Maize Seed Industries in Developing Countries 9781685851675

Citation preview

Maize Seed Industries in Developing Countries

Maize Seed Industries in Developing Countries

edited by Michael L. Morris

^

LYNNE RI ENN ER publishers

II

CIMMYT

Published in the United States of America in 1998 by Lynne Rienner Publishers, Inc. 1800 30th Street, Boulder, Colorado 80301 and in the United Kingdom by Lynne Rienner Publishers, Inc. 3 Henrietta Street, Covent Garden, London WC2E 8LU Published in association with the International Maize and Wheat Improvement Center (CIMMYT) Apdo. Postal 6-641, 06600 Mexico, D.F., Mexico © 1998 by CIMMYT. All rights reserved by the publisher

Library of Congress Cataloging-in-Publication Data Maize seed industries in developing countries / edited by Michael L. Morris, p. cm. Includes bibliographical references and index. ISBN 1-55587-789-3 (hardcover : alk. paper). — ISBN 1-55587-790-7 (pbk. : alk. paper) 1. Seed industry and trade—Developing countries. 2. Corn industry—Developing countries. 3. Corn—Developing countries— Seeds. 4. Corn—Varieties—Developing countries. 5. Green Revolution. I. Morris, Michael L. HD9019.S432D4457 1998 338.1731521'091724—dc21 97-40583 CIP British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library.

Printed and bound in the United States of America The paper used in this publication meets the requirements (S3) of the American National Standard for Permanence of Paper for Printed Library Materials Z39.48-1984. 5 4 3 2 1

Contents

List of Tables and Figures Acknowledgments

vii xi

Part 1 The Importance of Maize Seed Industries 1 2 3

Maize in the Developing World: Waiting for a Green Revolution Michael L. Morris

3

Overview of the World Maize Economy Michael L. Morris

13

Maize Seed Industries: A Conceptual Framework Michael L. Morris, Joseph Rusike, and Melinda Smale

35

Part 2 From Breeders' Plots to Farmers' Fields: Functions and Processes 4 5 6 7 8

Varietal Development: Conventional Plant Breeding Shivaji Pandey

57

Varietal Development: Applied Biotechnology David Hoisington, G. Michael Listman, and Michael L. Morris

77

Seed Multiplication, Conditioning, and Storage P. K. Agrawal, B. D. Agrawal, P. Venkat Rao, and Jai Singh

103

Seed Marketing and Distribution C. F. Krull, J. M. Prescott, and C. W. Crum

125

Economics of Hybrid Maize Adoption Paul W. Heisey, Michael L. Morris, Derek Byerlee, and Miguel A. Lopez-Pereira

143

v

vi

CONTENTS

9 Regulatory Issues: Varietal Registration and Seed Quality Control Robert Tripp 10 Industrial Policy Issues: Participation, Prices, and Property Rights Carl Pray and Robert Tripp

159

175

Part 3 Country Case Studies 11 The United States Donald N. Duvick

193

12 Brazil Joao Carlos Garcia

213

13

Mexico Pedro Aquino

231

14 India Suresh Pal, R. P. Singh, and Michael L. Morris

251

15 Thailand Michael L. Morris

269

16 17 18

Malawi Joseph Rusike and Melinda Smale

285

Zimbabwe Joseph Rusike

303

Turkey Carl Pray

321

19 China Carl Pray, Scott Rozelle, and Jikun Huang

335

Part 4 Summary and Conclusions 20 Future Directions for National Maize Seed Industries Michael L. Morris

355

References The Contributors Index About the Book

373 389 391 401

Tables and Figures

Tables 2.1 2.2 2.3 2.4 2.5 3.1 5.1 5.2 6.1

7.1 7.2

8.1 8.2 12.1 12.2

World Cereals Production Statistics, 1 9 9 3 - 1 9 9 5 Distribution of Maize Production Environments, Early 1990s Adoption of Improved Maize Germplasm (OPVs and Hybrids), Early 1990s World Maize Production Statistics, 1 9 9 3 - 1 9 9 5 World Maize Utilization Statistics, 1 9 9 2 - 1 9 9 4 Characteristics Associated with the Stages of Maize Seed Industry Development Q T L Mapping Studies in Maize for Selected Traits Importance of Maize Marker Systems for Molecular Genetic Applications Components of Double-Cross Hybrid Raw Seed Production Costs Reported by Private Seed Companies in Brazil and Mexico, Mid-1990s Maize Seed Prices, Hybrid Use, and Average Yields, Selected Developing Countries, Early 1990s Components of Retail Selling Price of Hybrid Maize Seed Produced by Private Seed Companies, Selected Countries, Mid-1990s Yields of Local Maize Varieties and Maize Hybrids in Unfertilized On-Farm Demonstrations, Malawi, 1 9 9 0 - 1 9 9 3 Effects of Variables That Affect Technology Demand and Supply on Area Planted to Hybrid Maize Area, Yield, and Production of Maize, by Region, Brazil, 1993-1995 Hybrid Maize Seed Prices and Numbers of Cultivars, Center-South Region, Brazil, 1995 vii

14 16 17 21 23 50 87 90

115 132

133 147 156 215 225

viii 12.3 13.1

14.1 14.2 14.3 14.4 14.5 14.6 14.7 15.1 15.2 15.3 17.1 18.1 18.2 18.3 18.4 19.1 19.2 19.3 19.4

TABLES AND FIGURES

Sales of Publicly and Privately Produced Maize Seed, by Type of Seed and Price Range, Brazil, 1995 Production of Improved Maize Seed by Public, Private, and Voluntary Organizations, Mexico, 1988-1995 Area, Yield, and Production of Maize, India, 1 9 5 0 - 1 9 9 2 Annual Growth in Area, Yield, and Production of Maize, India, 1 9 5 0 - 1 9 9 2 Public and Private Investment in Maize Research, India, 1993 Cost of Producing Double-Cross Hybrid Maize Seed, India, 1992 Seed-to-Grain Price Ratios in India and Selected Asian Countries, Early 1990s Public and Private Seed Sales, India, 1 9 8 1 - 1 9 8 3 to 1990-1992 Maize Area Under Improved Germplasm, India, 1995 Area, Yield, and Production of Maize, and Net Exports, Thailand, 1 9 6 0 - 1 9 9 5 Sources of Maize Seed, Thailand, 1 9 8 0 - 1 9 9 4 Maize Seed Prices, Thailand, 1994 Area, Yield, and Production of Maize, Smallholder and Commercial Farming Sectors, Zimbabwe, 1 9 6 5 - 1 9 9 5 Growth in Maize Area, Yield, and Production, Turkey, 1966-1995 Hybrid Maize Seed Production, Imports, Exports, and Maize Area, Turkey, 1 9 8 0 - 1 9 9 3 Maize Seed Production Plans, by Type of Seed Company, Turkey, 1987 and 1994 Hybrid Maize Seed Prices and Subsidies, Turkey, 1982-1993 Maize Imports and Exports, China, 1 9 8 5 - 1 9 9 5 Seed-to-Grain Price Ratios for Single-Cross Hybrid Maize Seed, China and Selected Developing Regions, 1990s Seed-to-Grain Price Ratios for Cereals, Zhejiang and Shandong, China, 1988 and 1995 Decomposition of Hybrid Maize Seed Costs, Selected Countries, Early 1990s

228

240 252 252 255 260 261 261 262 271 279 279 305 322 326 327 332 341 345 346 349

Figures 2.1 2.2

Distribution of World Maize Production, Early 1990s World Maize Production, 1 9 6 1 - 1 9 9 5

15

21

TABLES AND FIGURES

2.3 3.1 3.2 4.1 4.2 5.1 5.2 6.1 8.1

8.2 11.1 11.2 13.1 13.2 13.3 13.4 16.1 16.2 16.3 17.1 17.2 19.1 19.2

International Reference Price of Maize, 1961-1995 Universe of Organizational Types Showing Selected Seed Industry Organizations Universe of Organizational Types Showing Location of Selected Classes of Products Morphology of Maize Production of Different Kinds of Maize Hybrids Efficiency of Conventional Backcrossing Versus Marker-Assisted Selection Basic Genetic Engineering Process Common Planting Patterns Used in Production of Hybrid Maize Seed Yield Advantage over Farmer's Current Cultivar Required to Repay Additional Cost of Hybrid Seed and Generate 100% Marginal Rate of Return Hybrid Maize Seed-to-Grain Price Ratio, United States, 1940-1990 Number of Maize Seed Companies Founded or Reorganized, United States, 1925-1995 Adoption of Hybrid Maize, United States, 1930-1960 Maize Area, Yield, and Production, Mexico, 1961-1995 Sales of Improved Maize Seed, by Sector, Mexico, 1970-1995 Hybrid Maize Seed-to-Grain Price Ratio, Mexico, 1970-1995 Maize Seed Prices, by Source and Type of Seed, Mexico, 1995 Maize Area, Yield, and Production, Malawi, 1961-1995 Per Capita Maize Consumption, Malawi, 1961-1995 Diffusion of Hybrid Maize, Malawi, 1981-1997 Sales of Hybrid Maize Seed, Zimbabwe, 1950-1995 Adoption of Hybrid Maize, Zimbabwe, 1950-1990 Area, Production, and Yield of Wheat, Rice, and Maize, China, 1975-1995 Rice, Wheat, and Maize Prices, by Market, China, 1978-1995

ix

29 47 48 60 74 88 92 110

149 154 200 204 233 242 244 245 287 288 290 310 314 338 340

Acknowledgments

No book—especially no edited book such as this—can be completed without the help of many organizations and individuals. Although space limitations preclude mentioning all those who contributed, several deserve particular mention. First and foremost, I would like to thank the International Maize and Wheat Improvement Center (CIMMYT) for sponsoring the study leave during which most of the book was written and for committing substantial resources to its publication. My gratitude goes also to the International Food Policy Research Institute, which provided an enjoyable and intellectually stimulating environment in which to work during my study leave. On a more personal level, I owe a special debt to the contributing authors, who not only took time from their busy schedules to write sections of the manuscript, but also tolerated my sometimes less-than-subtle efforts to push and prod their writing into a consistent and coherent whole. Many current and former colleagues at CIMMYT provided information, answered questions, and reviewed drafts of the manuscript, including Greg Edmeades, David Beck, Javier Betrán, Surinder Vasal, Jim Lothrop, Suketoshi Taba, R. W. Wedderburn, Hugo Cordova, Daniel Jeffers, Martha Willcox, Natasha Bohorova, Mireille Khairallah, Daniel Grimanelli, Alma McNab, Prabhu Pingali, Jim MacMillan, Larry Harrington, Ricardo Calvo, Victor Hernández, and John Woolston. David Lambert and Robert Falasca of the American Seed Trade Association (ASTA) granted me access to ASTA archival materials. James Shaffer and A. Allan Schmid of Michigan State University provided useful comments on the conceptual framework chapter. Representatives of literally hundreds of public and private seed companies interviewed during various national seed industry surveys were unfailingly patient in answering what at times must have seemed like ignorant (but I hope not impertinent) questions. The manuscript was edited at CIMMYT by Kelly Cassaday. Miguel Mellado supervised the production of figures, which were prepared by xi

xii

ACKNOWLEDGMENTS

Marcelo Ortiz, Juan José Joven, and Eliot Sánchez. Liliana Santamaría provided secretarial and administrative support. Sally Glover and Lesli Brooks Athanasoulis at Lynne Rienner Publishers coordinated the publication process. Despite having gone to great lengths to ensure the production of a comprehensive, factually correct, and well-balanced book, the controversial nature of the subject matter makes me suspect that not everybody will be satisfied by the final result. I gladly accept responsibility for whatever errors and deficiencies remain. —Michael L. Morris

