RURAL DEVELOPMENT AND AGRICULTURAL GROWTH IN INDONESIA, THE PHILIPPINES AND THAILAND RURAL DEVELOPMENT AND AGRICULTURAL GROWTH IN INDONESIA, THE PHILIPPINES AND THAILAND Edited by Takamasa Alayarna and Donald F. Larson O 2004 The World Bank This work is copyright. Apart from those uses which may be permitted under the CopyvightAct 1968 as amended, no part may be reproduced by any process without written permission from the publishers. The views expressed in this book are those of the authors and not necessarily of the publishers. Published by Asia Pacific Press Asia Pacific School of Economics and Government The Australian National Unversity Canberra ACT 0200 Ph: 61-2-6125 4700 Fax: 61-2-6257 2886 Email: books@asiapacificpress.com Website: http://www.asiapacificpress.com National Library of Australia Cataloguing-in-Publication entry Rural development and agricultural growth in Indonesia, the Philippines and Thailand. Bibliography. Includes index. ISBN 0 7315 3786 6. 1. Rural development - Indonesia. 2. Rural development - Philippines. 3. Rural development - Thailand. 4. Agricultural productivity - Indonesia. 5. Agricultural productivity - Philippines. 6. Agricultural productivity - Thailand. 1. Larson, Donald F. 11. AkiYarna,Takarnasa. Editor: Asia Pacific Press Cover design: Annie di Nallo Printed in Australia by Paragon Printers CONTENTS Tables vii Figures xi Abbreviations xiv Contributors xviii Foreword, C. Peter xrnrner xxi 1 Lessons on rural development and agricultural growth fiom Indonesia, the Philippines and Thailand 1 Takamasa Akiyama and Donald F. Larson 2 An ecological and historical perspective on agricultural development in Southeast Asia 15 Yujiro Hayarni 3 Growth of the agricultural sector: are there peculiarities with Southeast Asia? 49 Takamasa Akiyama 4 Do different regimes distort differently? Maria Arnelina 5 Rural organisations and development: the social background for collective action 123 Shinichi Shigetomi 6 The political economy of rural development in Indonesia 162 Toshihiko Kawagoe 7 The political economy of Philippine rural development since the 1960s 214 Arsenio M. Balisacan, Nobuhiko Fuwa and Margarita H. Debuque 8 Four decades of development in Thailand's rural sector and the role of government 294 Shinichi Shigetomi 9 Effects of direct protection for agriculture in three Southeast Asian countries 380 Takamasa Akiyama and Kei Kajisa 1 0 Comparing agricultural exports from three Southeast Asian countries 415 Masayoshi Honma and Takehiko Hagino 1 1 Determinants of agricultural growth in Thailand, Indonesia, and the Philippines 455 Yair M u n u Donald E Larson and Rita Buaer 12 Intersectoral migration in Southeast Asia--evidence from Indonesia, Thailand and the Philippines 523 Rita Bumr, Yak Mundlakand Donald E Larson Index TABLES Land endowments for agricultural production in Indonesia, the Philippines and Thailand, 1965-96 Distribution of operational farm size and incidence of agricultural tenancy in Indonesia, the Philippines and Thailand, 1973-78 Growth of agricultural production in Indonesia, the Philippines and Thailand, 1961-95 Net exports as a percentage of total world exports of selected agricultural commodities from Indonesia, the Philippines and Thailand, 1961-95 Annual growth rate of agricultural GDP by regions Agricultural value added, land and labour Production, growth rate and land share for rice, by region Food supply in calories per capita per day and growth rate of food supply Share of pastureland in agriculture by region Gini coefficients for initial land distribution, by region, 1950-90 Annual growth rate of agricultural fured capital General protection rate from agricultural domestic price per border price, by region Transmission of border to domestic price, by region Relationships among !growth rates of agricultural GDP, manufacturing GDP, and GD1' Annual trends for agricultural exports, by region Elasticities of income gowth and sectoral growth, estimated using country dummies as fured effects Development of savings groups in Thailand Number and size of savings groups in Thailand, by region Size of cooperatives in Luzon region, June 2000 Distribution of different sized groups under NGO microfinance programs Distribution of Bank Desa and lumbung desd in Indonesia Number of desa with Bank Desa and lumbung &sa in Java, 1996 Repayment of loans from Bank Desa in Java, December 1999 Changes in economic indicators in Indonesia, 1961-96 165 Changes in demographic and social indicators in Indonesia, 1961-96 166 Percentage distribution of Indonesian GDP by industry, 1960-96 168 Chronology of rural development in Indonesia 180 Growth of rice production in Indonesia, 1965-90 189 Production and harvested area of major food crops, 1965-95 192 Value of major export crops, Indonesia, 1965-95 194 Production of major export crops in Indonesia, 1962-97 195 Production of major export crops and percentage of production, by type of farm operation, Indonesia, 1974 and 1994 197 Per capita GDP by Indonesian province, selected years at current prices 212 Contribution of gross regional product by industrial sector, by region, 1992 213 Per capita GDP and agricultural share of GDP in Thailand, Indonesia and the Philippines 215 Human development indicators, 1960-96 215 Agriculture and gross domestic product growth in Thailand, Indonesia and the Philippines, 1965-97 216 Aggregate economic performance in the Philippines, 1960-97 217 Growth of gross national product in the Philippines 217 Average !growth rate of gross value added (GVA) in agriculture, by sector, 1965-97 223 Crop or sector share in GVA to agriculture, 1965-97 224 Rural poverty estimates based on official measurements 228 Poverty by class of worker, 1997 229 Families by main source of income, 1985-97 229 Rural poverty, FIES and f ~ e dphysical areas, 1961-91 231 Rural poverty, Labor Force Survey data, 1977-83 233 Poverty estimates using official and preferred approaches, 1985-97 234 Access to water and sanitation, 1985-93 237 Trends in real wages 263 Urban-rural ~o~ulationchanges in the Philippines, 1960-90 292 Export value of agro-industrial commodities fiom Thailand, 1980-99 306 Export trends and share of exports, by commodity groups, Thailand, 1962-95 311 Per capita monthly food expenditures of rural and urban dwellers, 1962 and 1992 312 Exports of jute and jute-like fibres from Bangladesh and Thailand in comparison to world totals, 1961-97 320 Global tapioca exports, exports from Thailand and selected areas, Thai market share, and per cent of volume imported by Western Europe 324 Tapioca exports, 1981-97 327 Volume of chicken exported from +lhailand 331 Yield trends for seven crops, 5-year averages with base year of 100 341 Origin and share (per cent) of loan amounts during farm year per farm household 344 Average monthly expenditure per household by area 352 Migration patterns between rural and urban areas by age group 354 Share of employed persons by educational level 355 Share of crop income by commodity, 1988-99 360 Factors that promoted agricultural exports of major commodities 363 Period average of direct, indirect and total rates of nominal protection for selected commodities in Indonesia, the Philippines and Thailand 383 Short-run and long-run price instability of real border price and real wholesale price and estimates of the benefits of price stabilisation 387 Dynamic OLS estimation of rice and sugar, by country 393 Dynamic OLS estimation of maize and rubber, by country 395 Dynamic OLS estimation of oil palm and copra, by country 397 Range of value of over time for dif5erentcommodities in the Indonesia, Philippines, and Thailand 402 Agricultural eXp0I-t performance of Indonesia, the Philippines and Thailand Export quantity, value and unit value of major commodities, Indonesia, the Philippines and Thailand Composition of food exports fiom Indonesia, the Philippines and Thailand, by destination, 1965-97 ECM regressions to explain principal agricultural exports from Indonesia, the Philippines and Thailand Factors accounting for of agricultural exports fiom Indonesia, the Philippines and Thailand Regressions for the effects of direct and indirect protection policies on agricultural exports Estimates of export equations for major agricultural products Factors accounting for export growth of major commodities from Indonesia, the Philippines and Thailand Regressions for the effects of direct and indirect protection policies on exports of natural rubber from Indonesia and rice from Thailand Elasticities calculated from AIDS estimates of import demand equations for major commodities exported from Indonesia, Philippines and Thailand to selected markets Price transmission elasticities and ratios of import price to eXp0I-t price for major export products from Indonesia, the Philippines atld Thailand in selected markets Selected growth rates Growth rates in agricultural output and inputs Production function summary results Productivity, prices and shadow prices Sources of growth for Thailand, Indonesia and the Philippines Restricted production function: Thailand, 1971-95, without state vaiables Production function: Thailand, 1971-95, with state variables Production function: Thailand, 1971-95, base model Thailand: productivity, prices and shadow prices 11.1O Production function: Thailand, 1971-95, augmented base model 11.1 IProduction function: Indonesia, without state variables 11.12 Production function: Indonesia, with state variables 11.13 Production function: Indonesia, base model 11.I4 Production function: Indonesia, 1971-98, alternative specifications for land 11.15 Indonesia: productivity, prices and shadow prices 11.16 Sources of gowth for Indonesia 11.I 7 Production function: the Philippines, 1961-98, without state variables 11.18 Production function: the Philippines, 1961-98, with state variables 11.19 Production function: Philippines, 1961-98, with filtered data 11.20 Production function: Philippines, 1961-98, alternative specifications 11.21 Philippines: productivity, prices and shadow prices 11.22 Philippines: sources of growth using alternative parameters 12.1 Average migration rates for selected periods 12.2 Decade average income ratios, per worker, non-agriculture to agriculture 12.3 Principal components regression results for Thailand, 1962-99 12.4 Principal components and ordinary least squares regression results for the Philippines, 1962-98 12.5 Principal components and ordinary least squares regression results for Indonesia, 1972-99 FIGURES F.1 The 'Development Trilogy': three 'spheres' of activity, held together by agriculture 2.1 Map of Southeast Asia, indicating the continental and insular regions 5.1 Matrix for categorising microfinance organisations 5.2 Local administrative hierarchy in rural areas of Thailand 5.3 Local administration ladder in the Philippines Structure of local administration for rural finance organisations in Indonesia 143 Rural poverty and per capita GNP, 1961-97 232 Rural poverty incidence and per capita income growth, 1985-97 232 Share of agriculture in government expenditures and agricultural GVA 247 Share of agriculture in total loans granted by formal financial institutions 252 Titled lands distributed by administration, 1972-97 255 Nominal protection rate for various agricultural products, 1970-98 257 Trends in real exchange rate and debt-adjusted real exchange rate 262 Export share of primary Thai agricultural commodities 296 Changes in agriculture, agricultural policy, and the political economy in Thailand, 1945-97 299 Thailand 300 Rice production and consumption 3 12 Rice exports from major exporting countries 316 Expansion of Thai rice markets 317 Rubber exports from major ~roducingcountries 318 Destinations of Thai maize exports 322 Sugar exports from major exporting countries 328 Exports of canned pineapple from major producing countries 334 Pineapple-harvestedarea in the Philippines, Thailand and CGte d'Ivoire 336 Trend of farmland expansion and proportion of forest land 338 Incidence of rural poverty in Thailand 347 On-farm and off-firm income trends, 1976 prices 348 Distribution of per capita GDP and Gini coefficients among selected Asian countries 352 Predicted direct protection rates at mean value of real border prices in Indonesia 400 9.2 Predicted direct protection rates at mean value of real border prices in the Philippines 401 9.3 Predicted direct protection rates at mean value of real border prices in Thailand 401 9.4 Price stabilisation for rice in Thailand and the Philippines, 1961-95 403 10.1 Agricultural exports from Indonesia, the Philippines and Thailand, 1961-97 417 10.2 Agricultural exports from Indonesia, Thailand and the Philippines, 1961-97 419 10.3 Agricultural exports from Indonesia, Thailand, and the Philippines as a share of total commodity exports, 1961-97 420 11.1 Land-labour ratio 457 11.2 Ratio of irrigated land to agricultural land 458 11.3 Ratio of the value of agricultural capital to output 459 11.4 Ratio of fertiliser to land 460 11.5 Marginal productivity of fertiliser 471 11.6 Marginal productivity of irrigated land 471 11.7 Marginal productivity of capital 472 11.8 Marginal productivity of labour 472 11.9 Marginal productivity of rainfed land 473 11.10 Distortions in fertiliser markets 475 11.11 Labour share of agricultural GDP in Thailand 493 11.12 Labour share of non-agricultural GDP in Thailand 494 11.13 Ratio of approved investment to agricultural GDP in Indonesia 504 11.14 Labour share of agricultural GDP in the Philippines 511 11.15 Price ratio of fertiliser to crops in the Philippines 515 1I.16 Small firms and leased farms in the Philippines 519 12.1 Agriculture's share of total employment 524 12.2 Migration from agriculture in Thailand 526 12.3 Migration from agriculture in the Philippines 527 12.4 Migration from agriculture in Indonesia 527 12.5 Ratio of average income per worker, non-agricultureto agriculture 530 ABBREVIATIONS USED IN TABLES n.a. not applicable .. not available zero insignificant ABBREVIATIONS Anghatan Bersenjata Republik Indonesia [Indonesian armed forces] ACUR Australian Centre for International Agricultural Research ACPC Agricultural Credit Policy Council A&D alienable and disposable ADB Asian Development Bank AFMA Agriculture and Fisheries Moder~lizationAct AFTA APEC Free liade Agreement AIDS almost ideal demand system APEC Asia-Pacific Economic Cooperation ARDL autoregressive distributed lag AEU' Agricultural Rehabilitation Plan ASLW Association of Southeast Asian Nations BAAC Bank of Agriculture and Agricultural Co-operatives BAPPENAS Badan Perencanaan Pembangunan Nasional [National Development l'lanning Agency] BD Bank Desa BHS basic human needs BOI Board of Investments BOT Bank of Thailand BPS Baddn Pwat Statistik [Central Statistics Agency] BRI Bank Rakyat Indonesia BULOG Bahn Umsan Logistik [State Logistics Agency] BUUD Bahn IJnit Usaha Desa CAZF Comprehensive Agricultural Loan Fund CARD Center for Agricultural and Rural Development CARP Comprehensive Agrarian Keform Program CBN cost of basic needs CCSF Coconut Consumer Stabilization Fund CDA Cooperative Development Authority . CDD Community Development Department CSIS Center for Strategic and International Studies CMSA constant market share analysis CP Charoen Pokphand CPT Communist Party of Thailand CRIS Cooperative Information System Cl'RP Comprehensive Tax Reform Program DOF Department of Factory DOL Department of Land DPA Dewan Pertimbangan Agung [Supreme Advisory Council] DTI Department of Trade and Industry, the Philippines ECM error correction model ERP effective rate of protection ESCAP Economic and Social Commission for Asia and the Pacific Facoma Farmers Cooperative Marketing Associations FA0 Food and Agriculture Organization FIES Family Income and Expenditures Survey FM-OLS fully modified Hansen-Phillips least squares approach GATT General Agreement on Trades and Tariffs Golkar Golon~nI h y Uoint Secretariat of Functional Groups] GVA gross value added HPAE High Performing Asian Economies HYV high-yielding varieties W E CRC Institute of Agribusiness Development and Policy, Center for Research and Communication IFAF Integrated Farmers' Aid Fund IID independent identically distributed ILP import liberalisation program IOCR incremental output-capital ratios IPP Investments Priorities Plan I l T Indonesia, Thailand and the Philippines IRRl International Rice Research Institute ISH Integrated Survey of Households IS import substitution IS1 import substitution industrialisation JETRO Japan External liade Organization KUD Kopmusi Unit Desa [Village Unit Cooperative] LFS Labor Force Survey xvi RURAL DEVELOPMENT AND AGRICULTURAL GROWTH LKMD Lernbaga Kkahanun Mqarakat Desa willage Council for Development Planning and Guidance] LMD Lmbaga Musyawarab Desa pillage Consultative Council] MAV minimum access volume MDU Mobile Development Unit MOA Ministry of Agriculture MOAC Ministry of Agriculture and Cooperatives, Thailand MOC Ministry of Commerce, Thailand MOF Marketing Organization for Farmers MOF Ministry of Finance, Thailand MOI Ministry of Interior, Thailand MPR Majelis Perrnusyawaratan Rakyat [People's Consultative Assembly] NAIC newly agro-industrialising country NASUTRA National Sugar Trading Corporation NEDA National Economic and Development Authority NEDR National Economic Development Board NESDB National Economic and Social Development Board NFA National Food Authority NGA National Grains Authority NGO non-government organisation NPR nomind protection rate NRDP National Rural Development Program NSCB National Statistical Coordination Board 0AE Office of Agricultural Economics OLS ordinary least squares OUSS Office of the Under-Secretary of State, Thai Ministry of Agriculture PAP Poverty Area Program PC principal components PCMS Philippine Coalition for Microfinance Standards PKI Partai Kornunh Idnesia [Indonesian Communist party] PNB Philippine National Bank PO people's organisations PRRM Philippine Rural Reconstruction Movement PWO Public Warehouse Organization RAPS Restructuring and Production System Plan for Agriculture RBP real border prices RCA Rice and Corn Administration RFD Royal Forestry Department ROSCAS rotating savings and credit association RT rukun temngga,or precinct-level neighbourhood association RW rukun warp [ward-level neighbourhood association] SACOP Social Action Center of Panpangas SHG self-help group SN Samahang Nayon SRA Social Reform Agenda SSR self-sufficiency ratio SUR seemingly unrelated regressions TAO Tambon Administrative Organization TDRI Thailand Develoynlent Research Institute TFP total factor productivity TRI Cane Intensification Program TRP tariff reform program TSC Thailand Sugar Corporation TSEG two-step Engle-Granger TSTC Thai Sugar Trading Corporation UNDP United Nations Development Programme xviii RURAI. DEVELOPMENT AND AGRICULTURAL. GROWTH CONTRIBUTORS TakarnasaAkiyama is Senior Advisor at the Foundation for Advanced Study on International Development (FASID) and Visiting Professor at the National Graduate Institute for Policy Studies (GRIPS) in Tokyo. He was with the World Bank in Washington when this project started. While at the Bank,he worked on rural development issues. He is currently working on more general development assistance policy issues for the Japanese development community. Maria Amelina is a Senior Social DevelopmentSpecialist at the World Bank. She specialises in the political economyof !governancein transition countries, with an emphasison the dynamics of power in rural areas and the impact of agricultural policies. At the time of writing, Ms. Amelina was working on more general issues related to the political economy of agriculture in developing countries. Arsenio M. Balisacan is currently Director of the Southeast Asian Regional Center for Graduate Study and Research in Agriculture (SFARCA) and Professor of Economics at the University of the Philippines, Diliman. He has served with the Philippine Government as Agriculture Undersecretary (Deputy Minister) and with various United Nations agencies as adviser and consultant. His recent research has focused on poverty, economic inequality, agricultural and rural development, and the political economy of policy reforms. His articles have appeared in numerous books and professional journals. Rita Butzer is a PhD candidate in the Department of Economics at the Universityof Chicagowith interests in agricultural economics and economic grow&.Shewas a visitingfellow in the Deparunent of Economicsat Harvard University and has worked as a consultant with the World Bank and the Japanese Center for Economic Research. MargaritaH. Debuqueholdsan MAdegree in economicsfrom the University of the Philippineswhere she is finishingher doctoral studies. She is currently doing research on monetary policy and currency arrangements in emerging Asian economies. She is also an economic journalist and has consistently won awards for her investigative and analytical pieces. Nobuhiko Fuwa is International Research Fellow at International Rice Researh Institute (IRRI),Visiting Research Associate at the University of the Philippines at Diliman and Associate Professor of AgriculturalEconomics at Chiba Unviersity. He was on the staff of the World Bank when this project started. His recent research interests include poverty dynamics, household behaviour in rural areas and gender and development. Takehiko Hagino has an MA from Keio University and, during the drafting of this book, worked as an economist at the Asian Research Bureau and Research Service Department, Japan Center for Economic Research. He specialises in the economies of South East Asia. Yujiro Hayami is Chairman of FMID Graduate Faaculty of Internatio~~al Development Studies in Tokyo. He specialises in development and agricultural economics, with an emphasis on institutional and technological changes in East Asia in comparison with other developing regions. He has served as an economist at the International Rice Research Institute and is currently the president of the Asian Society of Agricultural Economists. Masayoshi Honma is Professor at the Graduate School of Agricultural and LifeSciences at the University ofTokyo. His areasof expertise are agricultural and international economics with an emphasis on agricultural trade, trade policy, and political economy of agricultural policies. Kei Kajisa is Faculty Fellow at the Foundation of Advanced Studies on International Development (FMID) and Associate Professor at the National Graduate Institutefor PolicyStudies (GRIPS)in Tokyo. His current research interest is collective action and rural development. ToshihikoKawagoeis Professorat theFacultyofEconomics,SeikeiUniversity, Tokyo. He specialises in rural development, with an emphasis on rural marketing, land reform, and agricultural policies in Asia, especially in xx RURAL DEVELOPMENTAND AGRICULTURAL GROWTH Indonesia and Japan. He had worked as an economist for the World Bank and UNESCAI? Donald F. Larson is a Senior Economist with the World Bank's Research Group. He specialises in development economics with an emphasis on the agricultural policy. His research focuses on the determinants of agricultural growth and on the roles of risk and markets on rural livelihoods. Yair Mundlak is retired. He formerlyserved as Ruth Hochberg Professor of Agricultural Economics, The Hebrew University of Jerusalem, and F.H. Prince Professor of Economics,The University of Chicago. He has worked on various aspects of agricultural development. Shinichi Shigetomi is a researcher at the Institute of Developing Economies (IDE),JETRO, Japan. His main research interests are rural organisations and agricultural economic$ in Thailand. Currently he is a visiting scholar at Thammasat University of Thailand and is conducting a research project on non-governmental organisations. FOREWORD This volume, the culminationof more than a half-decadeof research, provides convincing evidence that agriculture is 'back on the agenda' of developing countries and donor agencies alike. The evidence comes in two forms. First, the volume was sponsored by the World Bank, with Japanese funding, with the explicit objective of correcting a major oversight in The Eat Asian Miracle (1993): the failure to explain the role of agricultural and rural development in the spectacular success of these economies in the latter third of the 20th century. Second, the chapters in the volume themselves demonstrate just how important that role was-in both positive and negative w a y i n stimulating the rest of the economy. Much of the analysis focuses, appropriately, on different policy settings. Importantly, the historical and ecological contexts for the choices available, described by Yujiro Hayami, are clearly not determining. Even within these constraints, the three countries treated in detail in this volume pursued widely varying approaches to their rural economies. This variance, explained by the country authors primarily on social and political grounds, nonetheless had powerful, long-run consequences. In some sense, all three countries are living today with policy