358 possible improvements, 360 problem in, 357 rate contract, 362–​363 reducing delay, 360–​361 rule framework, 352–​353 shared service centres, 363–​366 writ remedy, 354 Program Evaluation Office (PEO), 389 program implementation, 7, 273–​274. See also behavioural change, better policymakers; policymaking and program design achievements and neutrality, 287 ambiguity, comfort with, 284–​286 avoid successor syndrome, 303–​304 baby ontology, 274 basic administration improvement, 305–​307 Bayesian updation, 298–​300 behavioural change, 273 best-​practices, 281–​283 consultants, 338 evaluations, 391–​393 experiences/​data points, judging correctly, 288–​289 iterative approach, 294–​295 keep humility alive, 307–​308 legalized private benefit, 277–​280 logical solutions not always, 280–​281 multiple hypotheses approach, 283–​284 partial and general equilibrium, 290–​292 policy choices negotiation, 296–​297 policymaking and program design, difference with, 24–​25, 274 scientific skepticism, 300–​301 skills comparison, 274 strengthening supervision, 311–​312 structural factors, 192 trust deficit connections, 275–​277 unintended consequences, 292–​294

Index  423 project management unit (PMU), 328–​329 public financial management (PFM) systems, 17 Public Interest Litigation (PIL), 147 public policy, 263 action plan for implementing, 266 GoI is, 266 PDIA strategy, 263 Public Relations Officer (PRO), 335 public service delivery data management, 324 four principal, 182 frontline problems, 183–​186 systemic coherence, 182–​183 systems approach to, 180–​181 theoretical framework, 182 World Development Report 2004, 181 public-​private partnerships (PPPs), 371 challenges, 374–​375 conceptual merit, 372 metro railways, operations and maintenance of, 373 pre-​requisites, 371–​372 United Kingdom, 376 Randomized Control Trials (RCTs), 393 limitations, 393–​397 rate contract, 362–​363 ratings, 378, 380 Real Time Governance System (RTGS), 314 recruitment methods ad hoc, 63 contractual appointments, 63 Reforms Commissions, 351 regulatory bodies, 41 constitutional position, 42 independent regulators, 45–​46, 47 institutions, 50–​52 legal aspect influence, 46–​47 need for, 43–​45 traditional state, regulation through, 45 regulatory institutions, 50–​52

framework for analysis, 52overcoming decision paralysis, 252–​253 regulatory state, 41 reports four themes, 374 India’s fire safety preparedness, 95 India’s judicial system, problem in, 94 NAO, 364 Outsourcing, 377 Sixth Central Pay Commission, 104 United Kingdom’s Institute for Government, 371 resource endowments, 5 Right to Information (RTI) Act 2005, 137 activist auditors and courts, 157 China, excessive transparency in organization, 161 civil servants, 157 civil service behaviour, effect on, 156 civil service behaviour, effects of, 160 comptroller and auditor general (see Comptroller and Auditor General [CAG]) corporate sector, 159–​160 debate in United States, 158–​159 finance function, 172–​175 officers, 156 procurement, 355 sick people, 157 Supreme Court’s observations, 158 vigilance and investigative agencies (see Prevention of Corruption Act) School Education Quality Index (SEQI), 401 school education reform, 405–​407 Sixth Central Pay Commission Report, 104 social media disadvantages, 334 government officials and agencies, 333–​334 PRO, 335 suggestions on managing, 335–​336 use, of, 333

424 Index Special Leave Petitions (SLPs), 150 standards certification, 51 State Administrative Training Institutes (ATIs), 206 state autonomy, 19 state capability, 17. See also policymaking and program design addressing failures, 28–​29 alternative approach, 289 assessment, 20–​21 best-​practice technology, 283 building, 409 challenges and weaknesses, 87 conventional wisdom, 289 decision paralysis, 137 deficiencies, 182 defined, 4, 19, 21 determinant of development outcomes, 4–​8 dimensions, 21 district and local administration, 205–​207 erosion, reasons for, 10 execution challenges, 371 good development outcomes, 6 government branches, issues effect on, 14 implementation, weakness in, 9–​12 importance, 19 law and order management, 93 partial and general equilibrium, 290–​292 personnel, 53 PFM, 17 policymaking and program design, weakness in, 8–​9 political process, 5 politicians and ruling parties, 102 problem, 93 SDGS, 17 size of Indian state (see bloated state) top-​down vs. bottom-​up perspectives, 289–​290 unintended consequences, 292–​294 weak, 116, 192, 388, 395 state governments employees categories, 60

intermediate and lower services of, 60–​64 modes of recruitment, 61–​63 structural reforms CSB creation, 221 curb post-​retirement opportunities, 221 enhanced delegation of authority, 202–​207 fragmentation, reduction in, 199–​201 non-​governmental inputs, 207 NPM approaches, 207–​210 performance-​based incentives, 210 personnel management, 223 restrict civil service entry, 220 separating from implementation, 201–​202 specific skills training, 221–​223 Street-​level bureaucrats, 26, 181, 213, 329, 330 substantive legitimacy, 123 Such Suo Motu writ petitions, 152 supervision DMU, 327 (see also delivery units [DUs]) electronic monitoring systems, 313 traditional methods, improvement, 311–​313 Supreme Court of India directions in administrative matters, 151 SLPs, 150 Sustainable Development Goals (SDGs), 17 Swachh Bharat Mission (SBM), 265 taking charge behaviour (TCB), 310 Tata Administrative Services generalist/​specialist, 133 thin state, causes personnel management, litigation on, 105–​109 possible solutions, 111–​115 recruitment, long delays in, 109 unpredictable and hidden fiscal costs of personnel, 109–​111 third-​party agencies, 378 third-​party audits, 379

Index  425 transfers and postings, 349–​351 personnel management, 349 stability of tenure, 351 transparency, 351 Union government, 53 Union Public Service Commission (UPSC), 53 United Nations Commission for International Trade Law (UNCITRAL) model, 353, 354 inadequate safety standards, actions against, 96 revision, 357

vigilance and investigative agencies, 162–​168 diligent decision, 167–​168 mala fide decision, 168 negligent decision, 168 Prevention of Corruption Act, 162 rules of thumb, 165–​166 wrong tendency, 164 Weberianness scale, 20 World Bank’s generalist/​specialist, 133 writ remedy, 354