PART 1

The Importance of Maize Seed Industries

1 Maize in the Developing World: Waiting for a Green Revolution MICHAEL L . MORRIS

Of all the inputs used in agriculture, none has the ability to affect productivity as much as seed. A seed is a living organism that carries the genetic properties of plants. These genetic properties place an upper limit on yield potential and influence the productivity of other inputs by determining the ability of plants to convert sunlight, water, air, soil, and other nutrients into biomass. Because plants vary in terms of their ability to perform this conversion, seed choice is critically important in agriculture. If farmers can obtain seed of cultivars that perform well under local conditions, the efficiency with which other inputs are converted into economically valuable outputs increases, and productivity rises. At the same time, improved seed can make a contribution to productivity independent of other inputs. Thus, it is not surprising that technical change in agriculture frequently is driven by changes in crop varieties, which in turn depend on farmers having reliable access to improved seed. Looking back over the global history of crop varietal changes, it is apparent that maize has followed a very different path from that followed by other leading cereals. The so-called green revolutions in rice and wheat are by now well-known. During the late 1960s and early 1970s, improved semidwarf varieties of rice and wheat were introduced into some of the developing world's most populated countries. When grown with increased levels of fertilizer and an assured water supply, these modern varieties (MVs) performed significantly better than the traditional varieties (TVs) they replaced, leading to substantial production increases and higher incomes for millions of farmers who adopted the technology. 1 Following their introduction, rice and wheat MVs spread rapidly throughout many of the irrigated zones where rice and wheat cultivation was concentrated, leading to dramatic increases in food production in a number of countries that a short time earlier had been haunted by prospects of widespread famine. In subsequent years, rice and wheat MVs spread gradually into 3

4

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

less favorable environments, including many nonirrigated areas with relatively modest production potential. Meanwhile, in the irrigated zones that had first been affected, the original MVs were replaced by second- and third-generation MVs, helping to fuel slow but steady yield increases on the order of 1% per year that have continued to this day. In the case of maize, a phenomenon equivalent to the green revolutions in rice and wheat has not yet occurred, at least not in most of the developing countries. Despite extensive efforts to promote improved production technologies for maize, about half of the developing world's maize area continues to be planted to TVs that have never received the attention of a formal plant breeding program. The proportion of maize area that is irrigated lags behind that of rice and wheat, and use of fertilizer and other purchased inputs on maize remains relatively modest. The stark contrast between the experiences of rice and wheat on the one hand and the experience of maize on the other is particularly puzzling in view of what has happened in the industrialized world, where improved maize hybrids have been extensively adopted and use of purchased inputs on maize is high. If much of the developing world has thus far failed to experience a green revolution in maize, it is not because improved production technologies have been lacking. Investment in maize research has been extensive, and this investment has produced results. Plant breeders have developed many improved varieties and hybrids that clearly outperform the materials grown by farmers. Similarly, crop management specialists have identified improved management practices capable of significantly boosting productivity. Yet even though these technologies have shown marked superiority under experimental conditions, many have failed to spread beyond demonstration plots. Thus, the problem is not that improved maize varieties and hybrids are lacking, as well as the improved management practices that allow them to express their full genetic potential, but rather that the improved technologies that are available have not reached the farm level. Why this is so, and what can be done to improve the situation, is the subject of this book.

Key Role of Maize Seed Industries Maize (Zea mays L.) differs from rice and wheat in several respects that have important implications for the development and spread of improved germplasm. Because they are self-pollinating crops, when rice and wheat reproduce, each generation of plants retains the essential genetic and physiological identity of the preceding generation. This means farmers can safely replant rice and wheat seed harvested from their own fields, giving them effective control over the technology embodied in improved

MAIZE IN THE DEVELOPING WORLD

5

germplasm. Farmers can set aside a portion of their harvest for use as seed in future cropping seasons, as long as they are careful to avoid mixing seed of different varieties. Furthermore, if they so choose they can easily distribute seed to other farmers. This is precisely what happened during the green revolutions in rice and wheat: After relatively small quantities of seed were released by public breeding programs, rice and wheat MVs quickly spread through farmer-to-farmer seed exchanges, with relatively little involvement by any sort of formal seed industry. Maize presents a different story, however. Because it is a naturally cross-pollinating crop, when maize reproduces, much depends on whether the pollen used to fertilize a given kernel comes from the same plant or from a different plant. Unlike rice and wheat plants, when maize plants self-fertilize, the resulting progeny are often characterized by undesirable traits, such as reduced plant size and low yields. 2 But when maize plants cross-fertilize, some of the resulting progeny tend to demonstrate desirable traits, such as increased plant size or high yields. Commonly referred to as "hybrid vigor," this phenomenon is attributable to the complementary action of favorable genes and is frequently exploited by plant breeders in their efforts to develop commercial materials. Unfortunately, the benefits of hybrid vigor do not persevere across generations. When seed harvested from cross-pollinated maize plants (known as Fj hybrids) is replanted, performance in the resulting progeny (known as F 2 hybrids) decreases because of the inbreeding phenomenon referred to earlier. The size of the decrease in performance depends on the nature of the original F, hybrids. If the Fi plants were single-cross hybrids formed from highly homozygous inbred lines, the yield loss in the F 2 generation may be as great as 35^40%. If the Fj plants were double-cross hybrids formed from less highly inbred parents, the yield loss will be more modest, usually on the order of 10-15%. 3 These two characteristics of maize—its tendency to deteriorate through inbreeding and its ability to demonstrate hybrid vigor—affect the degree of control exercised by farmers over the technology embodied in seed. Because of the decrease in performance between F! plants and subsequent generations, farmers who choose to grow hybrids in effect forfeit the option of saving a portion of their harvest to use as seed in the following cropping cycle. 4 Farmers who choose to grow open-pollinated varieties retain the option of saving a portion of their harvest to use as seed in the following cropping cycle, but they must be careful to maintain the genetic purity of successive crops grown from replanted seed (e.g., by physically isolating plots and by staggering planting dates to prevent cross-pollination between varieties). With on-farm seed production either technically difficult (in the case of hybrids) or inconvenient and costly (in the case of varieties), maize farmers must acquire fresh seed for each planting if they want to be certain

6

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

of maintaining a high level of genetic purity in their crops. In and of itself, the fact that maize farmers depend on external sources of seed does not necessarily affect the development and diffusion of improved germplasm. On the face of it, there seems to be no good reason maize varietal development, seed production, and seed distribution activities cannot be performed by specialized organizations. Nevertheless, the empirical record is not encouraging, at least not in most developing countries, where decades of sustained investment in maize research and development activities have resulted in relatively modest impacts. The implication is that past approaches have not worked very well and may have to be modified. For better or for worse, in many countries the need to rethink maize research and development strategies comes at a time when public support for agriculture is stagnating or even declining. Forced to re-evaluate the desirability of continuing long-standing government spending programs, policymakers frequently have chosen to relieve pressure on overburdened public-sector budgets by encouraging the privatization of selected activities traditionally carried out by public agencies. Plant breeding research, which offers obvious opportunities for commercial gain, has often been among the first activities targeted for privatization. Judging from the experience of many industrialized countries, the maize seed industry indeed appears to be a suitable candidate for privatization. In the United States, Canada, and most of the member states of the European Community in which maize is an important crop, maize varietal development, seed production, and seed distribution are now performed largely by private firms. The industry is populated by large transnational corporations whose names are familiar: Pioneer, Sandoz, DeKalb, Limagrain, Novartis, ICI-Zeneca, Cargill. The seed divisions of these corporations have grown and prospered because they have been able to identify the germplasm needs of commercial farmers, develop hybrids that successfully meet those needs, produce high-quality seed, and distribute that seed at attractive prices. The undeniable success of many seed companies has heightened the widespread perception that private companies are well equipped to handle maize varietal development and seed production. In many developing countries, the result has been a groundswell of support for privatizing the maize seed industry. Not everybody is comfortable with the idea of a global maize seed industry consisting exclusively of private companies, however. Some observers argue that privatization poses a risk because the quest for profits could lead private seed companies to act in ways that are not necessarily in the best interest of all producers and consumers. Private companies clearly have economic incentives to serve the needs of large-scale commercial producers who represent a potentially lucrative market for seed, but they probably have much less reason to address the needs of small-scale subsistence farmers who may be unable or unwilling to purchase seed regularly.

MAIZE IN THE DEVELOPING WORLD

7

The lack of success achieved by public-sector organizations in bringing about a green revolution in maize suggests that it may be desirable to encourage private companies or nongovernmental organizations to play a greater role in disseminating improved maize varieties and hybrids, particularly in the developing world. Complete withdrawal of public organizations from the maize seed industry, however, would probably be undesirable for several reasons. First, private firms are unlikely to engage in activities that offer no prospects for short- or medium-term profits, including germplasm collection and conservation, as well as many types of basic research. Furthermore, private seed companies will tend to neglect small-scale farmers in marginal production environments who, because of their dispersed distribution, special germplasm requirements, and modest purchasing power, do not represent an attractive market. Thus, increased privatization will have to be encouraged in ways that ensure that the needs of the widest possible range of producers and consumers are effectively addressed. Privatization of the maize seed industry will fail if the old state monopolies are simply replaced by new private ones.

Objectives Following a long period of relative stability, the world food economy has begun to change in ways that are likely to have important consequences for millions of producers and consumers of maize. In many developing countries, private seed companies are beginning to assume a leading role in generating improved maize varieties and hybrids, producing commercial seed, and distributing that seed to large numbers of farmers. Meanwhile, the government research institutes and state seed agencies that have traditionally performed those functions are often being forced to scale back their level of activity. As a result, responsibility for the development and dissemination of improved production technologies for maize—which in earlier times would have been considered too important to entrust to the private sector—has begun to shift away from the public domain. Thus far, moves to privatize national maize seed industries seem to have had largely positive results, as evidenced in many countries by recent dramatic growth in the adoption of hybrids, expansion in the area planted to improved germplasm, and a surge in national average maize yields. Yet despite these encouraging signs, it is legitimate to ask where the changes will ultimately lead. If current trends continue, will maize farmers in developing countries always be able to obtain a wide range of improved germplasm and related production technologies at affordable prices? Will consumers always have access to adequate supplies of maize at affordable prices? Will efficient forms of industrial organization automatically emerge for carrying out seed research, seed market development, seed production,

8

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

and seed marketing? Can input and output markets be relied upon to function well enough to ensure that innovation and entrepreneurship are adequately and fairly rewarded? Will legal systems see to it that unlawful and unethical expropriation of intellectual property is effectively discouraged? Must governments actively engage in maize research and seed production to ensure socially desirable outcomes, or will it be sufficient if they merely establish and enforce the rules of the game? These are important questions. In an age when further increases in maize production are likely to become more and more difficult to achieve, it will be necessary to create conditions that favor the generation and dissemination of improved maize production technologies in ways that are both efficient and equitable. With the government resources available to support agriculture becoming increasingly scarce, ways must be found to encourage increased private-sector participation in the seed industry. This participation, however, must be structured in a way that is conducive to socially acceptable outcomes. An extreme laissez-faire approach to seed industry reform based on wholesale privatization of research and development activities would be undesirable if blind reliance on market mechanisms fails to protect the interests of poor producers and consumers. This book draws on the experience of researchers, seed industry representatives, and policymakers in examining the organization and performance of maize seed industries in the developing world. Intended as a comprehensive reference for readers with an interest in agricultural development issues in general and maize in particular, the book seeks to describe and examine key technical, economic, and institutional issues that must be resolved for maize seed industries to function well. Based on a detailed technical and economic analysis of the activities carried out by the maize seed industry, an effort is made to distinguish between goods and services that are likely to be provided effectively and efficiently by private firms operating in a market-oriented environment and goods and services whose provision is likely to require other organizational structures. Case studies are presented from a number of developing countries to illustrate how certain institutional arrangements appear to be more effective than others at fostering rapid and equitable growth.