choices made decades before. Why should this be? One of the great challenges in economic history- and development economics-is to explain the apparent contradiction benveen a positive role for agriculture in the development process, and the resulting structural transformation. In this transformation, the share of agriculture in the economy declines while industry rises and urbanisation takes place, especially because of rural to urban migration. It is easy to see how agriculture would get short-changed in the policy process if only the end points of the transformation are understood, not its dynamics. Both the detailed country studies and the cross-country analyses demonstrate the power of the structural transformation and the complexity of its dynamics. Some very basic forces must be at work, but these forces play out in their own country-specific ways. This volume is especially helpfd in providing the reader a detailed panorama of this specificity. Still, it is useful to review briefly the basic role of agriculture in this process, to set the stage for the detailed studies. An easy way to organise the xxii RURAL DEVELOPMENT AND AGRICULTURAL GROWTH discussion is in terms of a 'development trilogy', a triangle with 'rapid economic growth' at one apex, 'poverty reduction' at a second, and food 'security' at the third. In the middle of the triangle, connecting all three topics, is the agricultural sector. Food security is equally important to national leaders and heads of households,and so is a good place to stm. A powerfd lesson from economic historyis that no country has sustained the economic process without guaranteeing food security to the great majority of its citizens. Food security has two basic dimensions: at the national, or market level; and at the household, or consumer level. At the national level, the main concern is for reliable availability, which in a market economy is signalled by the level of prices for the staple food- rice in Asia-in the major urban ~narkets.There is a secondary concern for the stability of these prices. Historically, stabilising rice prices has been a serious and difficult problem in Asia, but attempted by all countries nonetheless, including the three treated here. Instability of rice prices has been much less of a problem in the past decade because the world rice market has become much more heavily traded and stable, like the world markets for wheat and maize. The issues of both the price level and price stability also emphasise that food security is a trade-related topic, not just a production topic. Indeed, 'food self-sufficiency' as either an analytical topic or political objectiveshould be eliminated from the language! Only desperately poor people are truly self-suflicient. At the household level, food security is a matter of reliable access to food, from own production or from the market. This access to food is primarily a hnction of household incomes, of food prices, and of nutrition knowledge of the mother. This knowledge is what translates 'food access' into meals for the family and nutritional well-being for individuals. The link between the household and market levels of food security, and the link to economic growth and poverty reduction, is the productivity of domestic agriculture. To re-em~hasisethe point about trade, this does not necessarily mean how much rice the country grows. In poor countries, the majority of the population, and most of the poor, live in rural areas. Their primary vehicle for sustaining food security-at household and national levels-is rising agricultural productivity. In the long run, this means increasing labour productivity, and non- farm employment and migration are important mechanisms. But in the --- -- Figure El The'Development Trilogy': three'spheres' of activity,held together by agriculture - Rapid economic growth securiry reduction short run,the main vehicle for raising agricultural productivity is to raise the value of output per hectare. And this almost always means using new agricultural technology: high ~ieldingvarieties (HYVs) of seeds, improved livestock genetics and feeds, fertiliser, mechanical equipment, farmer knowledge and management skills, especially in diversifying into higher- value activities. This train of logichas been verified historically-includingin the chapters presented in this volume-and tested econometrically. It runs from improved agricultural technology to higher productivity, to food security in markets and households, and from there to sustained economic growth and the reduction in poverty. RURAL DYNAMICS AND ECONOMIC GROWTH There is a good deal of evidence about these links from agriculture to economic growth, and about why a dynamic rural economy makes this growth more 'pro-poor'. First, it is important to remember that almost anything that speeds up the long-run rate of growth will also speed up the reduction of poverty-the record of economic history is very clear on this relationship. Second, agriculture can play a key role in this process, especially in poor countries where it has a heavy weight in GDP and in the labour force. xriv RURAL DEVELOPMENT AND AGRICULTURAL GROWTH There are three sets of linkages between agriculture and economic growth discussed in the literature: direct, indirect, and roundabout. 'I'he direct linkageswork through markets.The factor market mechanisms, often called the 'Lewis Linkages' because VC! Arthur Lewis analysed them in his 'surplus labour' model of development, involve labour and capital. The product market mechanisms, often called the 'Johnston-Mellor Linkages' - because of their early analysis of them, involve providing food for cities, demand for domestic industrial goods, supply of raw materials to industry, and the earning of foreign exchange to import capital equipment and technology. Perhaps the most important market effect has been the declining rural-urban terms of trade--cheaper food-which is hugely important to the poor.This linkshould probably be named after D. GaleJohnson, because he documented its impact so clearly and stressed its importance. 'I'he indirect linkages work through higher total factor productivity (TFP). As emphasised in the chapter by Mundlak, Larson, and Butzer, growth in TFP is the ultimatesourceof sustainedeconomicgrowthand higher standards of living. Agriculture contributes in three ways. First, the efficiency of decisionmaking in firm households is higher than in the rest of the economy, a point stressed by TW. Schultz. Farmers allocate their scmce resources very carefully, and get maximum output from new resources. Second, there is low opportunity cost to resources availablein abundance to rural households, especially their labour. This means that finding more productive activities for rural labour costs the economy very little. Third, farmers are able to circumvent the low capacity for financial intermediation in rural areas by investing directly in productive activities on their firms. This leads, at the margin, to rising output with little measured capital input, but can only happen when farming is profitable. Higher output with few measured inputs is a sure way to raise total factor productivity! The roundabout linkages work primarily through the removal of urban bias, and should be named after Michael Lipton, for his work, WhyPoor People Stizy Poor: urban b h in heloping countries. Urban bias has a profound impact on both the rate of economic gowth and its distribution. Urban bias is a political economy problem, where rent-seeking and urban-based governance sharply distort rhe allocation of resources. The evidence shows that it is not just agriculture that suffers from urban bias; so too does the entire economy and, especially, the poor in both rural and urban areas. MAKING GROWTH PRO-POOR Why would 'Getting Agriculture Moving', to use the title of a famous book in the 1960s by Art Mosher, not only speed up economic growth, for the reasons just listed, but also make it more pro-poor, at least in Asia?To answer this question, John Mellor has modelled the growth experience seen in these economies. His model emphasises the role of producing rural nun-tmddbles that are locally consumed-processed foods, construction, trade, and small-scale manufactures-as the 'ladder' for underemployed workers in agriculture to begin the climb to modern jobs at higher wages. In most poor, rural economies this non-tradable sector is demand- constrained. That is, expanding it, and the number of jobs it creates, does not depend on better access to capital or to management skills, but to greater purchasing power among local consumers. Thus Mellor points to rising profitability of agriculture-through higher productivity, not higher prices (because higher prices just choke off demand except for farmers with significant surpluses to sell)-and also to the wages of workers in a rapidly growing manufacturing export sector. This Mellor model is basically a 3-sector version of the standard Lewis model of the dual economy. In his version there are two 'commercial' sectors-industry and agriculture (which now uses modern technology and is market-driven)-and the 'non-tradable' sector (mostly rural but also urban and informal). The commercial sectors are the 'engines of growth', but connecting them to the 'non-tradable' sector is the key to a high 'elasticity of connection' between overall economic growth and rapid poverty reduction. A rapidly growing rural non-tradablessector absorbs surplus labour from agriculture and causes wages of unskilled labour to rise. This is the key to rapid reductions in poverty, and clearly, agriculture is half the story. The structural transformation emphasises the importance of the industrial sector in the long-run growth process, rightly, but it is important not to forget the role of agriculture as well. C. Peter Timmer Vice President and Chief Economist Development Alternatives, Inc., Bethesda, M D LESSONS ON RURAL DEVELOPMENT AND AGRICULTURAL GROWTH FROM INDONESIA, THE PHILIPPINES AND THAILAND TakarnasaAkiyama and Donald E Larson Central to the study of development is the understanding of economic of which rural economic growth is almost always an important aspect. Historically, rural agriculture has employed most people in most countries; in many countries it continues to do so to this day, especially in poor countries, where most householdseither live in or have familial ties to a ricultural areas. Nevertheless, the causal relationshipsbetween economic g growth and gowth in agriculture remain poorly understood. The lack of understanding stems partly from a more general debate concerning the relative contributions of geography, technology, private and public investment, trade and institutions to economic growth. These factors, well studied at the national level, take on distinctivecharacteristics and links in rural areas, where accumulations of human capital, infrastructure and governinginstitutions are often limited. Untangling the causal relationships between economic growth and growth in agriculture is made all the more difficult by the constantly evolving ways in which the above-mentioned factors shape and are shaped by policy. 2 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Thechaptersin thisvolumeexplore the manyaspectsofeconomicgrowth in the agriculturalsectors of Indonesia, the Philippinesand Thailand. The study period is generous, allowing the researchers to draw comparisons across time as well as space. The countries themselves invite comparison, sharing a monsoon climate conducive to rice cultivation on smallholder farms and a particular organisation of rural households and communities. These factors meant that the three countries' agriculturalsectors were well positioned to take advantage of the emerging technologies of the Green Revolution. Hard-won gains in productivity made food more abundant. Exports, especiallyof perennialcrops, becameimportant sourcesof revenue, and developed along with evolving technologies.The ways in which these factors interacted with policy, investments, resource constraints and the reallocation of agriculturallabour are the underlying theme of this volume. Looking back 40 years, the three countries' similarities become even morestriking.The Philippineswaslargelyrural,and IndonesiaandThailand overwhelmingly so. Incomes in all three countries were similarly low, and the then-new technologiesof theGreen Revolutionspurred growth.Growth in agriculture led to declining rates of poverty; in this respect the three countries initiallyfollowed similar paths. Differencesemer ed in the 1980s, g when growth in agriculture and the economy as a whole faltered in the Philippines, but continued in Indonesia and Thailand, to the point where both countries participated in what became known as the 'East Asian Miracle' (World Bank 1993). As the followingchaptersdescribe, thesecountries'similar characteristics and similarearly growthexperiences masked eventualconstraintson growth and other differences among them. Implemented policies and their supporting institutions (including regime type, state ideology and political system) are rooted in specific political and socio-economic histories. For example, Indonesiaand the Philippineshave a long history of colonial rule, whileThailand remained largelyindependent. The populations ofThailand and the Philippines ate roughly the same, while Indonesia's is nearly three times this size. The ethnic makeups of the countries differ. Indonesia has more than 200 ethnic groups, whileThailand and the Philippines are more ethnicallyhomo eneous, though what inter-ethnic tensions exist are long- g standing For the study period, Indonesia was organised politically as a unitary republic under a strong presidency that emphasised unity and centrally controlled economic development. For much of the study period, the Philippines was organised politically as a unitary republic, but with RURAL DEVELOPMENT AND AGRICULTURAL GROWTH 3 numerous politicalparties;however,the countryalsoexperiencedasustained period when political participation was limited. Meanwhile, Thailand was pverned largelyas a constitutional monarchy,with a government dominated by military and bureaucratic tlites. These differenceswere reflected in differing national and sector policies, thus they became apparent even as governments pursued similar objectives. In turn, the consequences of policy accumulated in agriculture's productive factors-especially human and physical capital. Since policies also have political consequences, when the East Asian economies faltered in 1997, the sustainability of growth came into question. The chapters that follow document the interplay of endowments, technology, the accumulation of productive factors, policy, and advocacy in the rural sectors of these three countries. Each contributes in its own way to an explanation of the past, and it is the editors' hope that they also provide insights into how appropriate policy can improve the lives of rural people. ECOLOGY AND INSTITUTIONS The process of !growthfollows a path, and any study of growth must find a starting point at which to begin on this path. Becauseagriculture is a natural process, natural endowments of climate and geology and their interactions with production technolo iesand commerce playa crucial rolein the choice g and organisation of rural economicactivity. In turn, choice and organisation of rural economic activity give rise to traditions and institutions, and come to be embedded in accumulated factors of production. These relationships are dealt with in Chapter 2 by Yujiro Hayami, who examines the ecological and historical roots of agrarian structure in the three countries. Hayami begins his analysiswith a description of the practices that prevailed in rural Indonesia, the Philippines and Thailand prior to the 1860s, when new transportation technologies gave the region's goods better access to international markets. At the time, wet rice production was the dominant agricultural practice in the rural communities of the region, and, from an ecolo ical perspective, the three countries' differences related primarily to g vast areas of unused land. With the development of new transportation technologies came new sources of demand and incentives to open up previously inaccessible tracts of land. Thus, previously unused resources acquired new value, and economic growth occurred. In insular Indonesia and the Philippines, the 4 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH combination of ecology and prevailing technologies promoted the development of tree crops, while in Thailand it was rice exports that grew. For the past three decades, the ratio of permanent crops to paddy rice has remained higher-and more stable-in Indonesia and the Philippinesthan inThailand.As Hayami points out, certain permanent crops-for example, sugar and bananas-require scale economies in processing and marketing which promote plantation agricultureoversmallholderagricultureand thus lead to the concentration of rural land use. In a similar way, the expansion of canals required to access remote areas of cultivatable land in Thailand also led to a concentration of land ownership. Hayami argues, however, that prevailing institutions and policies-their roots both pre-colonial and colonial-also influenced the interplay of ecology and technology. In Thailand, methods of rice production and storage worked together with traditional land policies to limit the concentration of agriculturalland and tenant farming. In contrast,in the Philippines,the colonialSpanish policyof assertingproperty rightsover uncultivatedand communallandsled to higher levelsoflandlessness.Meanwhile,in Indonesia, Dutch landpoliciesestablished longterm leases for plantations but also left in place traditional communal land-sharing arrangements giving smallholders greater access to land. Twentieth-century advances in agriculture led to scarcer land resources and more intensive production. Hayami argues that these developments affected plantationsand smallholdersdifferently,with smallholderssuffering less in terms of constraints on productivity than plantations. Consequently, the more smallholder-friendly land policies of Thailand and Indonesia alloweda ricultureto continue to expand,whereasin the Philippinesgrowth g stalled. Moreover, Hayami argues, policies in Thailand and Indonesiawere more favourable for a riculture than policies in the Philippines, partly g reflecting differences in political power structures with their roots in differencesin the distribution of wealth and land. In Chapter 3,TakamasaAki~amaasksifadditionalfactors, whetherrelated to or independent of the re ion's ecological characteristics, distinguish g SoutheastAsia's agricultural sectors from those of other developing regions. In doing so, he revisits some of the themes discussed by Hayami and introduces new themes as well. For Southeast Asia as a whole, growth in agricultural income was sustained for nearly 40 years into the twentieth century. Within the region, Indonesia, Thailand and China experienced consistent gowth, while income in every country grew faster than RURAL DEVELOPMENT AND AGRICULTURAL GROWTH 5 agricultural labour supplies, though gowth in the Philippines after 1981 was slow. The only other developing region where per capita agricultural income grew more rapidly was Latin America. Akiyama notes that the rapid gowth of agriculture in the region as a whole was due in part to the importance of rice and the effects of rice- specific new technologies. He also points out, however, that in Indonesia and Thailand expanding exports played a similarly important role. Among these expanding exports, Akiyama highlights palm oil, rubber and cocoa; forThailand he could have mentioned sugaras well. He attributesThailand's success to its ability to shiftinto higher-valueproduction, whilefor Indonesia he identifiesexpandin volumes as the crucialfactor. It is also worth noting g that both these countriessuccessfullymatched smallholderproduction with the centralisedprocessingusuallyassociatedwith plantation crops.Examples include rubber and sugar in Thailand, and, from the 1980s onward, palm oil in Indonesia. Akiyama also notes that the agricultural sectors of Southeast Asia have tended to build up their capital stocks, even as arable land per worker has declined because of rapidly expanding populations. He claims that equity also played a role, with land distribution occurring generally more evenly in East Asia than in other regions, and growth in the region as a whole having a larger impact on poverty than in other regions. In Chapter 4, Maria Amelina uses case studies of sugar and rice policies, which were very similar in the three countries, to examine the relationship between agricultural policies and the bases of political power. For rice in particular, the similarities among the three countries derived partly from the technical aspects of opportunities provided by the Green Revolution. In time, key features of the policies' implementation diverged. Amelina characterisesthe relationshipbetween technicalprogressand politicalpower in Indonesia and the Philippines as the 'authoritarian bargain', implying regimelegitimacybased on the abilityto deliverdevelopment.ForThailand, on theother hand,shedescribesa more pluralisticsystem in which traditional support for the King and parliament serve as a basis of political power. All three regimes lent support to the development of new technologies and took measures to benefit from them. In Indonesia.and the Philippines, however, the success of such development depended directly on public support for regimes that lacked both traditional and democratic legitimacy. Moreover, the evolving bureaucraticcontrols developed to deliver technical 6 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH progress actually came to be used to redirect public resources to supporters. According to Arnelina, this abuse of power created both the urgent need for successand obstacles tosuccessat thesame time. InThailand, the checks and balances of pluralism limited the capacity of agricultural policies to placedemands on public resources,and unsuccessful programswereallowed to end. Once the authoritarian bargainhas been struck and internalsupport bargained for, however, any admission of failure will shake the very foundations of the regime's legitimacy. In Chapter 5, at the other end of the political spectrum, Shinichi Shigetomi examines self-organising rural organisations. Rural areas are organised administratively, and the resulting administrative units become the formal conduits for governance.Voluntary rural organisations, however, can occur in a multitude of settings and, since they are self-organising, need not correspond to spatially defined administrative units. Shigetomi compares micro-finance organisations in the three countries to draw inferences about underlyingrelationships.He finds that formal and informal institutions critically affect the structure, size, management style and performance of rural organisations. Mainly because of differences among their respective institutions and government policies, the three countries' micro-finance organisations differ significantly in spite of the similarities in their rural financing market conditions. In Thailand, because the administrative village coincides with the traditional indigenous village, most of the saving groups have relatively large memberships and management guided by collective consent, and can utilise formal institutions in their decision making. In the Philippines and Indonesia the situation is quite different. In the Philippines, the micro- finance organisations are ~redominantl~of the Grarneen type and small in membership (usually about five), and are guided mainly by collective consent. The development of this form of organisation stemmed from the difficultyof formin larger organisations in the Philippines, one reason for g which was diver ence of the administrative village unit from traditional g community lines. In Indonesia, besides the formal and highly institutionalised administrative-levelfinancial organisations, there also exist numerous informal organisations with membership of between 20 and 30. In many parts of Indonesia, traditional locality-based communities developedinto micro-financeorganisations,which, because of the mismatch betweenthe administrative village unit and traditionalcommunities, tended to establish linkages with large formal institutions. RURAL DEVELOPMENT AND AGRICULTURAL GROWTH 7 COUNTRY STLDIES The next three chapters of the book describe agriculture's changing role in each of the three study countries, with each of the authors providing a different perspective on the topic. In Chapter 6, Toshihiko Kawagoe examines the political economy of ruraldevelopmentin Indonesia,throu h a reviewofthedevelopmentprocess g since the colonial period and a look at current rural development policies. Kawagoeargues that the ideas of Indonesian policymakersin the earlystages ofthe NewOrder rulewereaffected