Organization This book is divided into four parts. Part 1 lays out a conceptual framework that provides a set of analytical tools to help guide the subsequent discussion. Part 2 includes seven chapters organized along thematic lines whose purpose is to describe key products and processes associated with maize seed industries, including varietal development, seed production and conditioning, seed marketing and distribution, adoption of improved

MAIZE IN THE DEVELOPING WORLD

9

germplasm, and industrial regulation. Part 3 includes nine country case studies selected to illustrate some of the many developmental paths (successful as well as unsuccessful) that have been followed by maize seed industries throughout the world. Part 4 builds on the findings of earlier chapters in discussing the likely future course of seed industry development and in examining the relationship between alternative organizational forms and probable performance of the industry. To provide an idea of the context in which maize seed industries operate, Chapter 2 presents an overview of the world maize economy, beginning with a description of the principal growing environments and the dominant production technologies found in industrialized and developing countries. Recent production and consumption trends are then summarized, along with trends in international maize trade. The chapter concludes with a short description of the global maize seed industry and a brief discussion of the future prospects for the world maize economy. Chapter 3 describes a conceptual framework for examining maize seed industries. Distinctive characteristics of maize seed are described that influence the economic incentives facing producers and consumers, thereby affecting the optimal organization of the industry. The chapter concludes by considering the process of industrial growth, focusing on the critical question of whether industrial evolution is deterministic (meaning that maize seed industries everywhere pass through the same predictable stages of development) or essentially eclectic (implying that in each country, the seed industry undergoes a unique and largely unpredictable series of transformations). Chapter 4 describes biological and genetic aspects of maize and summarizes technical aspects of modern crop improvement programs. The material, intended primarily for readers who are unfamiliar with plant breeding methods, is included to provide a sense of what is involved in applied maize breeding programs. Chapter 5 reviews recent developments in the field of biotechnology and discusses how they are likely to affect applied maize breeding in the short to medium run. Topics treated in particular detail include the use of molecular marker techniques to enhance selection in conventional plant breeding programs and genetic transformation techniques involving the introduction of alien genes into maize. Chapter 6 discusses technical and economic dimensions of commercial maize seed production, including seed multiplication, seed processing, and seed storage. Procedures involved in multiplying different types of commercial seed are explained (e.g., open-pollinated varieties, doublecross hybrids, three-way-cross hybrids, single-cross hybrids), as are the steps involved in processing and storing commercial seed. Chapter 7 discusses technical and economic dimensions of maize seed distribution, with emphasis on the marketing strategies commonly pursued

10

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

by public and private seed companies. Conventional seed distribution strategies are described, including strategies frequently used in developing countries at the wholesale and retail levels. Methods used by seed companies to capture and retain market share are discussed, as are the strategies used to differentiate branded products in the marketplace. Chapter 8 analyzes technical and economic factors affecting the adoption of improved varieties and hybrids. Availability of seed is a necessary condition for adoption, but it is not a sufficient condition: In the absence of adequate economic incentives, farmers will fail to adopt improved materials. The microlevel basis for the farmers' adoption decision is analyzed, and technical and economic factors are identified that determine whether farmers will adopt improved varieties and hybrids. An attempt is made to distinguish between factors that can be influenced by the seed industry and factors that must be considered exogenous. Chapter 9 discusses important regulatory issues affecting maize seed industries. Regulations may be necessary to protect consumers whose knowledge of or access to information may be limited, to protect society from undesirable consequences that can arise if individuals and organizations are given free rein to pursue their own objectives, and to stimulate industry development. Regulation, however, may also fall prey to bureaucratic inertia or be susceptible to capture by powerful interest groups. This chapter reviews the cases for and against common forms of seed industry regulation, including varietal registration and release and seed quality control. Likely costs and benefits of alternative regulatory mechanisms are evaluated, and distributional impacts are discussed. Chapter 10 focuses on industrial policy issues. For various reasons, governments the world over have shown a tendency to enact policies designed to influence the behavior of commercial seed companies, control national and international flows of germplasm, manipulate seed prices, and regulate intellectual property rights. This chapter reviews the cases for and against selected industrial policies and assesses the likely costs and benefits of alternative policy measures. Chapters 11 through 19 consist of case studies from the United States, Brazil, Mexico, India, Thailand, Malawi, Zimbabwe, Turkey, and China. The country case studies illustrate some of the many different growth paths that have been followed by maize seed industries in the developing world and compares them with the path followed by the U.S. maize seed industry, the world's first maize seed industry and arguably the most successful. Although each country case study tells a story that is in certain respects unique, some historical features seem to recur over and over, lending weight to the idea that seed industries develop in ways that are evolutionary and to some extent predictable. Chapter 20 integrates important conclusions from the conceptually oriented thematic chapters with key findings from the empirically oriented

MAIZE IN THE DEVELOPING WORLD

11

case studies in drawing out practical lessons for researchers, industry representatives, and policymakers. Based on the characteristics of seed industry products and services, the strengths and w e a k n e s s e s of alternative organizational f o r m s are discussed. Public, private, and participatory organizations all appear to have important roles to play in the future.

Notes 1. The term modem varieties as used here refers to semidwarf varieties of rice and wheat developed since 1960. Traditional varieties refers to older varieties, including many that have never been worked on by a formal plant breeding program. As Derek Byerlee (1994) has pointed out, the term modern variety is something of a misnomer, since some MVs are now more than 30 years old. It is preserved here, however, to maintain consistency with other publications on the subject. The term high-yielding varieties, often used in referring to the same varieties, is equally inaccurate, since many of these varieties were bred for characteristics other than yield potential. 2. Because of this tendency to deteriorate through inbreeding, maize varieties must reflect a considerable amount of genetic diversity to retain their desirable characteristics. Individual plants of a given variety grown within the same field are actually complex hybrids among genetically similar biotypes. 3. For a more detailed discussion of maize breeding practices and the formation of different types of hybrids, see Chapter 4. 4. On-farm production of hybrid seed is simply not a realistic option for most farmers. Even if they can master the technically difficult task of producing hybrid seed, the parent lines are rarely available; when they are available, they tend to be difficult to manage and maintain.

2 Overview of the World Maize Economy MICHAEL L . MORRIS

Maize is one of three cereals that dominate the world grain economy; the others are wheat and rice. In industrialized countries, / jj maize ranks second only to wheat in area planted and production (Table 2.1). In developing countries, where most of the world's rice production is concentrated, maize ranks third behind rice and wheat. Unlike wheat and rice, which are mainly consumed as human food, maize is a multipurpose crop that can be eaten by humans, fed to animals, or used as a raw input into industry. Most maize produced in industrialized countries is used as animal feed or for industrial purposes, but maize remains an important food staple in many developing regions, especially sub-Saharan Africa and Central America, where it is frequently the mainstay of human diets. Considering the wide geographical dispersion of maize, it is not surprising that the productivity of maize production systems varies from place to place and among groups of producers. Although productivity can be expressed in different ways, one commonly used measure is grain yield per unit land area. 1 With good management, commercial maize grain yields often reach 10 tons per hectare (t/ha) or more in favorable production environments such as the "corn belts" of the United States and Western Europe. In contrast, subsistence farmers in marginal areas in Central America, sub-Saharan Africa, and Asia frequently harvest 0.5 t/ha or less of grain from their maize plots. The variability in yields can be attributed to environmental, technological, and institutional factors that determine the physical potential of the crop, influence the availability of yield-enhancing technology, and affect economic incentives to adopt improved technology.

Overview of World Maize Production Maize Production

Environments

Maize is believed to have originated in North America about 6,000 to 7,000 years ago (Smith 1995). Domesticated maize was apparently developed by

13

14

Table 2.1

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

World Cereals Production Statistics, 1993-1995 Area (million ha)

Developing countries Rice Wheat Maize Sorghum Industrialized countries Rice Wheat Maize Sorghum World Rice Wheat Maize Sorghum

Yield (t/ha)

Production (million t)

142.3 102.2 89.4 39.4

2.41 a 2.46 2.70 1.07

343.2 a 251.6 241.3 42.1

4.5 117.6 45.6 4.5

3.79 a 2.49 6.11 3.53

17.0a 292.7 278.7 15.9

146.8 219.9 135.1 43.9

2.45 a 2.48 3.85 1.32

360.1 a 544.3 519.9 58.0

Source: FAO Agrostat database, Primary Crops Production. Note: a. Milled rice.

the inhabitants of present-day Mexico from teosinte, an annual wild grass. From its center of origin, maize spread gradually throughout the Americas and later was carried by European seamen to Europe, Africa, and Asia. Today, maize is the world's most widely grown cereal, reflecting its ability to adapt to a wide range of production environments. Maize is cultivated at latitudes ranging from the equator to approximately 50° North and South, at altitudes ranging from sea level to over 3,000 meters (m) elevation, under temperatures ranging from extremely cool to very hot, under moisture regimes ranging from extremely wet to semiarid, on terrain ranging from completely flat to precipitously steep, in many different types of soil, and using a wide range of production technologies (Figure 2.1). No universally recognized system exists for classifying maize production environments. The closest thing to a global classification system has been developed by the International Maize and Wheat Improvement Center (CIMMYT), which holds the global mandate for maize germplasm improvement in developing countries. CIMMYT recognizes five major maize production environments, known as megaenvironments: (1) lowland tropics, (2) subtropics, (3) mid-altitude tropical zones, (4) tropical highlands, and (5) temperate zones. These five megaenvironments, which are defined primarily in terms of climatic factors (e.g., mean temperature during the maize growing season, elevation above sea level, day length), are characterized theoretically by their relative within-class uniformity. Since the growth habits of maize plants are influenced by complex interactions among many different climatic factors, however, it is not always clear exactly where one megaenvironment ends and the next begins.

OVERVIEW OF THE WORLD MAIZE ECONOMY

15

Figure 2.1 Distribution of World Maize Production, Early 1990s

Source: Adapted from Espenshade, Hudson, and Morrison (1995).

In considering the relative importance of the five megaenvironments, it is important to note a fundamental dichotomy between where maize grows in industrialized compared to developing countries. Over 90% of the maize produced in industrialized countries is grown in temperate production environments, but only about 25% of the maize produced in developing countries is grown in temperate environments, largely in China and Argentina (Table 2.2). Of the maize produced in nontemperate production environments, about 53% is grown in lowland tropical environments, 26% in subtropical environments, 11% in tropical mid-altitude environments, and 10% in tropical highland environments (Table 2.2). The marked difference between the industrialized and developing countries' maize growing environments has important implications for the flow of improved technology. Maize germplasm that performs well in temperate regions generally cannot be introduced directly into nontemperate regions without undergoing extensive additional adaptation breeding. Most of the improved open-pollinated varieties (OPVs) and hybrids developed for use in the United States, Western Europe, and northern China (including the vast majority of the commercial hybrids developed by private seed companies) are of little direct use to maize fanners in developing countries. 2

m »-i o vo

ra h : S < :

ËP'S ra S C 'fi

C w

e o

3 T3 O

c 3 O U

M s s i .sS.H J? h 13 > c U 3 Q O M U

•S CS c

oo rJ n t i O oo .h >

i l

3

H

X¡

x¡

f3

¿ ë 2 g s "ra "ra 'rt S- 2 Í3 O O O O U o 'S.'S.'S. £ S^H o o 1o -e e t. >, 3 U h h h m h

U

17

OVERVIEW OF THE WORLD MAIZE ECONOMY

Maize Production

Technologies

Maize production technologies differ between continents, between countries, and even between regions within the same country. One obvious manifestation of this phenomenon is the geographical variability in the use of improved germplasm; the adoption of improved OPVs and hybrids is virtually complete in most industrialized countries, as well as in temperate areas within the developing world, but it is much less extensive in the tropical developing countries (Table 2.3). Differences in production technologies arise from many factors whose relative importance varies from place to place. The most important factors include agroclimatic conditions, the relative availability and cost of key inputs such as land and labor, farmers' technical knowledge and management expertise, government policies that affect the profitability of maize versus alternative crops, the importance of maize in the national economy, the role of maize in local diets, and last but not least, simple historical accident. Industrialized countries. In most industrialized countries in which maize is an important commercial crop, production technologies reflect the highly specialized, input-intensive nature of modern science-based agriculture. Whether maize is grown for use on the farm as an input into the livestock production enterprise or for sale as a cash crop, the scale of

Table 2.3

Adoption of Improved Maize Germplasm (OPVs and Hybrids), Early 1990s 1992 Total Maize Area (million ha)

Latin America 3 Sub-Saharan Africa West Asia and North Africa South, East, and Southeast Asia b Mainly nontemperate developing countries3'15 China Argentina and Chile All developing countries Industrialized countries World

Area Under Improved OPVs

(%)

Area Under Hybrids

Area Under Improved Materials

(%)

(%)

25.1 15.7 2.3 17.6

13 17 6 29

36 20 20 13

49 37 26 42

60.7 21.1 2.4 84.3 37.7 132.6

19 7 8 15 1 10

24 90 85 43 99 63

43 97 93 58 100 73

Sources: FAO Agrostat database, Primary Crops Production. CIMMYT maize releases database. Notes: a. Excluding Argentina (1.7 million ha) and Chile (0.1 million ha), b. Excluding China (22.1 million ha).

18

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

production is generally large: The typical U.S. maize farmer plants 75-100 ha and the typical European farmer about half that much. In industrialized countries most maize is grown under rain-fed conditions (the crop's relatively low value makes irrigation uneconomical) and maize production is generally concentrated in zones of abundant rainfall and fertile soils. Maize in industrialized countries is invariably monocropped, and it is most frequently grown every year on the same land or alternated with soybeans. Land preparation, planting, cultivation, and harvesting operations are completely mechanized, reflecting the generally high cost of farm labor relative to machinery. Virtually 100% of the area planted to maize is under high-yielding single-cross hybrids, and use of chemical fertilizer, herbicides, and pesticides is extensive. Developing countries. Maize production technologies in developing countries are far more variable, reflecting not only the wide range of production environments in which maize is grown but also disparities in farmer knowledge and access to resources (see Dowswell, Paliwal, and Cantrell 1996). Throughout most of Mexico and Central America, as well as in many localized regions within some of the Andean Zone countries of South America, maize is an important food staple that is produced and consumed by a large part of the rural population. Except in a relatively small commercial farming sector, most maize production systems in these countries are characterized by their small scale, complexity, subsistence orientation, and heavy reliance on animal traction and especially human labor. Maize is often grown in association with beans, squash, peppers, cassava, and other food crops destined for home consumption, and many farmers use little or no chemical fertilizer or pesticides. Use of improved varieties is often limited, either because farmers lack access to reliable sources of affordable seed or because they prefer to grow traditional maize varieties developed to meet specific food and feed requirements. Lacking the resources, technical knowledge, or both to acquire and manage modern inputs, many farmers respond to the increased demand for maize by expanding the area planted rather than increasing yields through intensifying their use of purchased inputs. Although maize production has generally kept pace with population growth (with the notable exception of Mexico), it is not clear that current levels of production are sustainable over the long term, since farmers have been forced to shift production into ever more marginal areas, especially hillside areas susceptible to erosion. Elsewhere in Latin America, the picture is different. In southern Brazil, Argentina, and Chile, maize is primarily a cash crop grown by large-scale commercial producers using extensive mechanization and (where profitable) high levels of purchased inputs. Many of the maize production environments found in these countries feature a temperate climate,

OVERVIEW OF THE WORLD MAIZE ECONOMY

19

so growers have been able directly to adopt commercial hybrids and improved management practices imported from North America and Europe. In Africa, maize is a relative newcomer among staple food crops: It arrived in the fifteenth century in the holds of Portuguese ships. With a few notable exceptions (e.g., parts of Ghana, Togo, Benin, Cote d'lvoire, and Nigeria), maize never became a leading crop in the lowland tropical environments of West Africa—in part because many of the varieties introduced directly from North America performed poorly in the hot, humid conditions and proved susceptible to local diseases and in part because alternative starchy foods were readily available. But in eastern and southern Africa, maize became the dominant food crop and the mainstay of rural diets, not only because it was well suited to the region's drier, cooler production environments but also because it was actively promoted by colonial governments. Africa's maize production systems vary widely in scale and productivity. At one end of the spectrum can be found large numbers of peasant farmers who cultivate small plots of maize using family labor or animal traction as the principal source of power. Many of these small-scale farmers grow traditional varieties, applying little or no fertilizer. Yields are generally low, averaging around 2 t/ha or less, and weather-induced crop failures are common. These small-scale producers account for around 70% of all the maize produced in Africa (and probably more than 90% if the Republic of South Africa is excluded). At the other end of the spectrum are a relatively small number of large-scale commercial farmers who grow maize for sale as a food crop or for use on the farm to feed livestock. The large-scale commercial farms, which are concentrated in eastern and especially southern Africa, are often former settler estates established during the colonial period. The production technologies used by Africa's commercial maize farmers are more similar to those found in industrialized countries: Land preparation, planting, and cultivation operations are generally mechanized; use of improved inputs is extensive; and average yields are high (although even on commercial farms yields are characterized by considerable year-to-year variability). Until the middle of the twentieth century, maize production in subSaharan Africa grew mainly through expansion in the area planted, a strategy made feasible by the relative abundance of uncultivated or underutilized land. More recently, the exhaustion of the land frontier has eliminated area expansion as a potential source of production growth, and increased attention has been focused on raising productivity. Unfortunately, the unpredictable climate characteristic of many African maize growing environments has discouraged investment in yield-enhancing technologies; in such areas, farmers who purchase improved seed, fertilizer, and other inputs risk losing their investment in years of inadequate rainfall. Consequently, productivity growth in Africa has lagged compared

20

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

to other regions, with the result that growth in maize production has failed to keep pace with population growth. Asia differs from other regions in the developing world in that most maize in Asia is produced to feed livestock (significant exceptions include parts of India, Nepal, southern China, Indonesia, and the Philippines, where people are the primary consumers of maize). Maize production statistics for Asia are dominated by the statistics for China, whose maize crop is the largest in the developing world and the second largest overall after that of the United States. Because most of China's maize crop is produced in the northern part of the country where a temperate climate prevails, Chinese breeders have been able to make extensive use of improved germplasm imported from the United States and Western Europe. Hybrids developed with the help of imported germplasm have been promoted successfully throughout China's major maize-producing zones, where hybrid adoption levels currently rival those found in the U.S. Corn Belt. With the help of government policies designed to encourage the use of fertilizer and other productivity-enhancing inputs, China's national average maize yields have grown rapidly to their current levels of around 5.0 t/ha, double the average level for all developing countries. Outside China, Asian maize production systems more closely resemble those found in sub-Saharan Africa and Central America: Farms are small, use of improved germplasm and purchased inputs is modest, and yields are generally low. In India, most maize is grown by smallholders to meet household consumption requirements, although commercial production of feed maize has increased in recent years in response to urban consumers' increasing demand for livestock products. In Vietnam, the Philippines, and Thailand, the bulk of the maize crop is now fed to livestock, especially pigs and poultry. Maize Production

Trends

Since the initial diffusion of hybrid technology before World War II, the world maize economy has expanded virtually without interruption, as technological breakthroughs have allowed production growth to keep pace with steadily increasing demand. Although production growth has been strong everywhere, the main sources of productivity growth have differed not only between industrialized and developing countries but also among regions of the developing world. Most maize production takes place in industrialized countries, which currently account for about 60% of total world production (Table 2.4). The gap between industrialized and developing countries is narrowing, however, as production picks up in developing countries (Figure 2.2). Based on current projections of global population and income growth, and in view of the large untapped potential to significantly increase productivity in

OVERVIEW OF THE WORLD MAIZE ECONOMY

Table 2.4

21

World Maize Production Statistics, 1993-1995 Area (million ha)

Developing regions Africa 3 Asia b Latin America 0 Industrialized countries World

26.0 39.4 28.2 41.4 135.1

Yield (t/ha)

Production (million t)

1.56 3.55 2.49 6.51 3.85

40.6 139.9 70.1 269.4 519.9

Source: FAO Agrostat database, Primary Crops Production. Notes: a. Includes all countries. b. Includes all countries except Japan and Israel. c. Includes all countries.

Figure 2.2 World Maize Production, 1961-1995 (million t) 600

500

400

300

200

100

0 1961 19631965196719691971 19731975197719791981 19831985198719891991 19931995 Source: FAO Agrostat database, Primary Crops Production.

many developing countries, it is expected that total production of maize in developing countries will eventually overtake production in industrialized countries. In recent decades, sources of production growth have differed between industrialized and developing countries. In most industrialized countries, where supplies of uncultivated land are effectively exhausted, the area planted to maize has for all intents and purposes ceased expanding; as a result, productivity gains in industrialized countries are achieved almost

22

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

exclusively by raising yields. 3 Yield gains are also an important source of production growth in many developing countries, but in the developing world farmers still achieve production gains by expanding the area planted to maize. Expansion in maize area has been made possible largely by shifting land out of other crops and by introducing maize into nontraditional areas or in nontraditional seasons, such as rabi (winter season) maize in northern India. Even though maize in developing countries is often grown in marginal environments with highly variable climates, aggregate production growth across all developing countries has been more stable than in industrialized countries. The 1980s and 1990s have been a particularly volatile period for maize production in industrialized countries, in part because of several years of unusually poor weather in North America but more importantly because drastic changes in the policy environment and in the price of maize relative to other competing crops have repeatedly altered the economic incentives to grow maize in the United States and Europe.

Overview of World Maize Utilization No other cereal can be used in as many ways as maize. Virtually every part of the maize plant has economic value. The grain can be consumed as human food, fermented to produce a wide range of foods and beverages, fed to livestock, and used as an industrial input in the production of starch, oil, sugar, protein, cellulose, and ethyl alcohol. The leaves, stalks, and tassels can be fed to livestock, either green (in the form of fodder or silage) or dried (in the form of stover). Even the roots can be used for mulching, incorporated into the soil to improve the physical structure, or dried and burned as fuel. Maize Grain Types Numerous cycles of selection by farmers and (more recently) within scientific breeding programs have led to a proliferation of maize grain types, which differ widely in size, shape, color, hardness, smell, taste, ease of processing, storage quality, and other characteristics. In international markets, grain types are commonly characterized in terms of two main traits: color and hardness. All types of maize, regardless of their color, are genetically and biologically similar, but the presence or absence of certain pigments in the cover of the kernel (pericarp) can produce a vivid spectrum of grain colors. Approximately 85% of the global maize crop consists of yellow-grained materials, an additional 10-12% consists of white-grained materials, and the remainder consists of red-, blue-, purple-, or black-grained materials. 4

23

OVERVIEW OF THE WORLD MAIZE ECONOMY

Depending on the composition of the kernel, the hardness of maize grain can vary considerably. In flint maizes, most of the kernel is composed of hard starches, giving the kernel a rounded, slightly shiny surface. 5 In dent maizes, the inside of the kernel is composed of soft starches that contract as they dry, giving the kernel a concave, opaque surface. Approximately 80% of the global maize crop consists of dent or semident materials; an additional 15% consists of flint or semiflint materials. The remaining 5% is composed of specialized grain types, such as the floury maizes of the Andean Zone and the waxy maizes originating from China. Uses of Maize World maize utilization figures appear in Table 2.5. Striking differences are evident between maize utilization patterns in industrialized versus developing countries, as well as across regions of the developing world. Industrialized countries. In most industrialized countries, maize has little significance as human food. Over three-quarters of the maize produced in these countries is fed to animals, principally cattle, hogs, and poultry; the remainder is used mainly as an input into a number of extractive industries. Because of its high starch and low fiber content, maize grain is an extremely concentrated energy source. Readily consumed by animals, it

Table 2.5

World Maize Utilization Statistics, 1992-1994 Utilization (kg/capita/year)

Developing regions Africa a North Africa Eastern and southern Africa West and central Africa Asia'' West Asia East Asia South Asia Southeast Asia Latin America 0 Central America and Mexico South America Industrialized countries World

Food Use

Feed Use

Other Uses

38 64 45 72 64 34 20 31 76 32 29 58 15 5 22

51 23 44 19 13 55 72 60 4 60 57 28 72 76 63

11 13 11 9 23 10 8 10 19 8 12 14 11 19 15

60 62 73 76 43 46 29 84 10 44 149 151 148 274 94

Source: FAO Agrostat database, Food Balance Sheets. Notes: a. Includes all countries. b. Includes all countries except Japan and Israel. c. Includes all countries.