by thecultural,historicaland ideological factors of the idea of rural pessimism, anti-colonial sentiments and the state ideology ofpancdcikz, as well as the social and ideological factors of state nationalism, the potential for cooperatives to bring benefit to economicallyweak groups, and family farming, as in the 1945 constitution. Over the study period, Kawagoe notes, agriculture declined in relative economicimportance. Politically, the sector was treated with ambivalence. Formerly Dutch-owned plantations were expropriatedand managed to the benefit of the state or ruling coalitions, whilesmallholder export crops, for example coffee and coconuts, were largely ignored up until the 1980s.' Kawagoe argues that this historical ne lect influenced policymakers' ideas, g encouragingruralpessimism,exportpessimismand anti-colonialsentiment. In contrast, rice was accorded a special place in policy, and significant political and financial resources were exploited to bring the benefits of the Green Revolution to Indonesian rice farmers. Rice yields increased dramatically, and rice self-sufficiency was attained in the mid 1980s. Kawagoe argues that the support given to rice had an emotional aspect, partly explainable by the family principles that prevailed in rural areas. He alsoargues, however, that agricultural policyimplementation was hampered by factors such as the state's distrust of markets and promotion of village cooperatives,and that the all-out drive towards national unification led to the neglect of regional diversity in agricultural production planning. In Chapter 7,ArsenioBalisacan,Nobuhiko Fuwaand MargaritaDebuque examine agricultural developmentin the Philippines,and discusswhat they describe as policy failures. In spite of the successes it achieved in the 1950s and early 1960s,during the subsequentfour decades the Philippineslagged far behind the other study countries in overall economic growth, poverty reduction, human capital development and structural transformation. A markeddifferencebetween ruraldevelopment performancesbeforeand after 8 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH the mid 1980s is discernible.During the 1960sand 1970s,both the overall economy and the agriculturalsector recorded respectable growth rates.As a result of capital-intensive and foreign debt-driven growth and political turmoil, the 1980s experienced a series of booms and busts accompanied bymacro-stabilisationattemptsand politicalinstability.Agricultural gowth also slowed during this period, because of a combination of factorssuch as stagnation in public investment, the apparent exhaustion of Green Revolution potentials, and negative price incentives due to government policies. Unlike the rates of overall and agricultural growth, the pace of rural poverty reduction was disappointingly slow, though it appeared to pick up at the end of the 1980s, despite the boom-and-bust cycle and relatively low average growth rates of that decade. The authors argue that the heavy emphasis placed in development strategies on import-substituting industrialisation was a source of rural discrimination and that additional anti-smallholder policies compounded the effects of thisdiscrimination.Treatingsuch policiesas natural outcomes of the Philippinespoliticaland economicsystem,which hadlongsuppressed competition and pluralism, the authors trace the roots of the system to the creation of a landed oligarchy during the Spanish colonial era. In Chapter 8, Shinichi Shigetomi discusses key aspects of rural developmentinThailand. In his analysis of agriculturalcommodityexports, Shigetomi highlights Thailand's strong export crop performance and its contributions insustaininggrowthand reducingpoverty.Attributing export success in general to the private sector's capacity to respond to market conditions, Shigetomi does not treat the role of the Thai government as negligible or negative, but on the contrary asserts that some governmental policies had positive and even crucial effects on some commodities. Beneficial policies include programs of credit provision in rural areas and extension services on new crops, especially for high value-added and processed products; provision of services unavailable through market transaction, such as seed improvement of open-pollinated plants and irrigation;investment promotion, tax exemptionlreduction,and negotiation with foreign governments to reduce import duty; subsidies to reward improved productivity, for example in rubber re-plantation programs; and establishment of quality standards and export crop inspection systems. Agribusiness res onded positively to these policies, and the composition of p agricultural exports shifted from traditional to more value-added crops. RURAL DEVELOPMENT AND AGRICULTURAL GROWTH 9 Another successful aspect of rural development in Thailand was the reduction of rural poverty in absolute terms. Before the 1970s, both agriculturaland non-agriculturaldevelopmenthelped bringabout reductions in poverty. From the 1980s onwards, however, farm income decreased in real terms, while non-farm income increased. Shigetomi argues that this income structure is the cause of the Thai government's failure to narrow the income gap between the rural and urban sectors.This followed naturally from the Thai farm sector's reliance on the market mechanism for development. At the end of the 1980s,world commodity pricesstagnated, and Thailand's farm sector, which relied heavily on the export market, was hit particularly hard. Meanwhile, the private sector actively led industrialisation throughout this entire period. Currently, the Thai rural sector finds itself in unfavourable conditions, and Shigetomi argues that rural Thai people need to develop their skills as farm operators or wage earners in order to overcome the difficulties facing them. THE ROLE OF TRADE Trade is widely viewed as East Asia's primary engine of growth. This view usually focuses on the region's growing manufacture exports; however, sustained agricultural growth in Thailand and Indonesia is also associated with the steady expansion of commodity exports, especially sugar, rubber and rice in Thailand and palm oil in Indonesia. Like domestic production and consumption, trade depends on a combination of factors, some related to internallydetermined policiesand others toexternallydeterminedpolicies. Examinations of domestic trade policy and external determinants are taken up in two complementary chapters. In Chapter 9, Takamasa Akiyama and Kei Kajisa analyse agricultural protection rates and find that indirect protection rates have remained relatively stable and negative over time, while direct protection rates for many commodities have increasedsince the early 1980s. By the mid 1990s, total average protection rates for Indonesia and Thailand had become positive. However, indirect protection in the Philippines dropped to very low levels, making total average protection rates negativein the late 1990s. Akiyama and Kajisa also find that, over the study period, direct protection rates for many commodities in the Philippines and Thailand converged to zero, bringing domestic agricultural commodity prices more into line with international prices.This reflectsthesecountries' liberalisationofagricultural 10 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH commodity markets and reduced protection due to the achievement of self-sufficiency.Conversely, protection rates in Indonesiatended to increase over time, reflecting the Indonesian government's policies promoting self- suficiency in food as opposed to palm oil and rubber exports. Comparison of the three countries' total agricultural protection rates suggests that one of the reasons the Philippines lags behind Indonesia and Thailand in agricultural production and export growth is the heavy taxes the government has imposed on the agriculturalsector over time. While total protection rates in Indonesia and Thailand have declined substantially, the Philippines'agriculturalsectorstillsuffersfromsignificant negativeprotection. Akiyama and Kajisa's analyses indicate that the main characteristicof the three countries' agriculturalpricing policies is their price-stabilisingeffects, especially on food crops, in both the short and the long run. Price stabilisation was achieved over time through effective progressive taxation of border prices, apparentlyshielding both consumers and producers from abrupt price changes in the world commodity markets in the short run. In Chapter 10, Masayoshi Honma and Takehiko Hagino examine the study countries' agriculturalexport performances in the light of changes in both external and internal factors that have occurred over the past four decades. Using a model that relates the growth and pattern of agricultural exports to external market conditions, competitiveness and commodity diversification, Honma and Hagino measure each factor's quantitative impact on export growth rates. In Indonesia, commodity diversification has played the most significant role in this respect. In Thailand, competitiveness has been the most important factor. In the Philippines, the growth rate of exportsdepended most heavilyon the expansion of markets. Honma and Hagino go on to explorethe structure of demand for exports in selected markets using a formal demand system model. Agricultural exports are often considered less promising because demand for them is inelastic to price and income. At the level of trade flows, however, exports present many opportunities, and the authors find the price elasticity of demand for the three countries' major exports to be significantly large in many importing markets.They also undertake a statistical examination of the transmissibility of export prices to import markets, and find that transmission elasticity between export and import prices was significantly less than unity in 24 out of 38 trade flows. For exporting countries to ensure maximum gains from exports, international and domestic distribution systems must become more efficient and transparent. RURAL DEVELOPMENT AND AGRICULTURAL GROWTH 11 Whether agricultural exports can continue to contribute to economic gowth and rural welfare improvement depends on the sector's ability to adjust to changes in relative resourcescarcityand comparativeadvantage in the course of economic development. Thai agriculture showed it had this ability in increasingexports even during a period of rapid industrial growth that lasted four decades. In recent decades, Indonesian a riculture showed g it could also adjust to changes in external and internal factors. Philippines agriculture, however,failed to adjust the composition of its exports to shifts in comparative advantage, hence the country fell far behind Thailand and Indonesia in export levels. DETERMINANTS OF PRODUCTIVITY AND RESOURCE FLOWS IN AGRICULTURE This book's final two chapters make quantitative comparisons of the determinants of growth and the reallocation of agricultural labour in the three countries, and show how similar improvements in technology can produce different results, partly because of naturally constraining factors. Importantly,thesepapersalsoshowhowpolicycan help minimiseconstraining factors and government investments in infrastructure and education can enhance growth. Because new technologies tend to reward land and capital more than labour, they also encouragethe movement of rural labour out of agriculture, a process universally associated with economic development. In Chapter 11, Yair Mundlak, Donald F. Larson and Rita Butzer use extended time-seriesdata to measure growth and productivity in the three study countries. They find that the new technologies associated with the Green Revolution affected returns to fertilisers, irrigated land and capital so that all came to prove scarce in varying degrees. Public and private investment, in part policy-driven, complimented these technology-related changes. The authors find that factor accumulation played an important role in output growth, and that accumulations from policy-driven investments in people and public infrastructure were important sources of productivity gains. They conclude that policies that ease constraints on factor marketsand promote public investment in people and infrastructure providethe best opportunities foragriculturalgrowth, though in this respect significant differences exist among the three countries and in time for the Philippines and Thailand. Two distinct periods characterise the Philippine experience. From 1961 to 1980, theagriculturalsectorexperienced rapidgrowthdriven byimproved 12 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH technologies, increased use of fertilisers, cultivation of new areas and expansion of irrigated areas. Private investments in agriculture grew significantly,and gains were made in education. From 1980 to 1998, both public and private investment in agriculture tapered off. Sectoral gowth slowed, and was driven primarily by a growing and better-educated labour force and continuing but slower powth in fertiliser use. The Thailand experience is also characterised by two distinct periods, though during both theseperiodsagricultureexperiencedsignificantgowth. Improvementsin education, the health of the population and infrastructure were important sources of gowth,as were high levels of private investment. From 1981 onwards, the growth rates of cultivated land and agricultural labourdroppedoffconsiderably.Privatecapitalgrowth,however,accelerated, thus aggregate resources devoted to agriculture increased. Consequently, output continued to growat higher than 3 per cent per annum, even as the labour growth rate fell to 0.42 per cent. In Indonesia, up until the recent period of financial and political instability,agriculturegrewconsistentlyfor manyyears. From 1971 to 1998, the sector expanded at higher than 3 per cent per annum. Improvementsin education and infrastructure and a declining mortality rate were consistent sources of gowth. Capital accumulation was fuelled in part by large investments in tree crops. Growth in agricultural labour was a much more importantfactor in Indonesiathan inThailand and the Philippines,because average returns to labour in Indonesia were high. The authors speculate that high average returns to labour could have been due to the constraints placed by scarcity of labour on agricultural expansion in Indonesia's outer islands. In Chapter 12, Butzer, Mundlak and Larson analyse sectoral migration within each of the study countries, and confirm earlier findings that the process of change for an a riculture-based labour supply is a lengthy one g that occurs over generations. Nonetheless, differences in opportunities matter,and the authors find significant responses to income differentialsin all three countries. Moreover, in all three countries, despite increasing agricultural productivity and persistent out-migration, average incomes in non-agricultural sectors have always been substantially higher than in agriculture.The differences are greatest in Thailand, where significant out- migration did not occur prior to the 1990s. RURAL DEVELOPMENT AND AGRICULTURAL GROWTH 13 CONCLUSION The papers in this volume describe a sustained agriculturaland economic gowth, rare among developing countries, from a variety of perspectives. They show how a natural comparative advantage in wet-rice cultivation put the three study countries in good positions to benefit from the Green Revolution. In all three countries, policymakers made the most of new opportunities, and new technologies helped bring about growth in rural incomes and created new demands for resources-especially fertiliser, irrigation and other formsof capital. Initial constraints boosted the shadow prices of these scarce resources and in some cases reduced the potential for !growth.Better management and more intensive use of land brought great rewards, and supported the development of a type of smallholder land ownership more typical of Thailand and Indonesia than the Philippines. Even in the outer islands of Indonesia, where land remained relatively abundant and labour scarce, smallholders engaged in new production of traditional plantation crops such as palm oil. Implementationofthe new technologiesofthe Green Revolutionbecame an important politicalobjective in all three countries. In Indonesia, the drive towards rice self-suff~ciencybecame a nationally uni&ing goal and also a measure of political legitimacy. All three countries, however, pursued macroeconomicpoliciesthatlargelydiscriminatedagainst tradableagriculture. At the same time, governments implemented commodity-specific policies reflectingthespecial politicaleconomyassociatedwitheachcrop. Rice received special treatment in all three countries. Historically rooted institutions also facilitated adaptation, the evidence suggesting that more pluralistic governmentsconstrained ill-advised policies at the same time as local self- organisingpoliticalinstitutionshelpedfarmersovercomeconstraintson credit. While the effects of the Green Revolution were dramatic, it was the accumulation of productivity factors that determined the sustainabilityof growth. Private sector activity and public investments contributed to the build-up of physical and human capital, with the incentivesshaping private sector activitieslocated partlyin globaleventsand partlyin domesticpolicy. Importantly, the associated accumulation of wealth and private capital was not independent of the three countries' initial ecological conditions, since so many of the rewards made possible by the new technologies accrued to land owners. 14 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH An important lesson to be drawn from these studies is that sustained agriculturalgrowth is a lengthy process that occurs over generations, partly because of the slow pace at which labour can practically move from agriculture to other sectors. It follows that sustained agricultural gowth requires a sustained effort by governments to contribute to this process materially-through investmentsin people, infrastructure and policiesand institutions that assist rather than penalise rural communities. NOTE ' For adiscussionof thewaysin which this kzissez-fdireapproachactuallyproved beneficial, see Akiyama and Nishio (1997). REFERENCES Akiyama, T. and Nishio, A., 1997. 'Sulawesi's cocoa boom: lessons of smallholder dynamismand a hands-offpolicy', TheBulletin ofIndonesian Economic Studies, 33(2):97-116. World Bank, 1993. The East Asian Miracle: economic growth and public poliqy, Oxford University Press, New York. AN ECOLOGICALAND HISTORICAL PERSPECTIVE ON AGRICULTURAL DEVELOPMENT IN SOUTHEAST ASIA Yujiro Hayami Ecology and historyare no doubt the hdamental constraints on the current course of economic development. Few studies, however, have investigated the process by which different ecologicalconditionsand historical trajectories have interacted to forge different social and cultural systems, resulting in major differences in development across economies. Drawing on agricultural development experience in Southeast Asia, this paper aims to shed light on this process. Southeast Asia's relatively strong agricultural growth during the last three- four decades is generally considered one of the factors underlying the 'economic miracle' in the region (World Bank 1993). Performances within the region, however, have varied significantly.This paper attempts to explain the variations mainly in terms of differences in the agrarian structures that were created by colonial regimes living in diverse ecological environments. Southeast Asia can be classified into two major ecolo ical zones-the g continental region, including Thailand, Vietnam and Myanmar; and the insular and peninsular region (henceforth called the 'insular' re ion), g including Indonesia, Malaysia and the Philippines (Figure 2.1). The former is characterised by major river deltas and the latter by tropical rain forests, among other attributes. Before the 1860s, when new transportation technology integrated the region with the rapidly industrialising West, people in Southeast Asia had made their living from wet rice production in small valleys or shifting cultivation in upland forests. A large part of the major deltas and thick rain forests remained unused, until growing demands 16 KURAL DEVELOPMENT AND AGRICULTURAL GROWTH from the West for tropical products lead to the conversion of deltas into paddy fields for commercial rice production and rain forests into plantations for export cash crops, establishing the basis for tent-for-surplus' growth.' Deltas continued to be dominated by peasants or small family farms, while insular areas were split between peasants cultivating rice in small valleys and coastal plains, and large plantations worked by hired labour. These differences in agrarian structure were rooted in significantly different ecological conditions. Equally significant were differences in land policies across political regimes. For example, the distribution of land ownership in the Philippines under Spanish colonialism became far more skewed than in Indonesia under Dutch colonialism, despite the fact that both these countries were located in the insular region. The &-reaching influenceof thesedifferencesin agrarian structure, formed along different historical trajectories under different ecological conditions, on agricultural development across Southeast Asia is the focus of this paper. While variables other than the agrarian structure, for example government Figure2.1 Map of Southeast Asia, indicatingthe continental and insular regions AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 17 policies relating to rural development, may also be important, the possibility that present policy choices are significantly influenced by the structure of the political economy, of which the agrarian structure forms a major component, must not be overlooked. The first section of this paper provides an overview of the resource characteristics, agrarian structures, growth in aggregate agricultural output, and major exports of the three economies. The second section reviews the process of vent-for-surplus development in Southeast Asia during the late nineteenth and early twentieth centuries, and examines the critical roles of major river deltas in the continental part of Southeast Asia and tropical rain forests in the insular part, along with the related trade patterns. The third section investigates the evolution of different agrarian structures in the three economies, and argues that the pre-emption of uncultivated but cultivable land by the power elite was the major force that skewed land distribution. The fourth section explains differences in agricultural growth rates across the three economies in terms of their different agrarian structures. The final section discusses government policy choices in relation to the formation of the three countries' political economies.In concluding,it argues that the approach taken in this paper may be applicable to comparisons not just within Southeast Asia but across regions, such as Africa and Asia, to provide insights into broader development issues. RECENT DEVELOPMENTS This section provides an overview of the resource characteristics, agrarian structures and agricultural production growth of Indonesia, the Philippines and Thailand over the past three decade^.^ Table 2.1 compares area of agricultural land in relation to population and labour force. Cropland includes both arable (annual) and perennial crops, as defined by the Food and Agriculture Organization of the United Nations (FAO). Arable land includes both lowland paddy and upland annual crops. Data on lowland paddy, based mainly on aerial photography, only recently became available, and has begun to appear in national statistical report^.