(%)

(%)

(%)

24

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

gives the highest conversion among all major cereals of dry matter to meat, milk, and eggs. In the United States and Europe, approximately half of the maize grain crop is retained by producers for use on their own farms to feed cattle or hogs. The remaining half is sold to the feed industry as an input into maize-based feed rations. Feed mills combine whole or cracked shelled grain with various sources of protein, vitamins, and minerals into specially formulated rations designed to meet the specific nutritional requirements of different animal species at particular stages of their growth cycles. The composition of these feed rations is determined based on least-cost pricing principles, so maize use fluctuates depending on its relative price and alternate sources of starch, protein, and other ingredients. Because maize is relatively inexpensive compared to readily available substitutes, however, demand for maize as a feed ingredient is fairly stable. In addition to the maize that is fed to animals in the form of grain, a significant portion of the crop in industrialized countries is fed to animals as forage. Forage uses of maize include fodder (leaves and stalks, tassels, husks), stover (dried stalks minus the ears), and silage (entire plant chopped and fermented). Between 10% and 30% of the maize planted in the United States and Europe is used as forage, primarily to feed beef and dairy cattle. Developing countries. In developing countries, maize is used primarily as human food and as animal feed. The relative importance of these two uses varies by region and country. In Mexico, Central America, and the Andean Zone of Latin America, most maize is consumed directly by humans, frequently in the form of tortillas (thin round cakes made from ground grain and water, cooked on a heated clay or metal surface). Other common forms of preparation include roasted or boiled green ears, as well as various baked or fried dishes made from ground grain. Countries in which annual per capita maize consumption is particularly high include Mexico (127 kg), Guatemala (103 kg), Honduras (98 kg), El Salvador (93 kg), Venezuela (68 kg), and Nicaragua (56 kg). Not unexpectedly, considering the central role of maize in human diets, Latin American consumers readily distinguish among a wide range of maize varieties on the basis of physical attributes such as appearance, texture, taste, smell, processing quality, cooking quality, and storage quality. Countless cycles of selection by farmers have yielded literally thousands of local varieties distinguished by their unique characteristics. By and large, Latin American maize consumers strongly prefer white-grained materials. Since little difference usually exists in the digestibility or nutritional value of white and yellow maize, this preference is apparently related to the appearance and taste of dishes prepared from white maize.

OVERVIEW OF THE WORLD MAIZE ECONOMY

25

The preference for white maize may also stem from historical experience: When imports have been necessary to meet local production shortfalls, many countries have imported yellow feed grain whose consumption qualities have frequently been unsatisfactory. Throughout large areas of sub-Saharan Africa, particularly eastern and southern Africa, maize is consumed directly as human food—most frequently in the form of porridge, soup, or gruel prepared from ground grain. Roasted or boiled green ears may also be consumed. In some parts of western and central Africa, grain-and-water mixtures are fermented prior to cooking. People directly consume about 85% of the maize produced in sub-Saharan Africa; only a small proportion of the crop is fed to animals, generally poultry. As in Latin America, throughout most of subSaharan Africa consumers strongly prefer white-grained materials. Once again, the source of this preference seems to be historical and cultural; it does not appear to come from any association between grain color and digestibility or nutritional quality. Annual per capita maize consumption levels can be extremely high, especially in Malawi (137 kg), Zambia (113 kg), South Africa (94 kg), Kenya (93 kg), Zimbabwe (89 kg), and Lesotho (87 kg). In Asia, where the principal food staple is often rice, maize rarely forms the mainstay of the human diet and is generally used to feed livestock. Depending on the type of livestock, as well as the availability of alternative feed sources, maize may be fed to animals as either forage or grain. Throughout most of East and Southeast Asia, where meat consumption is relatively high, producers retain much of the maize crop to use as an input into backyard hog and poultry operations. In South Asia, where meat consumption is much lower, maize is most commonly used as a source of fodder for dairy cattle. Maize Utilization Trends Maize resembles most other major cereals in that once consumer incomes rise above a certain threshold level, direct human consumption of maize declines with additional growth in income. Above the threshold income level, consumers respond to additional increases in purchasing power by consuming less maize and greater quantities of higher-priced luxury foods, especially meat and dairy products, fruits, and vegetables. But unlike most other cereals, at higher income levels declining direct demand for food maize is offset by rising indirect demand for feed maize, because wealthier consumers use greater quantities of livestock products whose production requires significant amounts of grain. This characteristic of maize distinguishes it from other cereals, because it means the demand for maize holds up across a wide range of income levels.

26

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

Industrialized countries. Since the early 1970s, in industrialized countries the main source of growth in demand for maize has shifted. A slower rate of population growth has been offset by income-driven changes in consumption patterns, which have heightened the demand for livestock products. Countries with excess maize production capacity, notably the United States and Western European nations, have intensified their efforts to increase demand for maize by promoting industrial uses such as starch extraction and the production of ethyl alcohol. In general, these efforts have failed to influence utilization in any significant way, and most of the crop in industrialized countries continues to end up in livestock feed. Future demand for maize is difficult to predict. Population growth in many industrialized countries has begun to slow, suggesting that population increases will dwindle in importance during the first half of the twenty-first century. Although income levels are still rising in many of these countries, consumption of meat and livestock products may have started to level off in some countries, perhaps because of health-related concerns surrounding excessive meat consumption. Direct food demand and even indirect feed demand for maize are likely to grow more slowly in the future. On the other hand, industrial demand for maize seems likely to increase as innovative uses are found for the crop. The International Food Policy Research Institute (IFPRI) projects that demand for maize in industrialized countries will grow at an average annual rate of 0.80% during the period 1990-2020, significantly faster than the projected demand for wheat and rice in these countries (Rosegrant, Agcaoli-Sombilla, and Perez 1995). Developing countries. Utilization of maize in developing countries is increasing steadily, although the sources of demand growth vary by region. In Africa, where per capita incomes are rising very slowly if at all, the main source of demand growth continues to be population increases. The situation is quite different in East and Southeast Asia, as well as in the Southern Cone of South America, where changes in consumption patterns attributable to income growth have caused a sharp surge in demand for feed maize. Mexico and Central America, as well as most of South Asia, present an intermediate scenario in which demand growth originates from both food and feed sources. Looking to the future, demand for maize in the developing world will continue to grow strongly, fueled by population growth (e.g., in subSaharan Africa, Mexico, and Central America), income growth (e.g., in East and Southeast Asia), or both population and income growth (e.g., South Asia). Currently, IFPRI projects that demand for maize in developing countries will grow at an average annual rate of 2.2% during the period 1990-2020, about as fast as projected growth in demand for wheat and significantly faster than projected growth in demand for rice (Rosegrant,

OVERVIEW OF THE WORLD MAIZE ECONOMY

27

Agcaoli-Sombilla, and Perez 1995). Since IFPRI's baseline projections do not account for changes in the income elasticity of demand for different cereals, they undoubtedly understate future growth in demand for maize relative to wheat and rice.

International Trade in Maize After wheat, maize is the world's most extensively traded cereal. Global trade in maize averages about 60-70 million tons per year, compared to about 110-120 million tons per year for wheat and wheat flour. 6 Although virtually all of the wheat traded internationally is destined for human consumption, most of the maize is used to feed livestock, primarily cattle and chickens. Global trade in maize increased significantly between 1950 and 1980, rising from just over 15 million tons to about 80 million tons per year. Most of this growth resulted from rapidly expanding production in temperate environments, particularly in the United States, which created large surpluses that were disposed of in international markets with the help of aggressive export promotion policies. After 1980, the volumes of maize moving through international markets declined, reflecting not only decreased demand from traditional importers (many of which had succeeded in increasing domestic production) but also decreased supply from traditional exporters (many of which had introduced policy reforms to curtail chronic overproduction). Since the early 1980s, global trade in maize has fluctuated at between 60 and 75 million tons per year. Underlying these relatively stable numbers, however, are significant changes in global trading patterns. The United States has maintained its position as the world's dominant exporter; U.S. exports have consistently made up about 60-65% of the total volume of international trade. The European Community (EC) has also been a consistent net exporter, although the true volume of the EC's trading activity is often overlooked because trade among EC members is not reflected in global trade statistics. Meanwhile, several developing countries have risen from relative obscurity to become players in the global maize trade, most notably Argentina and China, which have taken advantage of their positions as low-cost producers to capture a share of the global export market (in some cases only temporarily, however). Compared to the changes that have occurred on the demand side of the global maize market, the pattern of imports has been relatively stable. Asia remains the world's largest importing region, as strong population and income growth in countries such as Japan, South Korea, and Taiwan has spurred demand for feed grain imports. The European countries of the former USSR constitute another strong source of demand; maize imports into

28

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

those countries rose appreciably following the political upheavals of the late 1980s and early 1990s, which disrupted agricultural production and distribution systems. Import demand from these countries can be expected to diminish if and when political stability returns and effective agricultural support policies are re-established. Traditionally, industrialized countries and wealthier developing countries have dominated the world trade in maize; relatively few low-income developing countries have been consistent exporters or importers. The most significant exceptions are Mexico, which since the mid-1970s has relied increasingly on maize imports from the United States to make up chronic domestic production shortfalls, and the Republic of South Africa, which exports maize in years of average or above-average weather. Countries in Central America, as well as sub-Saharan Africa, have attempted to remain self-sufficient in maize, turning to international markets only as a last resort to make up periodic production shortfalls or to dispose of occasional surpluses. Low-income developing countries in which maize is an important food staple have sought to isolate themselves from international markets for two main reasons. First, reliance on international markets is often seen as undesirable for reasons of national food security. Second, in most countries where maize is a food staple, consumers prefer whitegrained maize, whereas most of the maize available in the international market is yellow-grained. 7 Countries that import maize often must accept yellow maize, although consumers may be reluctant to use it. Complicating any projections about the future patterns of global trade in maize are the uncertain prospects for China. Since the early 1980s, China has shown an increasing willingness to turn to international grain markets to make up differences between domestic supply and demand of cereals. By virtue of its enormous size, China can make a large impact whenever it enters international grain markets—even if the transaction involves a small proportion of China's internal market for grain. During the early 1990s, following economic reforms that encouraged farmers to produce grain for export, China suddenly became a major player in Asian maize markets, but Chinese exports proved short-lived. Whether China will return to international maize markets, and in what capacity, remains an open question. Some analysts predict that China cannot maintain current levels of self-sufficiency in food production and that the government will be forced to rely increasingly on grain imports to meet chronic domestic production shortfalls, resulting in a sharp drawdown in global grain stocks and strong upward pressure on international prices (Brown 1995). Others are more sanguine about China's ability to feed itself over the longer term and predict that traditional exporters such as the United States will be able to step up production to keep pace with slow growth in China's demand for imports (Rozelle, Huang, and Rosegrant 1996).

OVERVIEW OF THE WORLD MAIZE ECONOMY

29

International Maize Prices Adjusted for inflation, the main international reference price of maize has been declining steadily for decades (Figure 2.3). Over the long term, large fluctuations in production levels and significant growth in trade volumes scarcely seem to have affected the long-term trend. International prices have occasionally spiked upward for brief periods following unusual weather- or policy-related shocks (most notably in the early 1970s during the world oil crisis and more recently in 1996 following successive years of exceptionally poor weather), but periods of prolonged price increases have been rare. The historical pattern of international maize prices reveals several things about the global maize economy. First, the relative stability of maize prices compared to prices of many other widely traded commodities indicates that the global market is competitive and can effectively equilibrate supply and demand. Second, the relative absence of prolonged price spikes suggests that industrialized countries have the excess production capacity to respond quickly to the price rises that signal global shortages. Third, the long-term downward trend in international prices indicates that the benefits of technological change have largely been passed along to consumers in the form of lower prices.

Figure 2.3 International Reference Price of Maize, 1961-1995 (U.S.$/t)

30

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

Since many countries deliberately seek to isolate themselves from international markets, domestic maize prices often diverge significantly from international prices, especially in the developing world. Despite attempts to control domestic prices through policy interventions, effective stabilization is not always possible, especially in countries where the bulk of production comes from rain-fed areas subject to highly variable precipitation patterns. The relative stability of international prices thus contrasts sharply with domestic prices observed within individual countries.

The World Maize Seed Industry The growing size and increasing commercialization of the global maize economy have been accompanied by an expansion in the industries that provide inputs used in maize production—especially improved seed, chemical fertilizer and pesticides, and machinery. The steady intensification of maize production systems, initially in the industrialized countries of North America and Western Europe and later in selected developing countries, has created increased opportunities for input supply companies, some of which now rank among the largest in the world. Over time, as the global maize seed industry has matured, it has undergone a series of organizational changes. The nature and pace of these changes have varied among countries, reflecting differences in the structure of production from one country to the next, as well as differences in the economic, political, and institutional climates. The net result has been a global maize seed industry comprising many different types of national seed industries that vary widely in their organization and performance ( C I M M Y T 1992, 1994).