^ The area of upland annual cropland is the difference between arable land and lowland paddy areas. In 1996, per capita cropland area in Thailand was more than twice as large as in Indonesia and the Philippines; per farm worker, however, it was only marginally larger. From 1965 to 1996, cropland area increased by about 20 per cent in Indonesia, 40 per cent in the Philippines and 60 per w 00 Table 2.1 Land endowments for agriculturalproduction in Indonesia,the Philippines and Thailand,1965-96 Indonesia Philippines Thailand - - 1965 1996 139611965 1965 1996 199611965 1965 1996 199611965 ratio ratio ratio No. of farm workers ('OOO)a 29,006 47,713 1.64 7,363 12,128 1.65 12,450 20,824 1.67 Croplandb S F Total ('000 ha) 26,000 30,987 1.19 6,660 9,520 1.42 12,600 20,445 1.62 Per capita (ha) 0.24 0.15 0.63 0.21 0.14 0.67 0.41 0.35 0.85 2 Per farm worker (ha) 0.90 0.65 0.72 0.90 0.78 0.87 1.01 0.98 0.97 5 Pecentage of cropland 3 Lowland paddy' n.a. 27 n.a. n.a. 32 n.a. n.a. 53 n.a. 5u Annual upland cropsd n.a. 31 n.a. n.a. 22 n.a. n.a. 31 n.a. 8 Permanent crops 31 42 1.35 38 46 1.21 11 16 1.45 z [ 5 ~conomicallyactive population Arable (annual) land area plus area under permanent crops 'Lowland paddy areas are for 1995 in Indonesia, 1991 in the Philippines, and 1993 in Thailand Arable land area minus lowland paddy area Source: Food and Agriculture Organization of the United Nations, n.d. FAOSTAT, FA0 online database, http:llapps.fao.org/default.htm;except for G, 8 lowland paddy areas, which are taken from IndonesianStatisticalEarbook 1996(8,484,000ha in 1995),PhilippineStatisticalEarbook1998 (3,OO1,000 ha in 1991) andA~culturalStatistics1995/96for Thailand (10,934,000ha in 1993). z AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 19 cent in Thailand. In all three countries, however, the expansion rate of cropland was lower than the growth rates of both the population and the agricultural labour force. Thus, fiom 1965 to 1996 cropland area relative to population in Thailand decreased by 15 per cent; in Indonesia and the Philippines it decreased by more than 30 per cent. Meanwhile, cropland area relativeto agriculturallabour force remained about the same in Thailand, but decreased by about 30 per cent in Indonesia and about 10 per cent in the Philippines. These data suggest that until recently Thailand has been endowed with better conditions for expanding cropland than Indonesia and the Philippines. The important characteristic distinguishing Thailand from Indonesia and the Philippines is the high share of paddy in total cropland area- more than 50 per cent in Thailand compared to only about 30 per cent in Indonesia and the Philippines in 1996. On the other hand, more than 40 per cent of cropland in Indonesia and the Philippines is under permanent crops, compared to less than 20 per cent in Thailand. Permanent crops in these economies include tropical cash crops such as coffee, coconuts and rubber. The comparative share of tree crops in total cropland remains the same today as 30 years ag-much higher in Indonesia and the Philippines than in Thailand. These data reflect ecological differencesbetween the continental part of Southeast Asia, represented by Thailand, and the insular part, represented by Indonesia and the Philippines.That is, the continental region has major river deltas that are used almost exclusively for paddy, while the insular region was originally covered by tropical rain forests that could profitably be converted to plantations for tropical cash crops. The different types of agricultural production that arose in response to different environmental conditions in turn gave rise to different agrarian structures in the two regions. Like cereal production in other parts of the world, rice production in Thailand is dominated by peasants or small h i l y farms. In contrast, tropical cash crops are mainly grown on plantations or large estate farms worked by hired labour, though many peasants also grow cash crops. Table 2.2 compares distribution of operational farm size and incidence of tenancy across the three countries for the period 1973-78, before the influence of Philippine land reform became significant. In all three economies, farms smaller than 5 hectares in area accounted for between 70 and nearly 100 per cent of all farms and cultivated 4&70 20 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH per cent of all farmland. Farms larger than 50 hectares in area, considered to be agribusiness plantations, were few in number but accounted for 14 per cent of all farmland in Indonesia and the Philippines; in Thailand, however, they accounted for less than 1 per cent of all farmland. It is clear then that the agricultural sector in the insular region was divided between peasants subsisting on small parcels of land and large plantations with hired labour, while in the continental region it was characterised by a unimodal distribution of self-employed family farmers. In the Philippines, plantations were privately owned and managed, while in Indonesia they were mostly state enterprises expropriated from Dutch planters after independence. Incidence of tenancy also varied widely, being distinctly higher in the Philippines than in the other two countries, especially in terms of percentage area under pure tenancy. Table 2.2 Distribution of operationalfarm sizeand incidenceof agriculturaltenancy in Indonesia, the Philippinesand Thailand, 1973-78 Indonesia Philippines Thailand Year of survey Average operational farm size (ha) Percentageof farms and farmland < 5 ha Farms 98 85 72 Land area 69 48 39 > 5 ha Farms 0" 0.2 0" Land area 14 14 0.9 Gini coefficient of land concentration 0.56 0.51 0.45 Percentage of tenanted area in total farmland Pure tenancy 2 21 6 Totalb 24 33 16 Percentage of share tenancy in tenanted land 60 79 29 aLess than 0.05 per cent Area in pure tenancyfarms plus area in ownerltenant farms Source: Hayami and Otsuka, Keijiro,1993. The Economics of Contract Choice,St Martin's Press, New York:10-11 (based mainly on FAO, 1971. 1970 WorBCensusofAp'mlture: analysisand comparison of the results,Food and Agriculture Organizationof the United Nations, Rome). AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 21 Table 2.3 compares agricultural growth in the three economies from 1965 to 1995. Growth rates for total agricultural output were about the same for Indonesia and Thailand, but in Indonesia the per capita and per worker growth rates were somewhat higher. The growth rates for all three measureswere lowest in the Philippines. In production per hectare,Thailand's growth was much slower than Indonesia's and comparable to the Philippines'. Thailand's slow growth in land productivity was partially a result of a major expansion of the cultivation frontier in the northeast, a region characterised by poor soil and unstable rainfall, and partially due to the relatively low diffusion rate of HYV rice.* In terms of both environmental conditions and resource availability, Thailand's traditional comparative advantage was in rice production, while that of Indonesia and the Philippines was in tropical cash crops. It should not be surprising then that Thailand has long been the world's largest rice exporter, with its market share rising continually from 1961-65 to 1991-95 Table 2.3 Growth of agriculturalproductionin Indonesia, the Philippines and Thailand, 1961-95 Growth rate (per cent per year) Index 1976180 1991195 Indonesia Total Per capita Per farm worker Per haa Philippines Total Per capita Per farm worker Per haa Thailand Total Per capita Per farm worker Per haa " Per hectare of cropland (arable land plus land undel permanent crop) Source: Food and Agriculture Organization of the United Nations, n.d. FAOSTAT, FA0 online database, http:1/apps.fao.orgldefault.ht1n. 22 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH (Table 2.4). In contrast, Indonesia and the Philippines have{remainednet rice importers, though their import margins were significantly reduced by the Green Revolution. The effect of the Green Revolution was especially positive in Indonesia, where it made possible a high rate of growth in aggregate agricultural output despite relatively slow growth in cropland area (Tables 2.1 and 2.3). Thailand's surprising rise as an exporter of certain tropical cash crops was associated with declines in exports from Indonesia and, more conspicuously, the Philippines. Sugar is a typical example. Before the Second World War Thailand was a net importer of sugar and in the early 1960s was barely self- sufficient. By the 1990s, however, it had become the third-largest exporter of sugar in the world, after Brazil and Australia. Meanwhile, Indonesia and the Philippines, two traditional Asian exporters of sugar, almost completely lost their significance in the international market.Thailand had also exceeded Indonesia in rubber exports and the Philippines in exports of pineapple products by the 1990s. Indonesia, however, achieved a major increase in its world market share of coffee, and maintained high shares of palm oil and rubber exports. Meanwhile, the Philippines was the loser in most of the tropical cash crops in which its traditional comparative advantage was supposed to lie, especially from the 1970s onwards. The strengthening of Thai agriculture's competitive position and weakening of the position of Philippine agriculture are clearly revealed in market share data (though comparative cost data are not available). In the East Asian economic miracle that began in the 1960s and ended in the financial crisis of 1997, Indonesia and Thailandwere 'high-performing economies' (World Bank 1993). The Philippine economy, however, stagered, especially in the 'lost decade' of the 1980s. Thus, per capita GNP in Thailand, about the same as that in the Philippines in the 1970s, had doubled by the early 1990s. At the same time, Indonesia's per capita GNP increased from about one-half that of the Philippines to about the same level. It would be reasonable to expect that the differences among the agricultural performancesof the three economies,as summarised here, would be one of the significant factors underlying the differences in their growth rates. The following sections aim to identify ecological and historical determinants of the agrarian structure unique to each of the three countries, and discuss how their different agrarian structures might be related to their different agricultural and economic growth rates. AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 23 Table 2.4 Netexportsasa percentageof totalworld exportsof selectedagriculturalcommoditiesfromIndonesia, the Philippines and Thailand,'1961-95 Share of world market (percent) 1961165 1976180 1991195 Rice Indonesia Philippines Thailand Maize Indonesia Philippines Thailand Sugara Indonesia Philippines Thailand Coffeeb Indonesia Philippines Thailand Coconut Oil Indonesia Philippines Thailand Palm Oil Pndonesia Philippines Thailand Rubber Pndonesia Philippines Thailand Pineapplec Indonesia Philippines Thailand Banana Indonesia Philippines araw sugar equivalent green and roasted coffee canned pineapple Note: Negative numbers represent imports. F n r ~ r r ~ . EnnA I n A Arrr;rlllnlrp nmran;.r~+;nnnfr h p TTn;+pA kTqt;nnc nA E A n C T A T P A n 24 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH The basis of vent-for-surplus development A basic element in this analysis is the classification of Southeast Asia into continental and insular regions, the former characterised by major river ' deltas and the latter by tropical rain forests. Of course, such characterisation is subject to the hazard of oversimplification, since it disregards wide ecological variations within each regi~n.~ In Thailand, for example, the major delta of the Chao Phraya River encompasses only a part of the Central Plain, one of the country's four regions. The North is characterised by small river valleys amidst hills and mountains rising towards the Myanmar-China border, where irrigated rice farming can easily be practiced by tapping small streams, a system on which early Thai dynasties were built. The Northeast, which borders Laos, is characterised by undulating plateaux with sporadic rainfill and poor soil. It had remained relatively unsettled until recently, when Lao migrants settled in the region to take up risky rain-fed farming. The South, which borders Malaysia, was originally covered by rainforest has an environment similar to the insular region originally covered by rainforest. Ecological variations within the insular region are equallywide. Especially pronounced is the difference in Indonesia between Java (and Bali) and Outer Islands such as Kalimantan and Sumatra. While much of the land in the Outer Islands is typically covered by tropical rain forest, the environment of Java and Bali is characterised by volcanic slopes with fertile soil and a steady water supply, which make much of these islands suitable for irrigated rice farming. The environment of the Philippines is largely similar to that of Indonesia's Outer Islands but is mixed with volcanic terrain similar to that of Java. Despite these wide variations within the regions, in general it can be said that the continental region is characterised by major river deltas and the insular region by tropical rain forests. Despite the differences between them, both the continental and the insular regions formed the basis for economic development in Southeast Asia from the late nineteenth century to the early twentieth century. When expanded trade opportunities opened, it was the deltas of major rivers, such as Chao Phraya in Thailand, Irrawaddy in Myanmar and Mekong in Vietnam, that became the basis for vent-for- surplus development in continental Southeast Asia. In the insular region, on the other hand, it was rain forests that provided the basis for economic development. AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 25 Because the major river deltas in the continental region are very flat and low relative to sea level, they are almost completely submerged by floods during the rainy season but unable to store water in the dry season. Until the mid nineteenth century, floodplains in the major deltas defied human settlement, literally remaining empty land. Through major civil engineering works aimed at controllin flooding, the deltas were transformed into g habitable and agriculturally ~roductiveland. In Thailand, water control was achieved through the development of a network of canals connected to the Chao Phraya River. These canals guide floodwater more evenly over wider areas, facilitatin rice production, while their banks offer farmers g usually flood-free living areas. Canal constructionin the Chao Phraya deltawas begun by the government of the enlightened King Mongkut (Rama IVof the Chakri dynasty), shortly after the 1855 signing of the so-called BowringTreaty with Britain, which opened the kingdom to trade with the West. Soon, increased foreign demand for Thai rice significantly raised both the price of rice and the value of rice- producing land, and attracted private investment. In 1889, the Siam Canals Land and Irrigation Company, established by a group of influential courtiers and wealthy Chinese traders, secured a concession to dig canals in a vast tractofswampyland northeastof Bangkok,alongwith rights toland reclaimed alongside the canals. The company's operation was managed by the Chinese business klite and its consmaion work was carried out by Chinese migrant labourers who worked for wages, unlike the corvkelabour used in the King's previous projects. The farmers who settled on reclaimed land as tenants, however, were Thais who had migrated from other regions.' There is little doubt that the opening of the Chao Phraya delta to rice production established the basis of the vent-for-surplus growth of the Thai economy, which in the late nineteenth century specialised in rice production. In the period 1905-09, the area planted to rice in the Central Plain was as large as 6.8 million or 85 per cent of the kingdom's total rice-producin area. g Althoughcomparable data for earlier years are not available, this was certainly larger than the 1850 total rice-producing area of 5.8 million rai (Ingram 1971). The opening of the Irrawaddy and Mekong deltas to cultivation was no less important in the development of the Myanmar and Vietnamese economies in the same period. A comparable role in the vent-for-surplus development of the insular part of Southeast Asia was played by tropical rain forests. Long before the 26 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH mid nineteenth century, rain forests had supplied valuable products for trade, such as cinnamon, cloves, birds nests, deer horn, and hides. The high incidence of malaria and other tropical diseases, however, prevented human settlement inside thick tropical forests at low elevation. Typically, native people lived on sea coasts and occasionally entered the forests to collect and extract natural products for sale to foreign traders or their agents who sailed the coasts. It wasn't until the late nineteenth century that Western capital and entrepreneurship began to convert the forests to plantations of tropical export crops, relying heavily on migrant workers from China for labour. Prior to this period, Western colonial powers had attempted to collect tropical products from the natives by tax and other coercive means; it was largely after the mid nineteenth century, however, that Western colonisers themselves began to produce export crops by establishing plantations in Southeast Asia. Greater integration of Southeast Asia into the world economy in the late nineteenth century led to concurrent exploitation of continental deltas and insular forests. Even greater integration occurred when a free trade regime was established under the hegemony of Britain and ocean transport was revolutionised. Britain forcefully imposed the free trade system on native economies,either directly in its colonies or indirectly by forcing liberalisation on local sovereign and other colonial pavers. Under the Bowring Treaty, Thailand conceded to Britain not only exterritoriality but also lost financial autonomy. Export and import duties were fixed at the flat rate of 3 per cent ad valorem, and internal taxes such as excise taxes, transportation tolls and even land taxes were not allowed to change by will of the kingdom alone. Advisors from Britain caremy monitored public finance. For several decades afier signing the Bowring Treaty, the Kingdom of Siam was effectively something like a British protectorate. The free trade system, both internationally and internally, was imposed on the kingdom in a way similar to that in which it was imposed on British colonies like Burma and Malaya. Britain also ~ressedother Western colonies to adopt the free trade system. For example, its occupation of Manila from 1762 to 1764 during the Seven Years' War broke Spain's monopoly in the re-export of Chinese goods from Manila to Mexico. Manila was then opened up to other nations for transportation and commerce, and Britain continued to apply pressure on the Philippines to open up its ports until the mid nineteenth century (Larkin 1972). AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 27 The reason behind Britain's strong drive for free trade was its high manufacturing capacity, which established it as the 'workshop of the world' after the Industrial Revolution. During this period, British industries sought both markets for their products and sources of raw materials. Having established a modern fictory system that could produce industrial products at lower costs than local cottage industries in the tropics, Britain found it advantageous to trade their manufactured commodities for tropical agricultural products and raw materials. This trade met import demands and replaced the forced collection of tropical commodities through taxes and other means, common under earlier colonial regimes such as those of the Spanish conquistadors in the Philippines and the Dutch East India Company in Indonesia. The British approach soon came to be adopted by other Western nations as they followed Britain in industrialisation. As the West's industrial production capacity expanded, its demand for raw materials such as cotton, rubber and tin also increased. Moreover,as incomes and wages in the West rose, tropical delicacies such as pepper, coffee and tea, hitherto available only to the high-income elite, became common on the tables of ordinaryworking people. Thus, demand in the West for primary products from the tropics in general became incomparably larger. This tendencywas strengthened by major developments in ocean transport, that is, the introduction of the steamship and the opening of the Suez Canal in 1869. These two milestones made transportation costs from Bangkok to major ports in Europe lower than those from Bangkok to Thailand's old capital, Chiang Mai (Ingram1971). If such developments had not occurred, it would have been impossible for bulky SoutheastAsian commodities such as rice to find markets as far away as Europe (Furnivall 1948). While advances in ocean transportation reduced the prices of Southeast Asian commodities in the West, they also reduced the prices of Western commodities in Southeast Asia. Thus, industrial commodities from the West flowed into Southeast Asia, out-competing local handicraft industries. De-industrialisation became common in the region (Resnick 1970). For example, Thailand, an exporter of cotton products prior to the 1850s,soon becamea major importer (Ingram1971).A correspondingshift of indigenous labour from manufacturing to primary production for export occurred. Together with labour migration from China and India, this shift facilitated the exploitation of previously unused natural resources, thus drivin vent- g for-surplus development. 28 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH The impact of international trade on specialisationin primary production can be seen in the development of sugar production in Negros, the Philippines. Prior to the opening of nearby Iloilo City as an international port in 1855, Negros Island was sparsely populated and largely uncultivated. From 1855 onwards, land on the island was rapidly developed for sugar plantations. At the same time, local weaving industries that made textiles for export in the environs of Iloilo were devastated by imports of cheap British cloth (McCoy 1982). Under the global trade system created in the late nineteenth and early twentieth centuries, exchange took place not simply between industrial commodities in the West and primary commodities in Southeast Asia. Rice produced in the continental region was originally brought to Europe as cheap food for industriallabourers (with some re-exported to Latin America). Later, however, as plantations developed in the insular region, demand for rice as the basic subsistence food of plantation labourers expanded so fast that it could no longer be met by local suppliers. Correspondingly, the share of rice exported from the continental region to the insular region within Southeast Asia increased.Thus, the trade flows that emerged during this period were triangular: rice produced in the continental region was exported to the insular region, while tropical cash crops produced in the insular region by labourers fed on rice from the continental region were exported to Europe in exchange for industrial products. In this triangular trade flow, comparative advantage dictated regional specialisation. For example, when trade opened, sugar appeared to be a promising export industry in Thailand, however it was soon destroyed by imports from Indonesia and the Philippines (Ingram 1971). Thus, vent-for-surplus development in Southeast Asia based on the exploitation of hitherto unused land was reinforced by comparative advantage within the region, which was largely determined by ecological conditions.' EVOLUTION OF AGRARIAN SYSTEMS How the process of vent-for-surplus development influenced the formation of agrarian structures in Indonesia, the Philippines and Thailand (Table 2.2) may be surnmarised as follows. Thailand is characterised by a unimodal distribution of peasants or family farms, while large estate farms or plantations are insignificant and the incidence of tenancy is relatively low. AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 29 Indonesia and the Philippines are split between the peasant sector, which grows mainly subsistence crops, and the plantation sector, which grows tropical cash crops. In contrast to the other two countries, the incidence of tenancy in the Philippines is high. The share of the landless population in the rural sector is highest in the Philippines. The persistence of the peasant mode of production in the face of the emergence of the plantation system is commonly explained by the differing technological production requirements of subsistence food crops and export cash crops. Some would argue, however, that different firm sizes and the problem of landlessness in Southeast Asia (as well as in other parts in the world) stem from the theft of land by colonial and domestic 6lites rather than the technological factors of agricultural production. Conditions of the plantation system9 Conventional explanations for the emergence of the plantation system refer to the economies of scale inherent in the production of tropical export crops (Baldwin 1956). However, only a small number of crops are subject to sufficientlystrong economies of scale at the farm level to make a plantation necessary (Pim 1946; Wickizer 1951, 1960; Lim 1968; Hayami et al. 1990).1°In hct, examples of the successfd cultivation by peasants of every so-called plantation crop can be found at least somewhere in the world. Significant returns emerge only at the levels of processing and marketing. For a large-scale central processing and/or marketing system to be supplied with raw materials according to schedule, it must be vertically integrated with a large estate farm. A typical example is fermented black tea. The manufacturing of black tea at a standardised quality for export requires modern machinery into which fresh leaves must be fed within a few hours of being picked (Wickizer 1951, 1960). The need for close coordination between farm production and processing underlies the traditional use of the plantation system for black tea manufacture. In contrast, unfermented green tea remains predominantly the product of Chinese and Japanese peasants." In the case of bananas for export, harvested fnrits must be packed, sent to the wharf and loaded onto a refrigerated boat within a day. A boatload of bananas that can meet the quality standards of foreign buyers must be 30 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH collected within a few days. The production process from planting to harvesting must therefore be precisely controlled so as to meet shipment schedules. Although the plantation system has a decisive advantage over peasants in producing bananasfor export, bananas for domestic consumption are usually produced by peasants. For crops that do not require centralised processing and marketing, on the other hand, plantations have no significant advantage over peasants. Typical examples are cocoa and coconuts. The fermentation of cocoa and the drylng and smoking of coconut meat to make copra can be done in small lots with no large capital requirement beyond small indigenous tools and facilities. Thus, these two crops continue to be grown predominantly by peasants. Sugar is frequentlycited as a classiccase of an economy of scale stemming from the need to coordinate between farm production and lar e-scale central g processing (Binswanger and Rozenweig 1986). Efficient operation of a centrifugal sugar mill requires a steady supply of a large amount of cane. Coordinationis required from planting to harvesting to processing; however, this coordination need not be as stringent as it is for tea and bananas. If cane processing is delayed, then the rate of sugar extraction decreases; the loss incurred, however, is in no way comparable to the devastating damage that delayed processing can cause to the quality of tea and bananas for export. Because sugarcane can be transported relatively long distances and stored for several days, the need for vertical integration is not as great, and the necessary coordination can be achieved through contracts with cane growers on the time and quota of cane delivery. In fact, efficient sugar industries with smallholders have developed in Australia, Taiwan and, more recently, Thailand. Another explanation for the use of the plantation system is the advantage large estates farms have in access to capital. This access, it has been argued, gives plantations an advantage in tree crops that can take years to produce a saleable crop after planting (Binswanger and Rosenzweig 1986). However, the opportunity costs of labour and capital incurred by peasants in the production of tree crops are not necessarily high, because peasants typically plant trees on unused land. If this land is located near their home, they open new land for planting using family labour at low opportunity cost during the idle season for crops on land already in use. When peasants migrate to frontier areas, they typically slash and burn jungles and plant subsistence crops such as maize, potatoes and upland rice, together with tree seedlin s. Under the plantation system such complex intercropping, g AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 31 involvin the monitoring of labourers over a wide and ecologically varied g area, is very difficult to manage (Brewster 1950; Binswanger and Rosenzweig 1986; Hayami and Otsuka 1993). Even during the colonial export boom of tropical cash crops in the nineteenth and early twentieth centuries, the plantation system failed to make inroads in regions where the indigenous population had established family farms (Lewis 1970). Western traders found it more profitable to purchase tropical agricultural commodities from peasant producers in exchan e for imported manufactured commodities than to produce the g tropical crops themselves on plantations. The establishment of plantations in less developed economies became a necessity where regions physically suited to produce the tropical products for which demand in industrialised nations was rising had no significant peasant population that could produce them and trade in them. The opening of frontiers to produce new crops entailed high capital outlays: virgin land had to be cleared and developed, and physical infrastructure such as roads, irrigation systems, bridges and docking hchties constructed; capital in the form of machinery and equipment had to be imported and redesigned to suit local conditions; and labourers imported from more populous regions had to be trained. Plantations thus required huge initial capital investment. For investors to internalise gains from infrastructure investments, farms had to be large. From this perspective, the plantation system can be seen to have emer ed g not because it was more efficient than the peasant mode of production, but because it was the mode of agricultural organisation that could most effectively exploit sparsely populated virgin areas, typically in the process of vent-for-surplusdevelopment. From this perspective, it is easy to understand why the same crop tends to be grown by peasants in one place and plantations in another. For sugarcane production in the Philippines, for example, the peasant mode is more common in the older settled areas of Luzon, while the plantation system predominates in the newly opened Negros (Hayami et al. 1990). Usually, the peasants' share of cash crop exports rises as the initial land-opening stage continues and infrastructure is established with increased population density (Booth 1988). While the economic advantage of the plantation system at the vent-for- surplus stage is clear, plantations could not have been established if they had not been granted concessions to large tracts of virgin land for their exclusive use. Typically,such concessions were given by colonial governments 32 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH :o Western planters. For example, in response to a sharp rise in demand for tropical cash crops in the late nineteenth century, the Dutch colonial government, which had traditionally replated against land purchases by foreigners, implemented the Agricultural Land Law of 1870, granting Dutch planters long-term contracts to lease land owned by the government but used by native tribes. While this new institutional arrangement should have accelerated the development of 'empty land' for cash crop production, it served as an instrument to pre-empt land for the tlite and close smallholders' access to land. Similar public land-leasing arrangements made under the American colonial administration on frontier land in the Philippines, especially in Mindanao, created the conditions for the establishment of large plantations managed by multinational corporations (Hayami et al. 1990). Land pre-emption and tenancy The incidence of land tenancy is closely related to pre-emption of land, but land tenancy relationships can emerge among peasants in the absence of pre-emption. If a rural community isn't disturbed by external forces, one would expect its land tenure institutions to evolve gradually from communal to private ownership. Increased relative scarcity of land under mounting population pressure necessitates intensified land use, typically through shifiingfrom long furrows to short furrows, to annual cropping, and fUrther to multiple croppingeach year using irrigation (Boserup 1965). This process of a ricultural intensification requires major investments in land g improvement, from the removal of stones and roots from newly opened land, to land levelling and terracing, and irrigation and drainage. To encourage such investment, land users must be granted exclusive land-use rights. Thus, land tenure institutions normally evolve from communal ownership to private ownership by steps, from the periodic re-allotment of communal land among community members to the granting of life-long usufruct ri hts, inheritable usufruct rights and private property rights g amenable to market transactions. Because individual land tenure can bring about increased production efficiency by improving the relationship between land and labour (including entrepreneurship) as it becomes longer and more exclusive, land tenancy arrangements tend to become institutionalised over time. When a farmer AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 33 finds his family lacking the labour needed to cultivate a land parcel on which a long-term usufruct is established (becauseof illness or other reasons), he may rent a part of it to a farmer with an excess of labour relative to land. It is a Pareto improvement if the latter pays the former a rent equivalent to the mar inal productivity of the land. At the same time, land tenancy g associatedwith private property rights can work as an institution to increase inequality in income distribution and social hierarchy within a community. An entrepreneurial farmer may rent more land, increase his income and eventually buy the land. As he accumulates more land than his family labour can efficiently cultivate, he may rent out part of his land to someone who has become landless for whatever reason. The increased income from his rent revenue added to his farm income may motivate him to purchase more land to rent. This process could progress at a n increasingly fast rate as the relative scarcity of land rises under mounting population pressure. Indonesia Usually, however, the autonomous evolution of land property rights and tenancy relationships does not result in large-scale absentee landlordism as observed in several developing economies. Rather, it tends to stratify the peasantry along a continuousspectrum between landlord/ownerand owner1 tenant farmer.Although land tenancy is common, a majority of land remains under owner cultivation,and both non-cultivating landlordsand pure tenants are still a minority. This type of agrarian structure is typically found in the peasant sector in Indonesia. Unlike other colonial powers, the Dutch tried not to impose Western institutions such as private property rights, but rather to preserve and even strengthen traditional community institutions and organisations. The Agrarian Law of 1870 granted long-term leases of virgin public land to foreign planters, but did not allow them to purchase or rent cultivated land from individual peasants. Instead, sugar planters were allowed to lease rice land through contracts with village heads, normall y extending for fewer than 20 years. The lessee was allowed to occupy only one-third of the village land, which had to be rotated over three crop seasons. This rotation was designed to prevent planters from gaining a permanent hold on village land. Periodic re-allocation of village land under the direction of village headmen strengthened traditional tendencies toward communal landholding (Pelzer 1945). 34 RURAL DEVELOPME,NT AND AGRICUUrURAL GROWTH Philippines The Philippines presents a sharp contrast to Indonesia. From the time of conquest, theSpaniardsintroducedthe notion oflegaltitle toland (McLennan 1969). In the Philippines they applied the same principle they applied in other new territories: all land except that officially proved to be private or communal property belonged to the Spanish crown. The crown's property rights were established over vast areas of uncultivated land, including areas used as commons by native people. Much of the royal domain was granted to conquistadorsand monastic orders such as the Augustinianand Franciscan friars. This institutional development in the early Spanish era represented a wholesale pre-emption of usable land, closing access to native people. Later, as the population increased and foreign demand for Philippine products increased with trade liberalisation, large landholdings created from earlier royal grants became the basis of plantations in the uplands and rice haciendas manned by tenants in the lowlands. Native peasants, however, could gain no access to land ownership. For example, when the inner part of Central Luzon, previously covered by jungle and used only for cattle ranching, was finally converted to large rice haciendas in the late nineteenth century, many peasants migrated from the north in the belief that they had settled in no man's land. Having opened the jungle, they were told by the agents of landowners to pay rent as tenants on haciendas (Hesters and Mabun 1924). Pervasive landlordism in the Philippines is also rooted in relatively free land transactionscarried out under the Spanish regime. Chinese and Chinese mestizos who engaged in internal trade along littorals where native peasants held traditional land rights acquired land through money lending using land as collateral. A common arrangement had the borrower continuing to cultivate his land as sharecropper to his creditor during the loan period. If the borrower was unable to repay the loan at the end of the loan period, the land title shifted to the creditor and the borrower usually continued his sharecropping (McLennan 1969). The scale of landholding accumulated through this process in the coastal area was typically much smaller than that of haciendas in the inner part of Central Luzon (Haymi and Kikuchi 1981). Thus, before the Marcos land reform of the 1970s, most rice- producing land in the Philippines was cultivated by share tenants, who typically owned no land of their own. Pervasive landlordism in the rice sector and ~lantationsin the cash crop sector characteristicof the traditional AN ECOLOGICAIJ AND HISTORICAL PERSPECTIVE 35 indi enous agrarian structure were both rooted in the pre-emption of land g that occurred during the Spanish period. Thailand Pre-emption also occurred in Thailand at the vent-for-surplus stage, when land concessions were ganted to private canal builders in the Chao Phraya delta. As a result, there is a significant incidence of tenancy in the Central Plain, especially in the Rangsit area northeast of Bangkok, where the private company carried out intensive canal construction. Yet, tenancy in Thailand as a whole is of minor importance in comparison to tenancy in Indonesia and the Philippines, partly because of relatively abundant land but more importantly because of government ~olicies.According to ancient Thai custom, every man had the right to take as much land from the state as he and his family could cultivate, normally 25 rai (equivalent to 4 hectares). This institution was maintained even after the opening of trade with the West. The Consolidated Land Act of 1908 did not specifjr an exact area, but in practice the area ranged between 20 and 50 rai. The Land Act of 1936 specified 50 mi as the maximum a farmer could take. These laws kept land access open to ordinaryThais."This situation was very diEerent from that in the Philippines. Both the 1908 and 1936 Thai laws incorporated another old custom, which granted a cultivator title to land only after he had cultivated it for three years. This clause, along with land taxation applied to both cultivated and uncultivated land, discouraged the holding of idle land for speculation (Ingram 1971). The major differences in land policies between the Philippines and Thailandstemmed from the digerentcultures or value systems of the Spanish colonial rulers and the rulers of the independent kingdom. The reason for the Dutch colonial rulers' attempts to preserve traditional village institutions in Indonesia might have been their understanding that they needed to maintain social stability in order to be able to extract tropical agricultural products from the colony at minimal administrative cost, as argued by Furnivall (1944, 1948). It is also important to note that pre-emption of rice land through canal construction in Thailand foreshadowed the emergence of large-scale landlordism, but not the formation of plantations, as in the Philippines. The large holdings of landlords were usually subdivided into small parcels for rice cultivation by the family labour of landless peasants under tenancy 36 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH contracts. The owners of large tracts of rice land who had established their titles through land pre-emption, for example by obtainin canal-digging g concessions in the Chao Phraya delta, preferred tenancy to plantation operations. One reason for this preference may have been the daculty of task standardisation and hence of worker supervision in rice production. Another reason could have been the fact that paddy, unlike black tea and bananas for export, is storable, making close coordination between firm production and processinglmarketing unnecessary. Although rice milling and marketingfor export involvesignificant economies of scale, the operators of this business could secure an adequate supply of paddy through ordinary market transactions. As a result, there was no need to vertically integrate farm production with processing and marketing using a plantation or contract farming system. It seems reasonable to postulate that if the nature of rice milling technology were such that it re uired close coordination with paddy q production, large rice plantationswould have been establishedin the Rangsit area, where territorial concessions were granted to private canal builders. Outside the newly opened delta area, in the previously settled North region, the practice of tenancy is fairly common. The agrarian structure in Thailand's North, which did not experience pre-emption, is like the peasant sector in Indonesia in that it is characterised by a continuous spectrum from landlordlowner to ownerltenant firmers. AGRARIAN STRUCTURE AND AGRICULTURAL GROWTH RATE There is evidence to suggest that differences among the three economies' agricultural growth rates in recent decades can be explained, at least in part, by the differences among their agrarian structures, which emerged along different historical paths under different ecological conditions. This section addresses two questions. Why have Indonesia and the Philippines, which both had a strong comparative advantage in tropical cash crops such as sugar before the Second World War, lost ground to Thailand in world market share in more recent years (Table 2.4)? Why was the agricultural growth of the Philippines so much slower than Indonesia and Thailand's (Table 2.3)? AN ECOLOGICAL AND HISTORICAL PERSPECTIVE Losing ground to plantations We have argued that the efficiency of the plantation relative to that of the peasant system during the early stages of land-abundant and labour-scarce economies is high. After land opening, however, as tropical economies shift from the land-abundant to the land-scarce stage, several negative aspects of plantations become significant. Capital for labour. Because ecologically diversefarm operations are inherently difficult and supervision of wage labourers working over a wide area is problematic, and because plantations have relatively easy access to both private credit markets and government concessional loans, the plantation system tends to substitute capital for labour. However, in developing economies characterised by a high number of labourers relative to capital, this substitution can be socially inefficient. Less intensive cultivation. In the plantation system, which mainly employs wage labour and usually practices monoculture, agricultural land tends to be cultivated less intensively. Complicated intercropping and crop-livestock combinations are more difficult to manage in the command system, thus both labour input and income per hectare on plantations tend to be lower.l3 This is a source of inefficiency, since it means that with population growth land becomes scarce relative to labour. In contrast, small family farms tend to cultivate land more intensively. Crop specialisation. Plantations tend to specialise in a single crop. This tendency reduces their capacity to respond to changing demand by shifting to other crops. Moreover, continual monocropping tends to degrade soil and increase pests. Fertiliser and chemicals put serious stress on the environment and on human health, and they are expensive. Task specialisation. Specialisation of plantation workers in specific tasks inhibits the development of their managerial and entrepreneurial capacity (Baldwin 1956; Myint 1956; Beckford 1972). Class codict. The plantation system is a source of class conflict between labourers and managerslcapitalists.The presence of a plantation enclave in rural economies where the peasant mode of production predominates has often strained relations in rural communities. From the standpoint of social stability, the plantation system is no match for the system of relatively homogeneous small-scale producers owning small assets, however small they might be. 38 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Although Southeast Asia was traditionally endowed with relatively abundant land ready for resource exploitation, the ex losive po ulation p p growth experienced by developing economies after the Second World War led to the progressive closure of frontiers for opening new land. It seems reasonable to assume that the advantage of the plantation system declined while that of the peasant system increased correspondingly during this period. It is no wonder then that the smallholdingp-basedThai agricultural sector began to perform better than the plantation-based Indonesian and Philippine agricultural sectors.14Major increases in Thai exports of non- rice agricultural commodities such as rubber, kenaf (hibiscus cultivated for its fibre) and cassava tips were based entirely on smallholder production. While non-rice exports in Thailand were, to a significant extent, su ported p by open frontiers that enabled relatively fast increases in cultivation area (Table 2.1), the important point is that the frontiers were exploited by smallholders and not plantations. Relative increases in the efficiency of the peasant system were not limited to Thailand. The dramatic rise in Indonesia's share of world coffee and cocoa markets was also based entirely on smallholder production (Akiyarna and Nishio 1996). Although some Philippine coconut plantations continued to operate, especially in Mindanao, it was coconut oil extracted from copra produced by smallholders that contributed most to the Philippines' capacity to maintain its high world market share of coconut oil production. One advantage of the plantation system is better coordination between lar e-scale processingand marketing and farm-level production. The peasant g system could gain this same advantage by organising contract farming, in which an agribusiness firm manages processing and marketing but contracts peasant fslmers to supply farm products. In return for their agreement to supply crops, the firm provides the peasants with technical +dance, credit and other services. In this way, the system can take advantage of peasants in farm production without sacrificin economies of scale in processing and g marketing. Contract farming depends not only on farmers' labour, but also on the management ability of rural people in developing economies. The system has enabledThailand, a relative newcomer to canned pineapple production, to surpass the Philippines, the former leading exporter, whose pineapple production is based on large plantations in Mindanao.15 AN ECOLOGICAL AND HISTORICAL I'ERSPECTIVE The dilemma of land reform We now turn to the question of how landlessness in the Philippines and land reform programs might be related to the relatively poor performance of agriculture in this economy. Attempts to mitigate social unrest rooted in pervasive landlordism in the Philippines by means of redistribution extend back to the US colonial regime. The framework of reform applied in the past four decades, however, was established by the Agrarian Land Reform Code of 1963, enacted under President Macapagd (Hayami et al. 1990). The major feature of the Code was the creation of owner/cultivators in rice and maize. This involved two steps Operation Leasehold converted share tenancy to leasehold tenancy with rent fixed at 25 per cent of the average harvest value for the preceding three normal years Operation Land Transfer transferred land ownership to tenants. In the latter operation, the government expropriated land .in excess of a landlord's retention limit (75 hectares) and compensated landlords with 10 per cent of the land value in cash and the balance in interest-free redeemable Land Bank bonds. The land was then resold to tenants for annual amortisation payments, to be completed within 25 years. In 1971 under President Marcos, the Code was amended to extend land reform to the entire nation, with all share tenants being converted to leaseholders. The 1971 Code was enforced by Presidential Decrees No. 2 and No. 27 under the Martial Law proclaimed in 1972. The landlord's retention limit was successivel reduced from 75 to 7 hectares, and the period for y amortisation payments was shortened to 15 years. It is easy to enumerate the shortcomings of