Industrialized

Countries

In most industrialized countries, the maize seed industry is now largely in the hands of the private sector. The roles of public universities, research institutes, and extension organizations, which once dominated maize research and technology transfer activities, have diminished as private companies have steadily expanded their sphere of influence to take advantage of profit opportunities offered by an increasingly commercialized and input-dependent maize economy. Public organizations continue to play an important role in the technology development and transfer process, but they do so within an increasingly narrow and specialized realm. For example, the focus of publicly funded research has shifted "upstream" toward more basic research. Today, publicly funded maize researchers engage mainly in strategic research aimed at developing more efficient breeding methods or generating basic descriptive information about the

OVERVIEW OF THE WORLD MAIZE ECONOMY

31

characteristics of different breeding materials; the relatively small number still involved in breeding tend to concentrate on prebreeding (aimed at separating desirable from undesirable traits in base collections of breeding materials) or on inbreeding (aimed at developing elite inbred lines that can be used as parents in agroclimatically targeted hybrid development work). Very few publicly funded maize researchers now operate toward the applied end of the research spectrum, for example, developing and testing finished hybrids; private companies have assumed these functions. Just as most applied research activities have been assumed by private seed companies, so have many technology transfer activities. Because farmers now understand that modern hybrids fully express their genetic potential only in the presence of proper management, seed companies routinely provide technical assistance along with their germplasm products. This assistance has greatly reduced the need for publicly funded extension services, which have dwindled in size and scope. Today, maize farmers in the United States and Europe are likely to look first to their inputs dealer rather than the local government extension agent for technical advice on how to manage their crop. This shift has been reflected in a change in private-sector hiring patterns, and it is not unusual to find seed companies hiring more agronomists than breeders in any given year. If the steadily growing world maize economy has provided the impetus for the private seed industry to expand, economies of scale in research and seed production have contributed to its increasing concentration. The 1980s and 1990s have witnessed an unprecedented wave of mergers and consolidations during which the vast majority of small independent seed companies have been bought out by or merged with larger competitors. The result is a highly concentrated global seed industry. Although reliable data are difficult to come by, at latest count less than a dozen companies appeared to account for over three-quarters of the global seed trade. Some of these companies specialize in seed, but the majority are diversified conglomerates that also deal in other agricultural inputs, especially pesticides and fertilizer. A few also engage in large-scale grain trading and livestock production. Although it is true that a handful of transnationals currently dominate the global seed industry, often overlooked is the fact that smaller companies have not been entirely displaced. In the United States, for example, although the 7 largest companies currently control about 70% of the market for maize seed, nearly 300 other companies also produce and sell hybrid maize seed (Norskog 1995). Many of these companies are family-owned firms that sell maize seed along with seed of other important field crops. This small-scale segment of the industry is remarkably persistent if superficially unstable in that high turnover occurs among individual companies (i.e., new companies are constantly being formed, and old companies are constantly going out of business). The key to the success of these smaller

32

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

companies apparently lies in their ability to offer certain segments of the market products and services the larger companies are unwilling or unable to provide (see Chapter 11). Developing

Countries

In developing countries, the maize seed industry is more variable in organization and performance. Generally speaking, the balance between public- and private-sector involvement varies with the size and degree of commercialization of the maize economy. In countries (or in regions within countries) where maize is still produced mainly by small-scale, subsistence-oriented farmers using low levels of purchased inputs, private firms have demonstrated an understandable reluctance to enter the market. In these countries, maize research, seed production, and seed distribution are generally carried out by public organizations. But in countries (or in regions within countries) where farmers have begun to adopt improved production technologies, commercial opportunities have begun to appear, and private firms are becoming increasingly active. In some cases, private-sector involvement is still limited; believing it is unprofitable to invest in research, private firms restrict themselves to multiplying and distributing OPVs and hybrids produced by public breeding programs. In other cases, private-sector involvement is much more advanced, with private firms engaging in a wide range of research, seed production, and seed distribution activities. Just as the degree of private-sector participation varies among developing countries, the identities of the leading actors also vary. In some cases, the national maize seed industry is composed of local companies, with little or no participation by transnationals. In other cases, transnational have established a presence, sometimes to the extent of dominating the local market. The factors that explain these differences—and the implications for the developing world's maize farmers—are examined in greater detail later in this book.

Future Prospects for the World Maize Economy At the global level, long-term growth prospects for maize appear stronger than those for other major cereals. As noted previously, because demand for feed maize will continue to increase with rising incomes even at relatively high per capita income levels, the demand for maize is expected to grow strongly well into the twenty-first century and beyond. What will be the source of the production increases that will be needed to meet the future demand for maize? Steady future production growth can be expected in the industrialized countries of North America

OVERVIEW OF THE WORLD MAIZE ECONOMY

33

and Western Europe. Not only are these traditional maize exporters inherently well suited to maize production by virtue of their temperate climates, abundant rainfall, and fertile soils, but they are the beneficiaries of a long history of sustained investment in maize research that has left them with a wealth of improved production technologies. In addition, the ability of many commercial farmers in North America and Europe to switch between crops in response to price signals would allow them to respond to increased demand by expanding the area planted to maize. Although increasing concern for the environmental impacts of indiscriminate use of inorganic fertilizer and chemical pesticides will no doubt lead to changes in crop management practices, the traditional maize exporting countries should retain the capacity to increase maize production substantially. Additional growth in maize production will come from the developing countries of Latin America, Africa, and Asia. In the short to medium term, maize production increases in these countries will be made possible by greater dissemination of productivity-enhancing production technologies, spearheaded by improved germplasm, which will serve as the catalyst for the adoption of improved crop management practices. In the long term, maize production will have to become more economically rational—meaning productivity will have to rise far enough to provide competitive returns to land, labor, and capital. Recognizing the need to rationalize their agricultural economies, governments in many developing countries have already taken steps to expose maize producers to the winds of global competition. In many cases, this has meant reducing support to inefficient smallholder production systems and allowing a greater share of domestic consumption requirements to be met through imports. Even if the general outlook for the world maize economy appears promising, it is important to recognize that the benefits of future technical change will be distributed unevenly. In many countries in Latin America and sub-Saharan Africa in which maize is a primary food staple and the backbone of the rural economy, formidable barriers continue to inhibit the spread of improved production technologies. Because these barriers are unlikely to be overcome in the foreseeable future, certain groups of smallscale, subsistence-oriented producers will be unable to respond effectively to attempts to rationalize the agricultural sector. In this context, political necessity—not to mention common human decency—demands that continuing efforts be made to address the special needs of these most disadvantaged groups.

Notes 1. Technically speaking, yield per unit land area is a partial productivity measure, so productivity comparisons based on yields are valid only when all inputs

34

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

are held constant (which is generally not the case). Nonetheless, because yield per unit land area is easily calculated, it is often used as a measure of (total) productivity. 2. In keeping with its mandate to serve the poor in developing countries, CIMMYT concentrates its maize breeding efforts on nontemperate environments. Work on temperate materials is left to the private sector and to the national programs of the industrialized countries. 3. This is something of an oversimplification. In some industrialized countries, particularly the United States, the area planted to maize fluctuates from year to year as land is shifted among crops in response to changing policy incentives. 4. Maize production and utilization statistics are not normally disaggregated by grain color, so these figures should be regarded as indicative. 5. Popcorn, which is consumed as a snack food in many parts of the world, is an extremely flinty maize with unusually small kernels. 6. The figures for wheat include food aid, which accounts for nearly 25% of the total. 7. World trade in white maize ranges between 1.5 and 2.0 million tons in normal years and is highly regionalized, making reliance on international markets very risky. For a description of the world white maize economy, see CIMMYT/FAO (1997).

Maize Seed Industries: A Conceptual Framework MICHAEL L. MORRIS, JOSEPH RUSIKE, AND MELINDA SMALE

What is the best way to assign responsibility for plant breeding research, seed production, and seed distribution activities among different types of organizations? What factors determine the optimal form of seed industry organization? Does the optimal form of industry organization change through time? Is the path of change deterministic and hence predictable? To what extent can the path of change be influenced by policy measures? This chapter sets out a conceptual framework to help structure discussion of these and other basic questions about the maize seed industry. The conceptual framework is needed to introduce a common vocabulary and to establish a set of standardized concepts and analytical tools that can be used in examining seed industry organization and performance. Initially, no attempt will be made to formulate answers to the key questions. As will become clear, many of the questions are at least partly empirical, and meaningful answers can be found only by looking to the real world. For now, the goal is simply to determine the extent to which theory and logic can help to narrow the focus and sharpen the terms of discussion.

Important Attributes of Maize Seed It is useful to begin by considering the industry's principal product. Maize seed, like all seed, is really two things in one. First, it is a consumable input—something that farmers can combine with other inputs (e.g., land, labor, water, fertilizer, pesticide) to produce a crop. At the same time, maize seed is also a source of germplasm—a store of genetically encoded information that determines how other inputs combine and become transformed into useful products. Practically speaking, these two things are difficult to separate, in the sense that a farmer who purchases a bag of seed in order to acquire the consumable input needed for planting cannot 35

36

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

avoid purchasing the genetically encoded information contained in the seed. 1 The fact that maize seed represents two fundamentally different things is significant. Although maize seed considered as a consumable input and maize seed considered as a source of germplasm have many attributes in common, they differ with respect to two attributes that turn out to be important in shaping the incentives for seed organizations to produce seed and for farmers to consume seed: subtractability and excludability. Subtractability refers to the degree to which use of a good or service by one person precludes use by another person. Maize seed considered as a consumable input has high subtractability in the sense that two farmers cannot use the same bag of seed. If one farmer plants a bag of seed, that same bag of seed can no longer be planted by another farmer. But maize seed considered as a source of germplasm has low subtractability. If one farmer purchases a bag of seed in order to make use of a particular openpollinated variety (OPV) or hybrid, this does not prevent other farmers from using the same OPV or hybrid. Assuming the other farmers are able to purchase different bags of seed, their use of the same germplasm neither directly affects nor is directly affected by the first farmer's use. 2 - 3 Excludability refers to the ease with which the seller of a good or service can deny access to nonauthorized users. Maize seed considered as a consumable input has high excludability, meaning it is cheap to exclude unauthorized users. Sellers of maize seed can deny access to nonauthorized users simply by refusing to provide them with seed if the nonauthorized users do not pay the asking price. But the degree of excludability is quite different for maize seed considered as a source of germplasm. In principle, the germplasm contained in maize seed should have low excludability (meaning it is expensive to exclude nonauthorized users), since nonauthorized users can gain access to germplasm by obtaining a few kernels of seed. In practice, however, because of differences in the way OPVs and hybrids reproduce, the degree of excludability varies. The germplasm in OPV seed has low excludability, because OPV seed is easily reproducible; if a nonauthorized user manages to obtain a few seeds of an OPV, more seed can easily be produced. But the germplasm in hybrid seed has high excludability, because hybrid seed can be reproduced only by those who know the pedigree or have access to the parent lines. Although differing in terms of subtractability and excludability, maize seed considered as a consumable input and maize seed considered as a source of germplasm share a third attribute that affects production and consumption incentives. Transparency refers to the degree to which the characteristics of a product are clearly evident. Like all seed, maize seed is nontransparent in the sense that potential users cannot easily make quality determinations. It is very difficult to judge the quality of maize seed simply by looking, because seed has a number of characteristics that are difficult to recognize and measure. The nontransparency attribute pertains

A CONCEPTUAL FRAMEWORK

37

both to maize seed considered as a consumable input (in which case the difficulty lies in determining varietal purity, viability, health, and similar factors) and to maize seed considered as germplasm (in which case the difficulty lies in determining the genetic composition). The nontransparency characteristic of maize seed (which is unchanging) and the subtractability and excludability characteristics (which differ depending on whether seed is being considered as a consumable input or as a source of germplasm) have important implications for the optimal organization of national maize seed industries, because they affect the economic incentives facing seed producers and consumers. The nontransparency of maize seed creates an information asymmetry between sellers and buyers. In most negotiated market transactions involving seed, the seller knows the quality of the seed but the buyer does not. 4 Thus, transactions involving maize seed always involve some degree of uncertainty for the buyer. Under these circumstances, buyers of maize seed must either invest time and effort in determining the quality of the seed (which is costly), or they must rely on some other means of reassuring themselves that the seed is likely to be of good quality. One way to do this is through regulation, as when a government agency is empowered to enact and enforce seed quality standards. Farmers' associations can also serve to conduct independent quality checks on seed. Another way that buyers reassure themselves of seed quality is to purchase seed only from companies whose seed is known to have been good in the past (brand loyalty), on the assumption that these companies will continue to produce good seed in the future. Unlike the problems caused by lack of transparency, the problems caused by the subtractability and excludability characteristics of maize seed depend on the type of seed involved and the use to which it is put. Considered as a consumable input, maize seed has high subtractability and high excludability. But considered as a source of germplasm, maize seed has variable subtractability and excludability characteristics. While the degree of subtractability is always low, the degree of excludability depends on whether the seed is OPV seed or hybrid seed. The low subtractability of maize seed considered as a source of germplasm, coupled with the variable excludability (low for OPVs, high for hybrids), suggest that the institutional arrangements needed to effectively develop, produce, and distribute improved maize seed will tend to differ. The combination of high subtractability and high excludability is associated with products which are effectively handled through private market arrangements or exchanges. Sellers have incentives to provide the product, buyers have incentives to pay for the product, and market failures are unlikely. Because of the high subtractability, nonauthorized users have little opportunity to act as free riders; for example, a farmer cannot plant a bag of maize seed once it has been planted by someone else. 5 Because of the high excludability, sellers are able to realize a return on their investment,