the land reform programs carried out in the Philippines, yet there is no denying that large haciendas in Central Luzon were broken up and most tenants acquired the status of leaseholdersor arnortisingowners, although sizeable areas remained under the direct administration of landlords. It is clear that the beneficiaries of land reform captured a large economic surplus because of increased rice yields due to irrigation development, new varieties and fertiliser, while rent and amortisation payments remained fixed. Thus,land reform has been successful in transferringmuch of the economic benefits from absentee landlords to former sharecroppers. Om the other hand, land reform has created serious income inequality within village 40 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH communities, because strong population pressure has prevented wage increases despite agricultural productivity gains, and income of landless labourers has not risen. The regulatory nature of the reform programs applied in a discriminatory manner to a certain agricultural sector resulted in major distortions in resource allocation. The application of land reform mainly to land with tenants gave landlords a strong incentive to evict their tenants and cultivate their land themselves. However, because it is difficult to supervise wage labourers, agricultural production and labour income per hectare in small family farms are usually higher than in large farms that use hired labour. The exemption of land under the directadministrationof landlords, therefore, has reduced labour input per hectare to a below-optimum level and thereby reduced the income of labours. Equally serious were regulations on tenancy contracts-especially the prohibition of share tenancy and rent control-that reduced the incentive of large landholders to rent their land in small parcels. These regulations applied not only to landlords but also to the beneficiaries of land reform. As the i~?comeof formersharecropperswho had been converted into leaseholders or mortising owners rosesignificantly, many of them retreated from arduous farm work, leaving it to landless labourers. However, because their formal titles required them to work the land themselves, they hesitated to sublet their holdings to the labourers, since if the sub-lessees proved to the agrarian reform office that they actually tilled the land then the lease would be transferred. Thus, land reform beneficiaries had to continue to use hired labour in cultivation, even if they could not work themselves because of sickness, old age or involvement in non-farm activities. Inevitably, then, an imbalance between land and labour arose. The negative effects of land reform on production efficiency were also si nificant outside the rice and maize sectors. Although the cash crop sector g was not covered by reform programs (the Comprehensive Agrarian Reform Law of 1988 intendedto cover cash crops was not significantlyimplemented), landowners still fear the eventual expropriation of their land and thus have stopped investing in improvement of their land, including in tree planting. Some landowners have even preferred to leave their land lying idle rather than use it for agricultural production. This was often the case in frontier regions like Mindanao, which might be a reason for the Philippines' low rate of cropland expansion compared to that of Thailand (Table2.1). Clearly, 42 RURAL. DEVELOPMENT AND AGRICULTURAL GROWTH highly efficient at delivering imported goods to farmers. Their activities . were facilitated by major public investments in infrastructure, especially highways. This free trade system, supported by government provision of public goods, both increased the competitiveness of Thai rice production and encouraged the production of new export crops, thus promoting agricultural diversification (Siamwalla et al. 1990; World Bank 1987). Importantly, however, it is unlikely that such differences in government policies are ever independent of differences in agrarian structure, ecological conditions and historical development. For example, Hara (1994) argues that import substitution was pursued more stron ly in the Philippines g than in Thailand and other ASEAN economies because the business tlite who benefited from industrial protection in the Philippines originated from the landed oligarchy, therefore little countervailing power was mobilised against industrial protection. In contrast, the rural countervailing power in Indonesia, Malaysia and Thailand was comparatively strong, because the urban business tlite were predominantly ethnic Chinese. The rather harmonious division of labour between Thai farmers and Chinese traders that developed in the Siam kingdom may also have prevented the modern Thai government from adopting anti-market and anti-trader interventions. Further evidence of the remarkable success of the Green Revolution in Indonesia can be found in the rapid growth of land productivity in the country, the fastest among the three economies. This growth was based largely on the Suharto government'sstrong support of the rice sector, which continued for three decades and took the form of investments in irrigation anci agricultural research and extension, and subsidies for inputs and credits. These supports helped Indonesia overcome the sorts of problems that seriously damaged agriculture in oil-producing economies such as Nigeria in the 1970s-early 1980s (Hayami 1997). Apparently, Suharto's policy was designed to help maiitain stabilityin the country by protecting peasants, regarded as the stabilising block of society In the 1970s, the Masagam 99 Program in the Philippines promoted the Green Revolution by distributing new varieties, subsidised fertiliser and other inputs to farmers. In the absence in the Philippines of 'peasant fundamentalism' comparable to Indonesia's, however, the program only lasted about 10 years (Hayami and Kikuchi 2000), and the rice self- sufficiencyachieved in the Philippines in the 1970s could not be sustained in the 1980s. In contrast, Indonesia rose from being the world's largest rice importer in the 1970s to self-sufficiency in the 1980s (though, as a result AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 43 of the recent financial crisis associated with the downfall of Suharto, the country has again become a major importer). Meanwhile, in Thailand, under liberal trade policies, government support to promote feitiliser use was not effective (World Bank 1987). Theories about political economies are now largely conjecture. Yet, a nation's agrarian structure and value system, created under unique historical and ecolo ical conditions, should have a far-reaching influence on policy g choices. Analysis of these relationships will be a major challenge for future research. If such analysis was extended beyond a single region and used to compare regions, it could shed light on major questions in world development, for example, why does Africa lag behind in adopting innovations in agricultural technology comparable to the Green Revolution in Asia? ACKNOWLEDGMENTS The author gatefully acknowledges helpful comments from Takarnasa Akiyama,RobertAllen, Gershon Feder,and Yair Mundlak,as well as technical assistance from Kei Kajisa and Yue Yaguchi. The main content of this paper was published in the World Bank Research Observe Vol. 16, No. 2 (Fall 2001), under the title Ecology, History and Development: A Perspective from Rural Southeast Asia. NOTES ' The 'vent-for-surplus' theory (Myint 1965, 1971) focused on the development of 'empty land' with low population density, large unused tracts and abundant natural resources, typically found in Southeast Asia and East Africa at the onset of Western colonisation. When theseeconomieshad been integrated into international trade, their natural resources (previously of no value to indigenous people) acquired market value, since they could be used to produce primary commodities of high demand in the West. In this way, previously 'unused' resources became the source of economic development. For a more comprehensiveassessment of agricultural grow& in Asia, including Southeast Asia, see the Asian Development Bank's five-volume S d y of Ruralhia report. Especially relevant to this section are the report's overview: Rural Asia: beyond the Green Reuolution (Asian Development Bank 2000); volume 1: Tram$iming the Rural Rrian Economy: the unfinished reuolution (Rosegrant and Hazel1 2000); and volume 2: The Growth and SustainabilityofAgrculture in Asziz (Kaosa-ard and Rerkasem 2000). Another major study specifically addressed to Southeast Asian agriculture, entitled Qynamism of Rural Sector Growth:policyIessonsfrom EvtRFian countries,is in progressat the World Bank. 44 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Huke and Huke (1997)estimate paddy area in Indonesia,the Philippinesand Thailand in the mid 1990s at 9,441,000,3,456,000,and 9,806,000 hectares,respectively,though they do notspecifyto which years thesefigures pertain.The substitution of Hukes' estimatesfor the data used in Table 2.1 does not changethe conclusion of this paper. * In Thailand's flood-prone Chao Phraya delta areas and the drought-ptone areas in the country's northeast, short-statured modern varieties were difficult to grow. Also, Thai farmers were slow to adopt the modern varieties because of their low value as exports. The descriptions ofSoutheast Asian ecologicaland environmentalconditionsgiven here are mainly based on Takaya (1985). Whiiethe corvieobligationwas replaced by tax-in-kindor money,slaverywasalsogradually phasedout underthereignsofKingsMongkutand Chulalongkotn,endingin 1905 (Ingram 1971;Feeny 1982).Theeliminationofslaveryandthe corvkeshouldhavebeenan important factor contributingto the allocation of a greatershare of Thai labour to rice cultivation. ' One rai equals 0.16 hectares. Comparativeadvantage based on natural resource endowmentswas reinforced by colonial policies on farmlands and public investments in physical and institutional infrastructure. For example, a strong sugarcane research program organised by the Dutch colonial governmentsignificantlystrengthened the international competitivenessof the Indonesian sugar industry (Evenson 1976). This section draws heavily on Hayami (1994, 1996). l oAbsence of economies of scale in agriculture is also attested to by estimates of aggregate production based on inter-country, cross-sectional data (Hayami and Ruttan 1985). " It is not imperative to employ the plantation system in the manufacture of bladc tea. In Taiwan, smallholders produce both green and black tea with small-scale equipment. It can also be argued that plantationshave used the largefermentation plant as adevice to enforce work schedules and standardise product quality for export. In fact, farm production by smallholders based on the system of contractfarminghas recentlybeen developedin Kenya (Lamb and Mder 1982). l ZAll forestlands were state-owned but accessible to all, except valuable teak forests, which were an important source of the Kingdom's revenue (Feeny1999). ' Officialstatisticsoften record that ~ieldpet hectare of cash cropssuch as coffeeand rubber are higher on plantations than when grown by smallholders. However, these statistics do not take into account various ~roductsintercropped by smallholders with principal cash crops, whereas monoculture is the common practice on plantations. l 4In addition to this disadvantage, the ~lantationsector in post-independence Indonesia, whichexpropriatedtheestatesofDutch ~lanters,seems to havesufferedfromtheinefficiency common to state enterprises. Attempts to solve this problem include the 'nuclear estate' scheme, by which a state ~lantationacts as a processing and marketing centre with a AN ECOLOGICAL AND HISTORICAL PERSPECTILT 45 demonstration farm for technical extension, and smallholders are organised as in contract farming. These attempts have often been marred by the direct application of plantation technology and practicewithout an understandingof smallholder conditions (Barlowand l'omich 1991).ThecaseofIndonesiacontrastswiththatofMalaysia,whereprivateplantations are relativelyhighly efficientand are wellsupported by a cooperativeresearchand extension system. ' However, a high degree of entrepreneurialand managerialskill is needed to organiseand operatean efficientcontract farmingsystem. It is not easyto enforcecontractsthat stipulate quantity, quality and time of product delivery with a large number of smallholders. Agribusinesses that do not have the necessaryskill often fail at contract fanning. Thus, the ofcontract firmingso far has been mixed,even inThailand (Siamwalla1992). Thesameoperationalnecessitiesapplyin other regions, includingAfrica,whereitisreported that contract farming organised by government agencies is usually inefficient Oaffee and Morton 1995). REFERENCES AkiYama,T. and Nishio, A., 1996. Indonesia: Cocoa Boom: hand-offpolicy encourages smallholder dynamism, Policy Research Working Paper No. 1580, World Bank, Washington, DC. Ariff,M. and Hill, H., 1985. Export-Oriented Indwtrtalizdtion:the AS&W expm'ence,Allen and Unwin, Sydney. Asian Development Bank, 2000. Rural Asia: bqond the Green Revolution, Asian Development Bank, Manila. Available online at http:// www.adb.org/Documents/Books/default.asp#contents. Baldwin, R.E., 1956. 'Patterns of development in newly settled regions', Manchester School of Economics and Social Studies, 24(2):161-79. Barlow, C. and Tomich, T.P., 1991. 'Indonesian agricultural development: the awkward case of smallholder tree crops', Bulletin of Indonesian Economic Studies, 27(3):29-53. i Bautista, R.M., 1987. Production Incentive in Philippine Agrculture: effects of trade and exchange rate policies, International Food Policy Research Institute, Washington, DC. Beckford, G.L., 1972. Persistent Poverty: underdevelopment in plantation economies in the T&d Worh!,Oxford University Press, New York Binswanger, H.P. and Rosenzweig, M.R., 1986. 'Behavioral and material determinants of production relations in agriculture', Journal of Development Studies, 22(3):503-39. 46 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Booth, A., 1388. Agricultural Development in Indonesia, Allen and Unwin, Sydney. Boserup, E., 1965. The Conditions ofAgrtgrtculturalGrowth,Men and Unwin, London. Brewster, J.M., 1950. 'The machine process in agriculture and industry', journal of Farm Economics, 32(1):69-8 1. Evenson, R.E., 1976. 'International transmission of technology in the production of sugarcane', joumdl of Development Sdies, 12(2):208-31. Food and Agriculture Organization of the United Nations, 1971. 1970 Workd Censm of Apimlture: amhis and comparison of the results, Food and Agriculture Organization of the United Nations, Rome. -, n.d. FAOSTAT, FA0 online database, http://apps.fao.org/default.hun, Feeny, D., 1982. The Political Economy of Agricultural Productivity: Thai agricultural development, 188@-1975, University of British Colombia Press, Vancouver. -, 1999. 'Thailand and Japan: why was Japan first?', in Yujiro Hayarni and MasaoAoki (eds), The InstitutionalFoundztions ofEastA k n Economic Development, Macmillan, London:413-42. Furnivall, J.S., 1944. Netherhnd India: a study ofpluraleconomy, Cambridge University Press, Cambridge. -, 1948. ColonialPoliq and Pratice: a comparative study of B u m and Netherhnd India, Cambridge University Press, Cambridge. Hara, Y., 1994. Tonan Ajia Shokoku no Keizai Hatten [Economic Development of Southeast Asian Countries], Institute of Oriental Culture, University of Tokyo, Tokyo. Hayami, Y., 1994. 'Peasant and plantation in Asia', in G.M. Meier (ed.), From Chsical Economics to Development Economics, St Martin's Press, New York:121-34. ----, 1996. 'The peasant in economic modernization', American journal of Agricultural Economics, 78(5):1157-67. -, 1997. Development Economics:fiom thepoveq to the wealth of mtions, Oxford University Press, Oxford. -andKikuchi,M.,1981.AsianVillageEconomyattheCrossroaa2, University of Tokyo Press and Johns Hopkins University Press, Tokyo and Baltimore. -andRuttan,V.W.,1985.AgriculturalDeuehpment:aninternational perspective, rev. edn, Johns Hopkins University Press, Baltimore. AN ECOLOGICAL AND HISTORICAL PERSPECTIVE 47 -, Quisumbing, M.A. and Adriano, L.S., 1990. Toward an Alternative Land Reform Paradigm: a Philippine perspective, Ateneo de Manila University Press, Quezon City. - and Otsuka, K, 1993. The Economics of Contrdct Choice, St Martin's Press, New York. -andKikuchi,M.,2000.ARiceVilhgeSaga:threedecadesofGreen Revolution in rice in the Philippines, Macmillan, Barnes & Noble, and the International Rice Research Institute, London, New York, and Los Bafios. Hesters, E.D. and Mabun, E, 1924. 'Some economic and social aspects of Philippine tenancies', Philippine Agn'culturist, 12367444. Huke, R.E. and Huke, E.H., 1997. Rice Area by 7jpe of Cultu~:South, Southeast and East Axia, International Rice Research Institute, Manila. Ingram, J.C., 1971. Economic Change in Thaihnd 185@197Q, Stanford University Press, Stanford. Intal, PS. Jr and Power, J.H., 1989. TrdcZe, Exchange Rate, and Agn'mltural Pricing Policies in the Philippines, World Bank, Washington, DC. JafTee, S. and Morton,J., 1995. MarketingAficai High--Val& Food, Kendalll Hunt Publishing Co., Dubuque, Iowa. Kaosa-ard, Mingsarn Santikarnand Rerkasem, Benjavan, 2000. The Growth and Sustainability ofAgrimlture in Asia, Odord University Press (China), for the Asian Development Bank, Hong Kong. Available online at http:/ / w w w . a d b . o r g / D o c u m e n t s / / B o o k s / R u r a l ~ A s i a l Growth~Sustainability_Agriculture/default.asp. Lamb, G. and Muller, L., 1982. Control, Accountabili~and Incentives in a Successfil Development Institution: the Kenya Ted DevelopmentAuthority, World Bank, Washington, DC. Larkin, J.S., 1972. The Pampangans: colonial society in a Philippine province, University of California Press, Berkeley Lewis, W.A. (ed.), 1970. Tropical Development, 1880-1913: studies in economic progress, Allen and Unwin, London. Lim,Yongil, 1968. 'Input of tea industry on the growth of the Ceylonese economy', Social and Economic Studies, 17:453-67. McCoy, A.W., 1982. 'A queen dies slowly: the rise and decline of Iloilo City', in A.W. McCoy and E.C. de Jesus (eds), Philippine Socidl History: global t r d and local tran$owtions, Ateneo de Manila University Press, Quezon City:297-58. 48 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH McLennan, M.S., 1969. 'Land and tenancy in the central Luzon Plain', Philippine Studies, 17:651-82. Myint, H., 1965. The Economicsof DevelopingCountries,Praeger, New York -, 1971.EconomicTheoyand UnhAvelopedCountries,OxfordUniversity Press, New York. Pelzer, KJ., 1945. Pioneer Settlement in the Asiatic Tropics, American Geographical Society, New York. Pim, A., 1946. Colonial Agricultural Production, Royal Institute of International Affairs and Oxford University Press, London and New York. Resnick, SA., 1970. 'The decline in rural industry under export expansion: a comparison among Burma, Philippines and Thailand, 1870-1 938', Journal of Economic Histoy, 30(1):51-73. Rosegrant, M. and Hazell, P.B.R., 2000. Transforming the Rural Asian Economy: the unjnished revolution, Oxford University Press (China), for the Asian Development Bank, Hong Kong. Available online at http:// www.adb.org/Documents/Books/Rural~AsidUnfinished~Revolution/ default.asp. Siamwalla, Ammar, 1992. Myths, Demons and the Future of Thai Agriculture, Paper presented at the Year-End Conference on Thailand's Economic Structure: Toward Balanced Development, Thailand Development Research Institute, Bangkok, 12-13 December. -, Setboonsarng, Suthad and Patamasiriwat, Direk, 1990. Thai Agriculture: resources, institutions and policies, Thailand Development Research Institute, Bangkok. Takaya, K., 1985. TonanAjia no Shizen to Tochinyo [Nature and Land Use in Southeast Asia], Keiso Shobo, Tokyo. Warr, EG. (ed.), 1993. Thai Economy in Tramition, Cambridge University Press, Cambridge. - and Nidhiprabha, Bhanupong, 1995. Thaihndj MmoeconomicMimle: stable aajustment and sustained growth, World Bank, Washington, DC. Wickizer, VD., 1951. Coffee, Tea, and Cocoa: an economic and political analysis, Stanford University Press, Stanford. -, 1960. 'The smallholder in tropical export crop production', Food Research Institute Studies, 1:49-99. World Bank, 1987. Thaihnd:agro-industrialdivers;J;cation:issuesandprospects, World Bank, Washington, DC. -1993.TheEastAsianMiracle,OxfordUniversityPress,Oxford. , GROWTH OF THE AGRICLTLTLTRAL SECTOR: ARE THERE PECULIARITIES WITH SOUTHEAST ASIA? Agricultural gowth, which in developing regions has differed significantly over the past four decades, is widely regarded as an important basis for economic gowth (for example, Johnston and Mellor 1961; Hayami and Ruttan 1985; Mundlak 1986). For example, in most Asian countries the gowth rate of agricultural GDP far exceeded that of the population, while in many African countries the reverse was the case. Also, during the period 1972-92, while some countries' agricultural exports tripled, many others' declined significantly (World Bank 1996:Chapter 4). Only a few studies have quantitatively examined regional differences in agricultural gowth (for example, Hayami and Ruttan 1985; Stern 1996), however, and more are needed. Such researchwould help to identify some of the sector's regional and country-specific idiosyncrasies and some of the causes and consequences of its development, thus deepening our understanding of the mechanisms and dynamics of agricultural growth and its impact on overall economic growth and poverty reduction. The main objective of this chapter is to identify prominent characteristics of agricultural gowth in Southeast Asia in general and Indonesia, the Philippines and Thailand in particular, in comparison with other developing regions and countries. In the first section, we examine the movements of key statistics such as agricultural GDE labour, production, land, and land and labour productivity in order to make a regional comparison of sectoral development over the past 3 0 4 0 years. We then evaluate some of the variables that have influenced sectoral development and corresponding 50 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH policies, and in the same context examine investment and capital accumulation in agriculture and direct protection rates. In the third section, we quantitatively examine some of the consequences of agricultural development with a focus on overall income growth and poverty. THE SOUTHEAST ASIAN AGRICULTURAL SECTOR Agriculture in Southeast Asia has performed considerably better than in other developing regions over the past four decades (Table 3.1). For the period 1961-98, agricultural value added in Southeast Asia increased at 3 per cent per year, the highest growth rate among developing regions. This growth was driven by both area expansion and land productivity increases, the former contributing to about one-third of growth and the latter to about two-thirds. This was achieved with a relatively small increase in agricultural labour (1.6 per cent per year), resulting in a high growth rate for agricultural labour productivity, at 1.5 per cent per year the second highest figure among developing regions after Latin America. In the late 1990s, Southeast Asia's agricultural GDP per agricultural labourer was the second highest among developing regions after Latin America, and about double that of South Asia and West Africa and four times as much as that of East and Southern Africa (Table 3.2). The region's labour productivity, however, was much lower than that of Latin America, probably because of a considerablylower land-to-labour ratio (0.99 hectares per person in Southeast Asia and 17.1 hectares per person in Latin America). In Southeast Asia over the past two decades, despite a large increase in absolute terms, agricultural GDP's share of total GDP declined fiom 24 per cent to 15 per cent, a much larger decline than in other developing regions. This pattern of Southeast Asian agricultural growth and structural change matches the typical economic development pattern described by Timmer (1988) and Hayami (2001). In order to analyse the components of agricultural labour productivity, we examine the gowth of land productivity of, and area per, agricultural worker, a breakdown that indicates which components contributed to agricultural labour ~roductivitygrowth. This can be expressed algebraically as follows where gr is growth rate, q is agricultural GDE I is agricultural labour, and a is agricultural land. The first term on the right-hand side of the equation is GROWTH OF THE AGRICULTURAL SECTOR 51 Table3.1 Annualgrowthrateof agriculturalGDPby regions Growth rate Growth rate of Growth rate of of agricultural agricultural land a ricultural labour g value added (per cent) (per cent) (per cent) 1961 1981 1961 1961 1981 1961 1961 1981 1961 -80 -98 -98 -80 -98 -98 -80 -98 -98 1 m 1.54 1.54 1.67 Indonesia 1.11 1.86 1.64 Thailand 2.22 1.02 1.69 Philippines 1.98 1.25 1.57 China 1.90 1.25 1.66 Southeast Asia 1.52 1.47 1.62 South Asia 1.61 1.28 1.41 West Africa 1.30 1.44 1.32 Southern and East Africa Latin America Growth rate of Growth rate of Growth rate of agricultural value agricultural value agricultural land/ added/agricultural added/agricultural agricultural labour land (per cent) labour (per cent) (per cent) 1961 1981 1961 1961 1981 1961 1961 1981 1961 -80 -98 -98 -80 -98 -98 -80 -98 -98 IPT" 3.22 2.23 2.55 2.50 1.21 1.74 -0.72 -1.02-0.83 Indonesia 3.93 2.52 3.18 2.74 1.28 1.97 -1.19 -1.24-1.29 Thailand 1.99 2.89 2.28 2.44 2.32 2.29 0.45 -0.57 0.01 Philippines 2.25 1.15 1.28 1.83 0.18 0.90 -0.42 -0.95-0.35 China 2.28 3.57 2.79 1.67 3.34 2.55 -0.66 -0.23-0.25 Southeast Asia 2.92 1.89 2.16 2.54 1.14 1.49 -0.69 -0.75-0.70 South Asia 2.14 3.19 2.74 0.73 1.93 1.43 -1.37 -1.25-1.31 West Africa 0.50 2.80 1.51 -0.30 1.67 0.45 -1.21 -1.12-1.13 Southern and East Africa Latin America aIndonesia, the Philippines and Thailand. Notes: Southern and East Africa does not include South Africa. Latin America does not include Trinidad and Tobago.The growth rates were estimated using linear and semilog regression models and selecting the one with the highest RZstatistic. In the linear case, the growth rates where calculated dividing the coefficient by the mean of the period considered. Estimates that involve agriculture value-added for Southeast Asia, West Africa, Southern and East Africa, and Latin America begin in 1970 and 1965 for South Asia. Source: World Bank, n.d. World Deuelopment Indicators Online,http://worldbank.org/data/ onlinedatabases/onlinedatabases.html;Food and Agriculture Organization, n.d. FAOSTAT (FA0 Statistical Database), Food and Agriculture Organization of the United Nations, Washington, DC, http://apps.fao.org/. 