38

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

since they can easily deny access to nonauthorized users who refuse to pay the asking price. Thus, buyers have an incentive to demand seed through the market, and sellers have an incentive to supply seed through the market, and in the absence of distortions the market will clear at prices and quantities which maximize social welfare. The combination of low subtractability and low excludability is associated with products which cannot always be handled effectively through private market arrangements. Sellers are likely to have difficulty realizing a return on their investment because they cannot easily exclude nonauthorized users. Under these circumstances, private companies are likely to underinvest in OPV research, since they cannot expect to capture the full benefits through seed sales.6 Therefore, reliance on a market solution is likely to result in production and consumption of OPV germplasm at socially suboptimal levels. This situation suggests that some form of government or collective action will probably be needed to ensure that adequate OPV research takes place. Numerous options exist for ensuring investment in OPV research; one strategy is to combine public research on OPVs with public, private, or collective seed production, all subject to regulation and monitoring. No strategy will be perfect, however, because the combination of low subtractability and low excludability is fraught with potential institutional hazards. In addition to the problem of free-riding, difficulties exist with measuring the demand for and benefits of OPV germplasm. And whenever there is public funding of provision, opportunities arise for rentseeking behavior, since government decisions regarding OPV research confer unearned advantages to particular individuals or groups.7 The combination of low subtractability and high excludability is associated with products which can potentially be handled effectively through private market arrangements. Sellers will be assured a return on their investment, however, only if the high excludability can be maintained. Traditionally, the high excludability of hybrid maize seed has been maintained through the "trade secrets" method, under which seed companies decline to release information about pedigrees, refuse to provide access to parent inbred lines, or both. More recently, as advances in technology have made it easier (and cheaper) to defeat the trade secrets strategy by identifying and replicating successful commercial hybrids, seed organizations have turned increasingly to the courts in an effort to maintain the high excludability of their proprietary hybrids (i.e., through intellectual property rights laws). In addition, they use advertising to convince buyers that their brand is superior to other brands.

Organizational and Institutional Options The transparency, subtractability, and excludability characteristics of maize seed influence production and consumption incentives. The lack of

A CONCEPTUAL FRAMEWORK

39

transparency restricts the ability of potential buyers of maize seed to make quality determinations, which exposes them to the risk of unexpected losses and decreases their willingness to buy (in the absence of trust in the seller). The subtractability and excludability characteristics of maize seed affect the ability of potential providers to realize benefits through sales, which, depending on the circumstances, either increases or decreases their willingness to sell. 8 Given these links between product attributes and economic incentives, which combination of organizations and institutions is most likely to lead to the production and consumption of socially optimal quantities of maize seed? Social science can help here. Welfare economics, institutional economics, and the theory of public choice provide some useful tools for understanding the forces that are likely to condition the behavior of seed industry participants in responding to economic and other incentives. Knowledge of these forces can help us to identify which forms of industrial organization and which supporting institutional arrangements are likely to bring about good performance in the seed industry. The Neoclassical

Paradigm

What can neoclassical economics tell us about the optimal form of seed industry organization? According to neoclassical theory, the most efficient form of industrial organization is the competitive market, made up of large numbers of economic agents (individuals or firms). These economic agents seek to maximize their welfare by engaging in production, consumption, and exchange activities. Voluntary transactions serve to equilibrate supply and demand for goods and services at market-clearing prices that lie close to the low points on the industry's long-term average cost curves. Market participants have full and free access to information and are able to adjust their bid and offer prices instantly and at no cost. Guided by an invisible hand, the market always clears, the most efficient technology always prevails, and excess profits are always eliminated. Any divergence from this idealized scenario is explained in terms of market failure. Whenever real-world performance fails to conform to the predictions of the theory, neoclassical economists assert that the efficient functioning of the market economy is being impeded by exceptional circumstances (e.g., externalities, natural or contrived monopolies, wage and price stickiness, free riding). The fact that these exceptional circumstances often arise does not so much negate the validity of the theory as inhibit its normal operation. Institutionalist

Modifications

Despite its obvious usefulness for analyzing many economic phenomena, the neoclassical paradigm suffers from a major problem: The idealized

40

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

market economy depicted in textbooks often fails to correspond to the messy reality of everyday life. After it became clear that the discrepancy between neoclassical theory and reality is too large to ignore, attempts were made to modify the theory to increase its relevance for applied analysis. One particularly influential source of modifications has been a group of economists who have focused on the role played by institutions in shaping economic activity. Previously, institutional matters had largely been relegated to a "black box"—even when they had been acknowledged, institutions had rarely been integrated into analytical models (Bardhan 1989). Institutional economists recognize that the real world is not inhabited by rational, all-knowing, self-interested decisionmakers and that people do not engage in instantaneous, costless transactions that lead to efficient, utility-maximizing market outcomes. Instead of resembling the predictable, stylized actions depicted by neoclassical theory, people's actions are shaped by a wide range of factors that influence the way they absorb and evaluate information, take decisions, and act upon those decisions. In the institutionalist universe, economic agents still seek to maximize utility; however, the ability (or in some cases the desire) to pursue this goal is now compromised. As a result, economic systems often generate outcomes different from those predicted by neoclassical theory. To appreciate the importance of the factors that affect human behavior and to understand their relevance for the study of maize seed industries, it is useful to review briefly the work of several seminal thinkers in the field of institutional economics. Ronald Coase (1960) developed a framework for analyzing the organization and performance of industries that revolved around the concept of transactions costs. Coase argued that firms and markets represent alternative ways of organizing production and that firms replace markets when the cost of conducting transactions through the firm is lower than the cost of conducting them through the market. According to Coase, what firms decide to buy, produce, and sell is determined by the relative size of the cost of administering transactions internally versus the cost of conducting them through the market. If the cost of administering transactions internally is lower than the cost of conducting them through the market, the firm will internalize the transactions (this process is known as vertical integration). But if the cost of conducting transactions through the market is lower than the cost of administering them internally, the firm will engage in buying-and-selling activities. Coase's pioneering work on transactions costs was modified and extended by Oliver Williamson (1975, 1985). Williamson added a behavioralist dimension designed to take into account the influential role individuals play in determining the performance of economic systems. According to Williamson, to understand economic systems it is necessary to recognize two distinguishing characteristics of the people who populate

A CONCEPTUAL FRAMEWORK

41

them. First, people are boundedly rational, meaning there are limits to the amount of information they can assimilate and process. This implies that they cannot know everything about a transaction. Second, people are opportunistic, meaning they will act in ways that enhance their self-interest (e.g., they will misrepresent the quality of their products and will renege on contracts when they believe it is in their interest to do so). These two characteristics of human behavior have important implications for the performance of economic systems. Because of bounded rationality, all contracts are necessarily incomplete. Because of opportunism, all contracts are subject to rupture. Taken together, this means that when individuals acquire new information they will tend to reassess their contracts and possibly break them, resulting in increased transactions costs as both parties continually readjust to the succession of broken contracts. Williamson argues that to lower transactions costs over the long term, potential parties to transactions develop institutional supports (rituals, customs, or standard operating procedures) that enable them to economize on bounded rationality while securing protection against the hazards of opportunism. Whereas Coase concentrated on two alternative ways of organizing economic production (markets and firms), Williamson extended the typology, positing that four generic forms exist for organizing economic activity: (1) markets, (2) hierarchies, (3) government bureaus, and (4) "hybrids" (e.g., strategic alliances, joint ventures, licensing agreements). These four forms differ in their incentive structures, administrative costs, ability to adapt to disturbances, and legal requirements, making them differentially suited for different types of transactions. According to Williamson, institutional arrangements evolve that minimize transformation and transactions costs. This evolution can occur in several ways. Most often, inefficient forms of organization are eliminated through market competition. Less commonly, inefficient forms of organization are deliberately replaced in the wake of careful analysis and planning. Although it has been praised by many social theorists for offering useful analytical tools for examining ways of organizing economic activity, Williamson's approach has been criticized by some for inadequately explaining how forms of economic organization evolve over time (for a summary of this literature, see Rusike 1995). These critics suggest that institutions undergo a process of evolutionary change that cannot be attributed solely to the drive to minimize transactions costs but rather appears to be driven by interactions among a range of other factors, including technology. Douglass North (1990) offered a slightly different interpretation of the process driving institutional change. North started with three basic assumptions that are fairly similar to Williamson's ideas about bounded rationality and opportunism. First, individuals are concerned with maximizing

42

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

their own welfare; when information costs are high, individuals stand to gain by disobeying rules others are expected to obey. Second, it is costly for individuals to specify and enforce rules when information is imperfect. Third, socially conditioned belief systems enable individuals to economize on information costs while forming judgments about the fairness and legitimacy of alternative institutional arrangements (meaning their actions are often guided by ideology). Starting with these three basic assumptions, North explored the process of institutional change by focusing his attention on economic systems that were experiencing periods of disequilibrium, including systems undergoing development. North observed that transactions costs appear to increase during periods of economic development, because possibilities for opportunistic behavior increase. Against this background, markets and institutions change over time, because people invest in knowledge and skills that lead to revised evaluations of their opportunities. This induces them to alter the formal rules and informal constraints governing behavior, so that new ways of organizing economic activity eventually emerge. The process is not deterministic, however. If the overall institutional framework provides incentives for people to engage in productive activities, the economy will be driven along a path that leads to economic progress. But if the institutional framework provides incentives for people to engage in unproductive activities, the economy will stagnate. As economic development proceeds, information costs proliferate, and increasingly complex institutions are needed to hold down the costs of enforcing contracts. The Marxian

Perspective

In contrast to the institutional economists, who describe industrial change as driven by the desire to reduce costs, Marxian economists offer a somewhat different explanation. The Marxian view is expounded by Jack Kloppenburg (1988), who has interpreted industrial evolution in terms of fundamental historical processes associated with the political economy of capitalism. According to Kloppenburg, the two key processes driving organizational change in the seed industry are primitive accumulation (the separation of the worker from the means of production) and commodification (the extension of the commodity form to seed). Kloppenburg focuses on the evolution of the U.S. maize seed industry, arguing that as long as OPVs remained the dominant technology, farmers retained the ability to save seed from their own harvest to replant the following season. Effective control over the means of production was left in the hands of direct producers, and rent-seeking firms had no incentive to enter the seed industry. To create economic incentives (which are needed to pave the way for the expansion of capitalism and the advance of seed technology), farming had to be converted from a largely self-sufficient

A CONCEPTUAL FRAMEWORK

43

production process into one in which purchased inputs accounted for the bulk of the resources employed. In the Marxian view, this transformation can be seen as a process of primitive accumulation in which the farmer is progressively separated from the means of agricultural production, including seed. According to Kloppenburg, in the U.S. maize seed industry the key development in this process was the commodification of seed, which was made possible by the introduction of hybrid technology and the extension of property rights to germplasm. These developments dispossessed farmers from control over the means of production by preventing them from autonomously reproducing seed for their own use. In Kloppenburg's view, the U.S. maize seed industry has evolved to support a capitalist mode of production (agribusiness) characterized by the existence of a class of direct producers who have been dispossessed of the means of production. This process has affected not only the seed industry but agricultural science itself. Reviewing the history of maize research in the United States, Kloppenburg argues that over time a social division of labor occurred that progressively subordinated public research to serve the needs of private industry. To see this, we need only to look at maize breeding research: In stark contrast to the early twentieth century, when virtually all commercial OPVs and hybrids were developed and released out of the public sector, public research programs no longer release commercial materials. Kloppenburg is by no means neutral about the process he describes. He is alarmed by the penetration of capital into the seed industry, because what is most profitable for seed companies is not always what Kloppenburg regards as socially optimal. One example is the tendency of seed companies to achieve market domination by promoting large numbers of similar products (e.g., hybrids derived from multiple lines derived from a single parent), even though this practice may narrow the genetic diversity of the materials planted by farmers. 9 Further Institutionalist