52 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Table 3.2 Agricultural valueadded, land and labour Agricultural Agricultural Agricultural value added value added value added (dollars) (dollars/agriculturaI worker) (dollars/hectare) IPT Indonesia Thailand Philippines China Southeast Asia South Asia West Africa Southern and East Africa Latin America Agricultural Agricultural labour value added as as a share of Agricultural land share of total GDP total labour (halagworker) (per cent) (per cent) 1981-83 1996-98 1981-83 1996-98 1981-831996-98 IPT 1.0606 0.9159 23.8 15.4 59.0 50.6 Indonesia 1.0436 0.8740 27.0 16.8 57.3 50.4 Thailand 1.0982 1.0143 20.4 11.5 69.5 58.8 Philippines 1.0571 0.9136 22.8 20.5 51.0 41.4 China 1.0721 1.0522 39.5 18.8 73.5 68.3 Southeast Asia 1.1053 0.9903 24.1 14.8 58.0 49.1 South Asia 0.8113 0.6725 34.5 25.1 68.3 59.9 West Africa 6.8922 5.8045 29.3 31.6 65.2 54.6 Southern and East Africa 4.3370 3.0725 25.3 27.2 78.3 72.9 Latin America 15.7450 17.0867 7.9 7.3 32.9 21.4 Notes: Southern and East Africa does not includeSouthAfrica. Latin America does not includeTrinidad and Tobago. AU dollar figures are 1995 US$ millions. Source: World Bank,n.d. World Development IndicatorsOnline, http://www.worldbankorg/ &ta/onlinedatabases/online&tabases.htd; Food and Agriculture Organization of the United Nations, n.d. FAOSTAT (FA0 Statistical Database), Food and Agriculture Organization of the United Nations, Washington, DC, http://apps.fao.org/. GROWTH OF THE AGRICULTURAL SECTOR 53 land productivity growth and the second term is the growth rate of area per agricultural worker. These components are shown in Table 3.1, and they show that land ~roductivitygrowth in Southeast Asia over the past four decades was second hi hest among regions after South Asia. Another characteristic of the g Southeast Asian agricultural sector is that the growth rate of agricultural land per worker over the past four decades has declined at a rate of 0.7 per cent per year, much lower than the rates of South Asia (1.3 per cent), West Africa (1.1 per cent) and Southern and East Africa (2.2 per cent). The trends of the land-to-worker ratio vary significantly among the three study countries. Thailand's ratio declined modestly (0.6 per cent per year), while the declines in Indonesia (1.2 per cent per year) and the Philippines (1.0 per cent per year) over the past two decades have been more significant. Although the agricultural sector grew rapidly over the past four decades in Southeast Asia in general, important differences in sectoral development can be seen among the study countries. Thailand achieved the highest agricultural GDP growth rate of the three (Table 3.1) at 4.1 per cent per year, to which land expansion contributed 40 per cent. The contribution of land expansion to agricultural GDP growth in the Philippines was similar, but the rate of growth was much lower at 2.4 per cent per year. Conversely, Indonesia's growth of 3.7 per cent per year was almost all due to land productivity growth, which contributed 90 per cent. One of the most prominent characteristics of the agricultural sectors of many countries in East Asia is that they are strongly influenced by heavy monsoon rains during the summer and relatively less rain during other periods of the year (Oshima1987). Under this weather pattern, it is difficult to grow crops other than rice, hence rice is both the main crop and the main staple in these countries. The average rice share1 of a ricultural land g was over 60 per cent for Southeast Asia in the late 1990s (Table 3.3). The shares in the three target countries ranged between 58 and 64 per cent. It is notable that Indonesia, traditionally a rice-importing country, increased its rice share from 39 per cent in the early 1960s to 64 per cent in the late 1990s. This reflects Indonesian efforts to gain self-sufficiency in rice. In contrast, Thailand, the world's largest rice exporter, reduced its rice share from 60 per cent to 58 per cent during the same period, the result of the diversification of Thai agriculture over the past four decades. The rice area share in the Philippines has remained relatively stable since the early 1960s at around 65 per cent. The rice share of South Asia was about one-half that 54 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Table 3.3 Production,growth rateand land sharefor rice,by region Rice share of Rice production Rice production land arable (metric tons1 ('000 metric tons) (per cent) hectare) I r n Indonesia Thailand Philippines China Southeast Asia South Asia West Africa Southern and East Africab Latin Americac Annual growth rate Annual growth rate of rice production of rice yield (per cent) (per cent) 1961-80 1981-98 1961-98 1961-80 1981-981961-98 I P F Indonesia Thailand Philippines China Southeast Asia South Asia West Africa Southern and East Africab Latin Americac aIndonesia,the Philippines and Thailand Southern and East Africa does not includeSouth Africa Latin Americadoes not includeTrinidadand Tobago Source: Food and A riculture Organization of the United Nations, n.d. FAOSTAT (FA0 g Statistical Database),Food and Agriculture Organization of the United Nations, Washington, DC, htrp://apps.fao.org/. GROWTH OF THE AGRICULTURAL SECTOR 55 of Southeast Asia in the late 1990s, but it has increased since the early 1960s. The shares for African regions and Latin America have been small, ranging between 4-7 per cent over the past four decades. Because of the importance of this crop, productivity growth for rice has had a great impact on overall agricultural productivity and growth of the SoutheastAsian region. Thanks to high-yieldingvarieties (HYVs),the region's rice production has increased substantially over the past four decades, especially in the 1960s and 1970s,at a rate of almost 4 per cent per year. It should be noted, however, that about 40 per cent of this growth was due to land expansion. Thanks to the rapid growth of rice production, Southeast Asia achieved a growth rate of food supply in terms of calories/capita/dayof 1.23 per cent per year, the highest among developing regions (Table 3.4). This rate is at least double that in other regions. This achievement was probably one of the foundations of Southeast Asia's overall economic growth. Among the three study countries and over time, however, the rates varied significantly. During the period 1961-98, Indonesia had the highest overall food supply growth rate, followed by the Philippines and Thailand. Thailand's low rate corresponds to the low growth rate of its rice production. The rate in the Philippines was high during the period 1961-80 (1.34 per cent per year), indicating the early introduction of H W rice, but declined significantly during the period 1981-98 (0.56 per cent per year). Southeast Asia's monsoon climate and high population density have precluded development of animal husbandry in the region (Oshima 1987:25). Nor is there sufficient land to grow both food and feed. The region's pastureland share has been declining; by the late 1990s it was only 16 per cent (Table 3.5). Among the three countries, Indonesia has the largest share of pastureland, but its share has declined over time and by the late 1990s was down to 27 per cent. The pastureland shares of African regions and Latin America of almost 80 per cent are more than five times higher than that of Southeast Asia. REGIONAL DIFFERENCES I N PRODUCTMTY GROWTH Agricultural productivity and its growth are critically dependent on climatic and soil conditions and cultural and historical developments (see Chapter 3 on the impact of historical developments on the agricultural performance of the three study countries). Other critical factors include government policies, both sector-specific and sector-neutral, and the politics and 56 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Table 3.4 Food supply in caloriesper capita per day and growth rate of foodsupply Food supply Annual growth rate (calories/capita/day) of food supply (per cent) 1961-63 1981-83 1996-98 1961-80 1981-98 1961-98 IPT 1,757 2,247 2,699 1.21 1.28 1.27 Indonesia 1,697 2,269 2,883 1.34 1.68 1.65 Thailand 2,011 2,228 2,439 0.65 0.60 0.33 Philippines 1,720 2,194 2,392 1.34 0.56 1.00 China 1,717 2,477 2,936 1.35 0.98 1.45 Southeast Asia 1,792 2,271 2,710 1.17 1.24 1.23 South Asia 2,030 2,107 2,422 0.06 0.87 0.55 West Africa 2,142 2,038 2,552 -0.39 1.73 0.46 Southern and East Africa 2,150 2,156 1,911 0.17 -0.83 -0.36 South America 2,329 2,594 2,808 0.59 0.53 0.47 Notes: West Africa does not include CentralAfrican Republic, Chad or Republic of the Congo, but now includes Cape Verde, Guinea and Liberia. The growth rates were estimated using linear and semilogregression models and selecting the one with highest R2 statistic. In the linear case, the gro-wthrates were calculated by dividing the coefficient by the mean of the period considered. Source: Food and Agriculture Organization of the United Nations, n.d. FAOSTAT (FA0 Statistical Database), Food and Agriculture Organization of the United Nations, Washington, DC, http://apps.fao.org/. Table 3.5 Shareof pastureland in agricultureby region(percent) IPT Indonesia Thailand Philippines China Southeast Asia South Asia West Africa Southern and East Africa Latin America Note: Southern and East Africadoes not includeSouthAfrica. Source: Food and Agriculture Organization of the United Nations, n.d. FAOSTAT (FA0 Statistical Database), Food and Agriculture Organization of the United Nations, Washington, DC, http://apps.fao.org/. GROWTH OF THE AGRICULTURAL SECTOR 57 institutions that underlie these policies. Details of these policies for the three study countries are discussed in other chapters in this volume. In this section, we compzre some of the quantifiable variables considered to affect growth of the agricultural sector-land distribution, investments and capital accumulation in the agricultural sector, pricing, and exchange rate policies. Initial conditions A number of studies have found initial conditions, especially those related to education, income and wealth inequality, to be extremely important in economic development in general and perhaps the most important factors in the rapid economic growth of East Asia in particular (Rodrik 1994). Clearly, in terms of educational levels and other human capital indicators, East Asia stands out (Rodrik 1994:17). Another initial condition that seems to have profound influence on subsequent economic growth is wealth inequality usually measured by land ownership. Recent studies suggest that, although the relationship between income inequality and lower subsequent growth may be tenuous, wealth inequality appears to be a major determinant of economic growth (Deininger and Squire 1997; Deininger and Olinto 2000). Although these studies use land ownership as a proxy for asset or wealth ownership because of data availability, their findings still point to the importance of asset inequality in rural areas and hence the importance of rural areas to economic development. These recent studies contrast with the well-known Kuznets' hypothesis and a number of past efforts to confirm the hypothesis (Deininger and Squire 1997; Griffin and Ickowitz 2000). Theoretical explanations for this relationship proceed through two channels,first that a high level of inequality in wealth implies that a large number of poor people are unable to obtain credit to make profitable investments, for example in schooling, and second that it implies that a large number of poor are unable to participate in political bargaining. Deininger and Squire (1997) consider the former to be more important. Table 3.6 shows Gini coefficients for the initial land distribution for several regions and the three study countries. It is evident from these data that land distribution in the East Asia and the Pacific region is considerably more equal than in other regions. Among the three target countries, Thdand's land distribution is most equal, and its Gini coefficient decried 58 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH over time. This is probably due to the colonial histories of Indonesia and the Philippines and the importance of plantations to the agricultural sector, while Thailand was never colonised and has never had a plantation system. Investment and capital accumulation Agricultural productivity can be increased through the introduction of technologies and machines that are either land saving, labour saving, or both. This requires considerable investment. Mundlak et al. (1997)collected agricultural investment data for over 70 countries and analysed the relationship between agricultural capital and productivity growth. They found that the two variables are strongly related and that prices have limited direct effects on productivity in the short term but significant effects in the long run through their impact on investment. Table 3.6 Ginicoefficientsfor initial land distribution, by region, 1950-90 (decadal medians for regions and specific year data for individual countries) 1950s Sub-Saharan Africa East Asia and Pacific 44.84 OECD and high- income countries 58.43 South Asia 67.81 Middle East and North Africa 78.30 Latin America 82.00 Eastern Europe 62.03 Indonesia Philippines Thailand " Data for 1963 Data for 1973 'Datafor1960 Data for 1971 'Datafor1980 Data for 1963 gAverageof data for 1980 and 1988 Source: Deininger, K and Squire, L., 1997. 'Economic growth and income inequality: re- examiningthe links', Financeand Dewehpmmt, 34(1):3841;World Bank, n.d. World Development Indicators Online, World Bank, Washington, DC. Available online at htq~:ll www.worldbank.org/data/onlinedatabaseslonlinedatab~es.h~. GROWTH OF THE AGRICULTURAL SECTOR 59 Growth rates for agricultural capital in several developing regions and the target countries are given in Table 3.7.' It appears that many developing countries invested much more in the agricultural sector during the 1970s than during the 1980s. All regions had positive trends in agricultural capital during the 1970s, but all except Southeast and East Asia turned negative during the 1980s. The sharp decline in world agricultural commodity prices that occurred throughout the 1980s could have been an important factor affecting the large negative growth rates for the Atiican regions, South Asia, and Latin America. The rate of capital accumulation in Southeast Asia was high, although it declined from 6.9 per cent per year in the 1970s to 2.8 per cent per year in the 1980s. High capital accumulation was probably one of the main reasons for the high growth rates of agricultural production in Indonesia and Thailand during the 1980s. Conversely, the stagnant performance of the Philippine agricultural sector was probably due to the negative agricultural capital growth during the 1980s. Protection policies As discussed by AkiYatnaand Kajisa (Chapter 9, this volume), direct price intervention has a profound impact on the performance of the agricultural sector. Governments intervene in agricultural product pricing for various reasons, including food security, stabilisation of producer prices, and the necessity of taxation. These policies are often cited as critical in explaining Table 3.7 Annual growth rate of agriculturalfixed capital IPTlSoutheast Asia Indonesia Thailand Philippines East Asia South Asia Southern and East Africa Latin America Note: Data for Thailand are from agriculmal fmed capital because data for total agricultural capital are not available for Thailand. Furthermore, the calculation for IIT is based on this fixed capital data for Thailand,and total agricultural capital for Indonesiaand Philippines. Source: Larson, D.F., Crego,A, Butzer, R. and Mundlak, Y., 2000. X cross-country database for sector investmentand capital', World Bank Economic &im, 14(2):371-91. 60 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH differences among developing countries in the performance of agricultural production and exports (Schiff and Valdes 1992). A straightfonvard evaluation of the extent of taxation and subsidies can be conducted by examining the ratio of domestic to border prices, ignoring domestic transportation and marketing costs (Table 3.8). For the period examined, average domestic agricultural prices in Southeast Asia have been higher than average border prices, and the ratios were considerably higher than those of African regions. The negative growth rate of this ratio during the 1970s indicates increasing taxation on exported products at a time of booming world commodity prices. Another way to evaluate the extent of government intervention, as discussed by Mundlak and Larson (1992) and Akiyarna and Kajisa (Chapter 9, this volume), is to perform the following regression equation to examine the relationship between domestic and border prices where DP is domestic price and BP is border price. The coefficient 6 measures the extent of the response of domestic prices to border price--that is, how much domestic price chan es in response to g changes in border price. It also indicates the extent of price intervention- thus, the doser the value of 6 to unity, the lower the extent of intervention, and vice versa. Mundlakand Larson (1992)argue that the R2 of the regression is another good indicator of the extent of the linkage between domestic and world prices and hence of government intervention.This statistic indicates how closely the two prices move together. The regression in Equation 3.2 was run on the prices of a number of countries and commodities for the period 1961-95 (Table 3.9). The results show that the values of 6 for Southeast Asia and Latin America are much closer to unity than those for South Asia, East and Southern Africa, and West Africa. The values of the R2 statistics for SoutheastAsia are lower than those for East and Southern Africa and Latin America, but significantly higher than those for South Asia and West Africa. The transmission elasticity and R2 are higher for the average of the three study countries than for Southeast Asia, because for the latter the averages are lowered by inclusion of Laos and Myanmar. The implication of this analysis is that agricultural product prices in Southeast Asia, especially the three target countries, and Latin America have been linked to world prices considerably more closely that those of South Asia and West Afiica. GROWTH OF THE AGRICULTURAL SECTOR 61 -- Table 3.8 Generalprotectionratefromagriculturaldomesticprice perborder price,byregion Annual growth rate of Agriculture domestic price/ agriculture domestic price1 agriculture border price agriculture border price IPT Indonesia Thailand Philippines China Southeast Asia South Asia West Africa Southern and East Africa 0.98 1.17 1.05 2.35 -2.35 0.40 Latin America 0.91 1.27 1.40 1.52 -0.66 1.59 Note: A price adjustment in Ghana at the beginningof 1980 prevents us from estimatingthe true characteristics of the economy, thus we have removed Ghana from the observationof West Africa in 1981-83. Data for Thailand are from agricultural f d capital because data for total agriculturalcapital are not available for Thailand. Furthermore, the cdculationfor IPT is based on this f ~ e capital data for Thailand,and data for total agricultural capitalfor d Indonesia and Philippines. Source: Food and AgricultureOrganization of the United Nations, n.d. FAOSTAT (FA0 Statistical Database), Food and Agriculture Organization of the United Nations, Washington, DC, http://apps.fao.org/. Table3.9 Transmissionof border todomesticprice,by region Transmission elasticity Average R2 IPT 1.02 0.83 Indonesia 0.95 1.oo Thailand 1.10 0.70 Philippines 1.00 0.79 China 1.00 0.79 Southeast Asia 1.02 0.76 South Asia 0.77 0.57 West Africa 0.78 0.69 Southern and East Africa 0.90 0.85 Latin America 0.97 0.81 Source: Food and Agriculture Organization of the United Nations, n.d. FAOSTAT (FA0 StatisticalDatabase),Food and Agriculture Organization of the United Nations, Washington, DC, http://apps.fao.org/. 62 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH CONSEQUENCES OF GROWTH IN AGRICULTURAL PRODUCTIVITY A number of development economists consider development of the agricultural sector to be a prerequisite for development of the overall economy. Recognition from economists of the agricultural sector's importance, especially in terms of supplying food, dates back to Malthus and Ricardo. More recently, economists such as Ranis and Fei (1961), Jorgenson (1961), and Johnston and Mellor (1961) developed and confirmed the essence of this idea. Also, because the agricultural sector makes up a large part of the overall economy in most developing countries, it is difficult to achieve economicgowth and subsequent economicstructural changes without a certain period of healthy growth in the agriculturalsector. In this section we examine the impact of agriculturalsector performance on GDP, exports, and poverty in several developing regions. Agricultural sector and GDP growth According to classical economic thought on the theme of the agricultural sector's role in overalleconomic growth, such as that of Adam Smith,Thomas Malthus,John Stuart Mill and David Ricardo, an economy will not develop if there is not an adequate and increasing food supply to meet the demand arising from population growth. This 'Ricardian trap' may be avoided through international trade. However, as Hayami (2001) shows using cross- country data for the period 1965-95, there is a strong correlation between per capita gowth of GDP and per capita food production. As discussed above, the per capita calorie supply in developing regions shown in Table 3.4 also supports the importance of food availability to economic development. To examine the relationship between the agricultural sector and overall economic growth rates, we conducted a cross-section regression on the gowth rates of GDP and manufacturing GDP using agricultural GDP as an explanatory variable for the periods 196698 for 60 developingcountries and 1980-98 for 73 developing countries (Table 3.10). It is clear that the gowth rates of overall economic and non-agricultural activities are closely correlated with those of the agricultural sector. Statistically, the agricultural sector growth variable is highlysignificant f-brall the cases examined. In the more recent period, however, the significance and hence the relationship weakened. This is probably because the non-agricultural sector's share of GDP in many developing countries has increased substantially in recent years, reducing the impact of the agriculturalsector on the overall economy. GROWTH OF THE AGRICULTURAL SECTOR 63 Table 3.10 Relationships among growth rates of agricultural GDP, manufacturing GDP, and GDP Dependent variable GDP GDP Manufacturing GDP Explanatory Non-agricultural variable Agricultural GDP GDP Agricultural GDP 1966-98 1980-98 1966-981980-98 1966-98 1980-98 Coefficient 1.15 0.44 0.8 0.75 1.55 0.7 t-value 6.79 3.03 32.51 35.2 3.88 2.3 Observations 49 68 49 68 34 53 R2 0.5 0.12 0.9 0.95 0.32 0.09 Source: Author's calculations. Performance of agricultural product exports At the pre-industrialisation stage of development, agricultural products are often the main earner of foreign exchange, which is necessaq for many developing countries to import capital goods and products to enhance their welfsre. High agriculturalproductivity growth is usually necessary to support high growth of agriculturalexports. southeast Asia achieved high agricultural export value growth not only through high export volumes made feasible by high agricultural production growth, but also through switching to higher-value agricultural products. For the period 1961-97, agricultural export volume from Southeast Asia increased at the highest rate among developing regions, though for the period 1981-97 it was surpassed by South Asia (Table 3.11). There are significant differences in the export performance of agricultural product volume among the three study countries-that of Indonesia and Thailand increased at over 5 per cent per year during 1961-97, while that of the Philippines recorded a low rate of 1.1 per cent per year during the same period. The figures for the more recent period of 1981-97 indicate that differences were even greater- Indonesia increased at 7.6 per cent per year while the rates for Thailand and the Philippines were O per cent and 4 . 