Modifications

The groundbreaking work of Coase, Williamson, and North has been complemented in recent years by ideas associated with a new school of thought generally referred to as the "new institutional economics" (NIE). The somewhat disparate literature of the NIE has concentrated particularly on issues related to interest group formation, collective action, organization theory, bounded rationality, transactions costs, and the relationship between technological change and institutional change. Mustapha Nabli and Jeffrey Nugent (1989a, 1989b) aptly describe the subject matter of NIE— institutions and their determinants—as the "very battleground" of the neoclassical and Marxist paradigms. Many NIE concepts are derived from ideas advanced earlier by mainstream institutionalists, such as the idea that the main force driving

44

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

organizational change is the push to save on transactions costs. Allan Schmid (1987) and others take this idea and build on it, arguing that the size and distribution of transactions costs are influenced not only by market structural conditions, availability of information, and technology, but also to a considerable extent by the characteristics of products themselves (see also Picciotto 1995; Ostrom, Schroeder, and Wynne 1993; Ostrom, Gardner, and Walker 1994). The proponents of NIE point out correctly that traditional approaches to policy analysis are frequently unhelpful, because they ignore important differences between types of goods. A principal argument made by these proponents is that when organizational options are described using a simple public versus private dichotomy, policy analysis is severely constrained, and the range of diagnostic and prescriptive options is reduced to just two: "market failure/government must intervene" and "state failure/government must privatize" (Ostrom, Schroeder, and Wynne 1993; Ostrom, Gardner, and Walker 1994; Klitgaard 1995). According to the NIE school, the public versus private dichotomy is inadequate, because forms of economic organization are myriad and many institutional configurations cannot be described accurately using such a simplistic classification. Recent theoretical work has expanded the traditional institutionalist framework. Elinor Ostrom (1994) and others describe arrangements under which participants voluntarily agree to cooperate to achieve some common objective without any expectation of remuneration. Because voluntary cooperation involves neither administrative fiat nor negotiated exchange, this approach to organizing economic activity cannot readily be described using the conventional categories of state (public) and market (private). A similar approach appears in the work of Chris Gerrard (1995), who proposes an alternative "pure type" classification of institutional arrangem e n t s consisting of hierarchy,

market,

and collective

action,

w h i c h he

refers to as allocation mechanisms for goods and services. Hierarchy is characterized by command and control from the top down and is a feature of governments and other large organizations. Political and sociological theory is often used to explain behavior within hierarchies. Markets are characterized by voluntary, impersonal exchanges of goods and services between buyers and sellers, such as among private agents or firms. Neoclassical economic theory is typically used to explain market behavior. A third, albeit less well understood, type of arrangement is characterized by collective action or participatory decisionmaking based on conventions, customs, or codes of conduct. According to Ostrom, Schroeder, and Wynne (1993), individuals will unite into formal and informal organizations for collective action to create jointly used goods and services if they perceive that the resulting benefits (including benefits shared with others) exceed the costs of the resources invested. Mancur Olson (1971) and others have identified some of the determinants of success in collective action.

A CONCEPTUAL FRAMEWORK

Relevance

to the Study of Maize Seed

45

Industries

What is the relevance of these various social science paradigms to the study of maize seed industries? The neoclassical model accurately attributes a utility-maximization motive to economic activity and correctly depicts the process by which supply and demand are equilibrated through negotiated exchanges taking place in a market context, making it a powerful tool for understanding any industry, including the seed industry. Its usefulness for applied analysis is limited, however, by the fact that it fails to account adequately for transactions costs. The depiction of fully informed, completely rational economic agents working within well-functioning, frictionless institutions is often misleading. The problem lies not in the assumption that decisionmakers act rationally but rather in the assumptions that information flows are perfect and transactions are costless. When these latter two assumptions are violated (as they nearly always are), numerous institutional problems arise that market forces in themselves cannot supersede (Klitgaard 1995). The institutionalist model avoids this problem by recognizing many of the institutional factors that influence behavior in the real world, but it does not fully explain where these institutions come from and how they change. 10 This shortcoming is particularly limiting in the context of economic development, because economic development is a process by which institutions undergo change over time. The Marxian model introduces dynamism into the analysis by allowing for the possibility of institutional change, but it asserts that change is driven by a deterministic historical dialectic that is essentially impervious to marginal collective action (according to the Marxian view, nothing short of revolution is capable of effecting fundamental change). The Marxian view thus leaves little role for discretionary incremental policy, limiting its value as a tool for applied policy analysis. This leaves the new institutional economics. The NIE paradigm incorporates many desirable features of the other three paradigms. It accommodates the utility-maximizing nature of most economic activity, recognizes the many institutional factors that influence real-world behavior, and endogenizes the process of institutional change by focusing on the rules and norms that determine the incentives for and constraints on the decisions of economic agents. In addition, it explicitly takes into account the important relationship between the characteristics of products and alternative forms of economic organization. Approaches based on NIE have been used to analyze labor markets, contract markets, and insurance markets in agriculture (see Bardhan 1989; Nabli and Nugent 1989a, 1989b). They have not been used extensively to analyze seed markets, however (significant exceptions include Butler and Marion 1983; Schmid 1987; Butler and Schmid 1984). The relative lack of

46

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

attention to seed markets is curious, because NIE concepts, when combined with ideas from the neoclassical and the Marxist paradigms, are extremely well suited for analyzing seed industries. Valuable insights can be gained by evaluating alternative forms of seed industry organization in terms of their ability to economize on transformation and transactions costs, given the basic attributes of hybrid and OPV seed. Knowledge of the relationships among product characteristics, organizations, institutions, and economic incentives can help in the formulation of hypotheses about the various outcomes that are likely to ensue from alternative forms of seed industry organization.

An Integrated Conceptual Framework An integrated conceptual framework combining ideas drawn from the various traditions might look something as follows. Individuals maximize their utility subject to economic and noneconomic constraints. They often find it advantageous to organize into groups (i.e., to form organizations) to pursue common objectives. The relationships among the members of an organization, between members and nonmembers, and among different organizations are governed by formal and informal rules. These rules or behavioral regularities constitute institutions. Although institutions can be defined in different ways, according to most definitions institutions are configurations of rules that (1) shape individual choices, (2) govern voluntarily or coercively the relations among individuals and groups, and (3) recur in predictable ways (Nabli and Nugent 1989b). Organizations of importance to the maize seed industry include public organizations, private organizations, and voluntary organizations, as well as others representing combinations of these "pure types." The three pure types provide a useful starting point for analysis by forming a triangular surface that represents the universe of possible organizational types. Most of the organizations associated with maize seed industries fall somewhere between the three extremes (Figure 3.1). Relationships within and between organizations are influenced by institutions. Institutions, which are shaped by historical circumstances and organizational efforts, include not only the formal rules, laws, and regulations that explicitly govern the activities of economic agents but also the wider set of legal regulations, cultural beliefs, and ethical norms that influence all human behavior. The institutions found in any given country at any given moment are determined, among other factors, by the ability of competing interest groups to exercise political, economic, and social power, which they use to put in place and perpetuate institutions that serve their own interests. Different types of organizations are better suited than others to certain forms of decisionmaking and tend to rely more heavily on those forms to

47

A CONCEPTUAL FRAMEWORK

Figure 3.1 Universe of Organizational Types Showing Selected Seed Industry Organizations Public

organizations

organizations

which they are particularly well suited. Economic transactions tend to be conducted hierarchically in public organizations, through negotiation in private organizations, and collectively in voluntary organizations. The relationship between organizational type and decisionmaking form is not deterministic, however. For example, public plant breeding institutes typically rely on hierarchical decisionmaking to guide production and consumption activities, but administrators frequently take market factors into account in determining the supply of inputs and assessing the demand for outputs. Similarly, private seed companies normally rely on market transactions to coordinate production and exchange activities, but within the firm decisions are often made hierarchically. Seed cooperatives are usually organized and sustained through collective action, but their management may pursue profit-maximization goals normally associated with private firms. Whereas alternative organizational-institutional combinations may or may not differ in their capacity to reduce transformation and transaction costs, they certainly differ in their ability to shift around transformation and transaction costs in ways that benefit some interest groups and hurt others. In the maize seed industry, as in any industry, the levels and distribution of benefits and costs depend on the interaction among the attributes of products, the organization used to provide those products, the institutional arrangements that shape the behavior of economic agents, market structural conditions, technology, and other factors. Since seed industry performance is measured in terms of the levels and distributions of cost

48

THE IMPORTANCE OF MAIZE SEED INDUSTRIES

and benefits, "good" performance (which is subjectively defined) depends on the ability of a particular combination of organizations and institutions to produce a desired set of outcomes. The likelihood that a desired set of outcomes will be achieved varies according to where different activities are distributed across the triangle (Figure 3.2). Whether a given type of organization actually succeeds in producing a desired outcome depends not only on the attributes of the product involved but also on the set of supporting institutions that are in place, on market structural conditions, and on technology. As these factors change, the appropriateness of the organizational type can also be expected to change.

Stylized Stages of Seed Industry Development The triangle representing the universe of possible organizational types can be used to depict a particular seed industry. Organizations active in the seed industry can be sketched in, with their location determined by their proximity to or distance from each of the three pure organizational types (public, private, voluntary). Although the resulting diagram provides a useful tool for seed industry analysis, it has at least one major limitation. Since the diagram represents a snapshot of a particular set of organizations

Figure 3.2 Universe of Organizational Types Showing Location of Selected Classes of Products Public

organizations

organizations

A CONCEPTUAL FRAMEWORK

49

taken at a fixed point in time, the relationships it depicts are essentially static. But for any given seed industry the diagram can be redrawn at different stages, providing a more dynamic view of the way the industry grows and develops. From a policy perspective, the important question is whether changes in the optimal organization of maize seed industries occur in some sort of deterministic manner, since this would imply that maize seed industries develop along a predictable path. Here, the empirical record may provide some guidance. Generalizing across countries, maize seed industry development often seems to follow the same basic path. Several analysts have advanced life cycle theories of industrial development based on the observed similarities. Four stages of growth are commonly described, although in reality these "stages" represent arbitrarily selected points along a growth continuum (Douglas 1980; Desai 1985; Pray and Ramaswami 1991; Jaffee and Srivastava 1992; Cromwell, Friis-Hansen, and Turner 1992; Rusike 1995; Dowswell, Paliwal, and Cantrell 1996). Characteristics of these stages are summarized in Table 3.1. Stage 1: Preindustrial

Stage

During the Preindustrial Stage, the organizations and institutions associated with a formal seed industry are absent. All of the materials grown are local OP Vs. Germplasm improvement occurs at the farm level as farmers (and their family members) carefully select superior ears from their own production to use as seed stock in the following planting season. Most farmers use seed saved from their own harvest, although some may also exchange seed with relatives or neighbors. Individual farmers are the dominant actors, because seed production technology is readily accessible to them and few incentives exist for commercial seed production. Despite their informal character (or perhaps because of it), preindustrial seed systems have been shown to be dynamic, flexible, and efficient at performing the essential activities of germplasm maintenance and improvement, seed production, and seed distribution (Almekinders, Louwaars, and de Bruijn 1994). The particular virtues of many preindustrial seed systems include their ability to provide small-scale producers with affordable seed of familiar materials in a timely and cost-effective manner (Cromwell 1990). Stage 2: Emergence

Stage

During the Emergence Stage, the realization that specialized knowledge and skills are needed to carry out germplasm improvement work leads to the formation of specialized research organizations. Since the predominant

O. B o u 00 « Ä

3

•P

C

o &•.' S S

o,

a 3 < K

ta

3 C C
Oh

o

O

J

c

2 M

«> C

B S

S B0

u z

O

>>

W) O

¿H

-O

ts >