6 per cent per year, respectively Significant differences in the growth of export unit values among regions and countries are influenced by the movements of world prices for major exports. Southeast Asia recorded the highest export value growth among all regions during all the periods examined. It is clear from Table 3.11 that these high export value growths were achieved by high growth in export volume and relatively stable unit values. In contrast, export unit values of 64 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Sub-SaharanAti-icaand Latin America fluctuated wildly.The large difference in the export unit values of Sub-Saharan Africa and Latin America between the periods 1961-80 and 1981-97 reflects the large share of coffee and cocoa in these regions' agricultural exports and the sharp declines in these commodities' prices in world markets between the two periods. It is impressive that Thailand achieved such high growth in its agriculturalexport unit value during 1981-97, while other regions suffered fiom sharp price declines for many major commodities throughout the 1980s and early 1 9 9 0 ~ . ~ The trends of export values and average unit values show interesting differences among the threestudy countries. Indonesia achieved high growth of export value through high export volume growth, including palm oil, rubber, and cocoa. Conversely, Thailand's high export value growth was due to increases in theaverage unit value,suggestinga shiftof export products from low to high value. It appears that Thailand's agriculturalexport growth followed Myint's (1965) 'vent for surplus' theory and Watkins' (1963) 'staple the or^'.^ Table 3.11 Annual trends for agricultural exports, by region Growth race of Growth rate of Growth race of value quantities unit values 1961 1981 1961 1961 1981 1961 1961 1981 1961 -80 -97 -97 -80 -97 -97 -80 -97 -97 IPT 11.1 5.6 8.4 6.47 1.99 4.70 4.59 3.13 3.59 Indonesia 10.8 7.7 8.8 5.35 7.57 5.17 5.16 0.33 3.32 Thailand 12.7 5.7 10.1 8.53 -0.09 5.96 3.95 5.53 3.86 Philippines 9.5 0.7 4.3 4.42 -0.62 1.09 5.15 0.75 2.99 China 11.6 5.7 8.7 6.79 3.23 5.51 4.42 1.71 2.52 Southeast Asia 11.2 5.5 8.5 6.68 3.25 5.41 4.22 1.64 2.52 South Asia 5.8 3.5 5.1 2.05 4.66 2.66 3.78 -1.44 2.29 West Africa 9.1 1.0 4.5 -0.84 2.64 0.34 10.28 -1.87 4.03 Southern and East Africa Latin America Source: Food and Agriculture Organization of the United Nations, n.d. FAOSTAT (FA0 Statistical Database), Food and Agriculture Organization of the United Nations, Washington, DC, http://apps.fio.org/. GROWTH OF THE AGRICULTURAL SECTOR Poverty reduction Poverty reduction has recently come to be recopised as one of the most important goals of economic development (World Bank 2001). The link between economic development and poverty reduction, however, has been hotly debated, especially since Kuznets proposed his now famous Kuznets' Curve. Recent findings suggest that economicgrowth and poverty reduction are compatible even in developing countries (World Bank 2001:Chapter 2), and that agriculturallrural sector development is critical in reducing poverty because 'nearly three quarters of the poor will continue to live in rural areas well into the next century' and '[tlhe significant majority of the rural poor depend on agriculture' (World Bank 1997:l). Analyses by Ravallion and Datt (1996) on Indian data, and Gallup et al. (1998) and Timmer (1997) on cross-country data show that agriculturallruralsector gowth tends to have a more significant impact on poverty reduction than the growth of other sectors. Further, Timmer argues that the impact on poverty is less significant when income inequality is severe. An interesting recent finding is that 'high initial inequality reduces the poverty impact of a given economic growth' (World Bank 2001:56; see also Timmer 1997). In the context of the agriculturalsector, the initial condition - effect may be due in part to the difference in the agrarian system. Rural and overall income inequality is usually larger in countries with a plantation system than in countries where agriculture is carried out predominantly by smallholders. In a plantation system, a large portion of agricultural income growth usually benefits large-scale landowners. Conversely, in a smallholder system, poor smallholders benefit more directly hom agricultural growth. As Hayami (Chapter 2, this volume) argues, smatkolder farming tends to be more labour intensive than the plantation system, meaning sectoral growth benefits landless hmers more under the former system, thus making- a greater contribution to poverty reduction in rural areas. We analysed the impact of agriculturaland non-agriculturalsector growth on poverty by regions using the data and approach of Timmer (1997) (Table3.12). The results show the income elasticitiesof five income quintiles for agricultural and non-agricultural GDP growth in several developing regions. The first income quintile represents the 20 per cent of the countries' population with the lowest income and the fifth income quintile represents the 20 per cent of the countries' population with the highest income. The income elasticities given in the table indicate the percentage change i~ income of a certain income quintile per percentage change in sectoral GDI? 66 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH For the estimation, fured effects as country dummies are used to account for non-systematic differences between countries within regions. The estimation is restricted so that the average income elasticity for each region over both sectors (agriculture and non-agriculture) is equal to unity. Two regions, Southeast Asia and Africa, show trends in the elasticities that indicate the impact of agricultural GDP gowth on the poor is stronger than that of non-agricultural GDP growth. Also, agricultural GDP growth has a stronger impact on the poorer quintiles than on the richer quintiles. This is especially the case for Africa, where the impact of agricultural GDP growth is considerably larger than that of non-agricultural GDP growth on the first-fourth quintiles. These findings are consistent with the fact that the agricultural sector is the most important sector in this region and the majority of farming is conducted by smallholders. For Southeast Asia, the same trend can be found for the second-fourth quintiles, with the trend broken by the richest (fifth) and the poorest (first) quintiles. It must be noted, however, that these two quintiles represent the extremes of income distribution and are therefore more likely to show distorted results. The other three regions under consideration do not show such trends. The trend in Latin America is in the opposite direction; for example, Table 3.12 Elasticitiesof incomegrowthand sectoralgrowth, estimatedusingcountrydummiesasfixedeffects 1st 2nd 3rd 4th 5th quintile quintile quintile quintile quintile Africa, agriculture 1.913 1.193 1.076 1.009 0.285 Africa, non-agriculture 1.355 0.967 0.900 0.880 0.420 Southeast Asia, agriculture 0.190 1.335 1.222 1.094 1.318 Southeast Asia, non-agriculture 0.457 1.122 1.108 1.008 1.146 East Asia, agriculture 0.647 0.655 0.821 1.178 1.814 East Asia, non-agriculture 0.757 0.784 0.871 1.064 1.410 South Asia, agriculture 1.101 0.965 0.973 1.031 0.898 South Asia, non-agriculture 1.189 1.049 0.963 0.915 0.916 Latin America, agriculture 0.410 0.938 1.192 1.383 1.221 Latin America, non-agriculture 0.483 0.911 1.I02 1.359 1.001 Source:Author's estimates based on data and model used in Timer, C.R, 1997. How Well Do the Poor Connect to the Growth Process?, CAER I1Discussion Paper No. 17, Harvatd Institute for International Development, Cambridge, Massachusetts. GROWTH OF THE AGRICULTURAL SECTOR 67 agricultural GDP gowth has a larger impact on the income growth of the richer income quintiles. This probably reflects the prevalence of the plantation system in the region. South Asia does not seem to have any obvious trends. The results of this analysis are consistent with previous studies that show the importance of agricultural sector gowth to poverty reduction but also suggest that the degree of the agricultural sector's impact is dependent on the type of agrarian system. CONCLUDING REMARKS Examination of variables related to agricultural sector performance in various developing regions and countries over the past four decades suggests that agriculturalsector growth is important for overalleconomicgo&, especially at an early stage of development. Strong correlation among growth of the overall economy, agricultural sector, food production, and agricultural product exports suggests that it would be difficult for many developing countries to achieve sustainable economic growth with a weak agricultural sector. The growth ofSoutheastAsia's agriculturalsector over the past four decades was very high compared with that of other developing regions' agricultural sectors. Because rice has been the most important crop for all countries in the region, the rapid increase in rice yields due to adoption of HYVs in the 1960s and 1970s appears to have formed a basis for the high growth of the re ion's agriculturalsector. Governmentsof SoutheastAsian countriesseemed g to have played an important role by implementing policies encouraging the adoption of the new varieties through large investments in irrigation facilities and extension services. This contrasts with the situations of Sub- Saharan Africa and Latin America, where government policies have generallyl not favoured smallholder farmers. In addition, many farmers in Southeast Asia have been able to invest in agricultural land, probably because land tends to be distributed more equally in Southeast Asia than in other developing regions. The performance of agricultural production and exports in Southeast Asia countries, especially Thailand, indicates that these countries not only increased production volumes of traditional crops but also succeeded in diversifying into commodities in high demand, especially higher-valued agricultural products. This enabled them to escape from the vagaries of fluctuating world agricultural commodity prices, as in the late 1980s and 68 RURAL DEVELOPMENT AND AGRJCULTURAL GROWTH early 1990s,and points to the importancefor agriculturalsector performance of flexibility and adaptability to changing market conditions. Another difference between policies implemented by Southeast Asian countries and policies implemented in other regions is that the former appear to have introduced less distortion for agricultural product prices, except for a few commodities such as rice. Akiyama and Kajisa (Chapter 9, this volume) argue that direct protection policies in Southeast Asia were aimed more at short and long-run price stabilisation, in contrast to the heavily distortionist policies implemented in many Sub-Saharan African countries up until recently (see Bates 1981, Lipton 1977, Akiyama et al. 2001). The differences among the regions in key statistics related to agriculture can also be observed in the three Southeast Asian countries studied in this volume. Agricultural sector performance has been considerably better in Indonesia and Thailand than in the Philippines. This can be traced to the larger investments in rural areas and more favourable protection policies in Indonesia and Thailand than in the Philippines. Along with its impact on overall economic growth, the agriculturalsector has significant effects on poverty because a majority of the poor in most developing countries live in rural areas. The effect is likely to be especially great in countries where the agricultural sector is large relative to other sectors, where the agricultural sector is predominantly smallholder-based, and where initial income inequality is small. Causes of Southeast Asia's greater success in reducing poverty relative to other developing regions are probably the high growth of its agricultural sector and the prevalence of smallholders in its agricultural sector. The development of Southeast Asia's agricultural sector appears to have followed the path described by many development economists for its performance, role and impacts. Triggered by 'miracle rice', the sector contributed greatly to the region's plentiful food supply, which in turn allowed the sector to earn foreign exchange, reduce poverty and form an important base for the growth that followed in the industrial sector. Such processes were possible because of the flexibility and adaptability of government policies and private-sector stakeholders in the region. NOTES Averagesin this chapter are kept as simple algebraicones to avoid the dominant influence of large countries. GROWTH OF THE AGRICULTURAL SECTOR 69 The number of countries for which data are available is considerablylimited and there are no data for West Africa. Also, because the sources of most data are from each individual country, strict inter-country or inter-region comparisons might be difficult. See, for example, various issues of the World Bank publication GlobalEconomic Prospects. See Chapter 10 of this volume for Honma and Hagino's analysis of changes in the three countries' agricultural exports. REFERENCES AkiYama,T, Baffes,J., Larson, D. and Varangis, I?,2001. Commodity Market Reforms: lessons of two decades, World Bank, Washington, DC. Bates, R.H., 1981. Markets and States in Tropical Afica: the political basis of agricultural policies, University of California Press, Berkeley Deininger, D. and Olinto, l?,2000. Asset Distrt'htion,Ineqzldli$ and Growth, Policy Research Paper W S 2375, World Bank, Washington, DC. Deininger, K. and Squire, L., 1997. 'Economic growth and income inequality: re-examining the links', Finance and DeveLopmt, 34(1):38- 41. Food and Agriculture Organization of the United Nations, n.d. FAOSTAT (FA0 Statistical Database), Food and Agriculture Organization of the United Nations, Washington, DC, http://apps.fao.org/. Gallup, J., Radelet, A. and Warner, A., 1998. Economic Growth and the Income of the Poor, CAER I1 Discussion Paper No. 36, Harvard Institute for International Development, Cambridge, MA. Griffin, K. and Ickowitz, A., 2000. 'The distribution of wealth and the pace of development', in K. Griffin (ed.), Studies in Devebpment Strategy and Systemic Transformation,MacMillan Press and St. Martin's Press, New York:66-9 1. Hayarni, Y., 2001. Development Econornics:fiom the poverty to the wealth of nations, 2nd ed., Oxford University Press, Oxford and New York. -andRuttan,W.W.,1985.AgriculturalDevelopment:aninternational perpective, Johns Hopkins University Press, Baltimore. Johnston, B.F. and Mellor,J.W., 1961. 'The role of agricultureand economic development', American Economic Review, 5 1(4):56693. Jorgenson, D.W., 1961. 'The development of a dual economy', Economic Journal, 71(282):309-34. Larson, D.F., Crego, A., Butzer, R and Mundlak, Y., 2000. 'A cross-country database for sector investment and capital', World Bank Economic him, 14(2):371-91. 70 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Lipton, M., 1977. WhyPoor People Stay Poor, Australian National University Press, Canberra. Mundlak, Y., 1986. Agriculture and economic growth: theory and measurement, lecture notes, University of Chicago, Chicago. -, and Larson, D.F., 1992. 'On the transmission of world agricultural prices', The World Bank Economic Review, 6(3):399422. -, Larson, D.F. and Butzer, R, 1997. The Deteminants of Agn'cultural Production: a cross-counq amlyis, Policy Research Working Paper No. 1827, World Bank, Washington, DC. Myint, H., 1965. The Economics of Developing Countries, Praeger, New York Oshima, H.T., 1987. Economic Growth in Monsoonh, of Tokyo University Press, Tokyo. Ranis, G. and Fei, J.C.H., 1961. 'A theory of economic development', American Economic Review, 51(2):533-65. Ravallion, M. and Datt, G., 1996. 'How important to India's poor is the sectoralcomposition of economic growth?', World Bank Economic him, lO(1):l-25. Rodrik, D., 1994. 'King Kong meets Godzilla: the World Bank and the East Asian miracle', in A. Fishlow, C. Gwin, S. Haggard, D. Rodrik and R Wade (eds), Miracle or Design? Lessonsfrom the Eat Asian expmience, Policy Essay No. I I, Overseas Development Council, London:13-54. SchiE, M. and Valdes, A, 1992. The Political Economy ofAgn'culturalPricing Poliq, vol. 4, a Synthesis of the Economies in Developing Countries, Johns Hopkins University Press, Baltimore. Stern, N., 1996. Growth Theories, Old and New and the Role ofAgngnculture in Economic Development, Food and Agriculture Organization of the United Nations, Rome. T i e r , C. E, 1997. How Well Do the Poor Connect to the Growth Process?, CAER I1 Discussion Paper No. 17, Harvard Institute for International Development, Cambridge, Massachusetts. -, 1988. 'The agricultural transformation', in H. Chenery and T.N. Srinivasan (eds), Handbook of Development Economics, vol. 1, North- Holland, Amsterdam:275-332. Watkins, M., 1963. 'A staple theory of economic growth', CadianJouml of Economics and Political Science, 29(2):141-58. World Bank, 2001. World Development Report 2000/01: attacking poverty, Oxford University Press, New York GROWTH OF THE AGRICULTURAL SECTOR 71 -, 1997. Rural Development: from vision to action, World Bank, Washington, DC. -, 1996. a b a l Economic Prospects and the Developing Countries 1996, World Bank, Washington, DC. -, various issues. Global Economic Prospectsand the Developing Countries, World Bank, Washington, DC. -, n.d. World Development Indicators Online, World Bank, Washington, DC. Available online at http://worldbank.org/data/onlinedatabases/ onlinedatabases.htm1. APPENDIX-DEFINITIONS &cultural area or agriculturalland is arable land and land under permanent crops. Agricultural GDP is drawn from a database at the World Bank. The base year is 1995. Agricultural labour is defined as all persons depending for their livelihood on agriculture, hunting, fishing or forestry. This estimate comprises all persons actively engaged in agriculture and their non-working dependants. &cultural production. Unless otherwise shown, the default multiplication factor for aggregate components is 1.00. Exceptions are displayed where they exist. The base year is the average from the period 1989-91. Pasture is permanent pasture (thousands of hectares) in the FA0 element 131. It is land used permanently (for five years or more) for herbaceous forage crops, either cultivated or growing wild (wild prairie or grazing land). The dividing line between this category and the category 'Forests and woodland' is rather indefinite, especially in the case of shrubs, savannah and so on, which may have been reported under either of these two categories. For 1995 and all later years, there will be no data for this element. Rice area is rice paddy o y a spp., mainly o y a sativa. Rice is grain after threshing and winnowing, also known as rice in the husk. Data refer to the area from which a crop is gathered. This harvested area, therefore, excludes the area from which, whether sown or planted, there was no harvest due to damage, crop failure, and so on. It is usually net for temporaq crops and gross for permanent crops. Net area differs from gross area insofar as the latter includes uncultivated patches, footpaths, ditches, headlands, shoulders, shelter belts and so on. 72 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Rice production refers to actual harvested production from the field or orchard and gardens, excluding harvesting and threshing losses and that part of a crop not harvested for any reason. Production therefore includes quantities sold in the market (marketed production) and quantities consumed or used by producers. When the available production data refers to a production period falling in two successive calendar years and it is not possible to allocate the relative production to each of them, it is usual to refer production data to that year in which the bulk of the production falls. Crop production data are stated in metric tons (MT). DETAILS ON OBSERVATIONS The following information describes the data used to construct the tables in this paper. A g r i c u l d GDP: 59 observations IPT-Indonesia, Philippines, and Thailand Eathid-China, Fiji, Indonesia, South Korea, Malaysia, Papua New Guinea, Philippines, and Thailand. Southeathid-Indonesia, Malaysia, Philippines, and Thailand. South hid-Bangladesh, India, Nepal, Pakistan, and Sri Lanka. Southern and Eat Africa-Burundi, Democratic Republic of Congo, Kenya, Madagascar, Malawi, Mauritius, Rwanda, South Africa, Zambia, and Zimbabwe. West Africa-Benin, Burkina Faso, Central African Republic, Chad, Republic of Congo, Cote d'Ivoire, Gambia, Ghana, Guinea-Bissau, Mali, Mauritania, Niger, Nigeria, Senegal, Sierra Leone, and Togo. Ltin Ame~icd-Argentina, Belize, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, Guatemala, Haiti, Honduras, Jamaica, Mexico, Nicaragua, Paraguzy, Peru, Trinidad and Tobago, Uruguay, and Republica Bolvariana de Venezuela. Rice: 57 observations As for Agricultural GDP, but excluding Zambia and Mauritius Trade: 76 observations As for Agricultural GDP, but excludingChina, Fiji, Papua New Guinea and Zambia and including Mozambique, Sudan, Swaziland, Tanzania, Uganda, Cameroon, Gabon, Guinea, Sao Tome and Principe, Bahamas, Barbados, Bolivia, Cuba, Dominica, Grenada, Panama, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, and Suriname in Agricultural GDP. GROWTH OF THE AGRICU1,TURAL SECTOR 73 Agricultural capital: 27 observations Exclzlding China, Fiji, Malaysia, Papua New Guinea, Bangladesh, Nepal, Burundi, Democratic Republic of Congo, Rwanda, Zambia, Belize, Brazil, Ecuador, Haiti, Mexico, Nicaragua and Paraguay from Agricultural GDP Inclzlding Tanzania, Sudan, Swaziland, Tanzania, Uganda, Cameroon, Gabon, Guinea, Sao Tome and Principe, Bahamas, Barbados, Bolivia, Cuba, Dominica, Grenada, Panama, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines and Suriname in Agricultural GDP Note that the Thailand observation uses fured capital and there is no West Africa. Border price: 75 observations Exclzlding Fiji, Democratic Republic of Congo, Gambia, Argentina, Brazil, Mexico, Nicaragua, Peru, Uruguay, and Guinea-Bissau &om Agricultural GDP Inchding Brunei, Lao PDR, Myanmar, Afghanistan, Botswana, Mozambique, Nambia, Reunion, Sudan, Swaziland, Tanzania, Uganda, Cameroon, Gabon, Guinea, Liberia, Dominica, Guyana, Panama, and Suriname in Agricultural GDI? Note that otherwise, observations are the same for agricultural GDP in pasture, agricultural land, and agricultural labour. DO DIFFERENT REGIMES DISTORT Maria Amelina In the wake of the 1997 East Asian crisis, the 'miracle' of East and Southeast Asian development has been reassessed. Many Asian states were, to borrow the metaphors of Peter Evans, downgraded from successful 'midwives' of development to 'predators' that promoted crony capitalism (Evans 1995). However, as Evans has suggested, the crisis should not lead us to a wholesale disregard for three decades of very substantial economic achievement made by states in the region (Evans1998). Rather, it should be taken as an occasion to try to understand with more precision and country-specific discretion 'what features of the multifaceted East Asian institutional context might constituteprerequisitesfor successful policy implementation' (Evans 1998:67). This chapter explores one dimension of this institutional context-the relationships between political regimes, institutions of agricultural policymaking, and economic outcomes. The exploration of these relationships traverses three decades (mid 1960s-mid 1990s) of evolving commodity policy, specifically, sugar and rice policies in three Southeast Asian states: Indonesia, the Philippines and Thailand. For all three countries, sugar and rice are key economiccommodities, and rice is also an emblematic and subsistence commodity. This chapter seeks to understand the effect of relatively pluralistic versus relatively authoritarian political structures on the more developmental versus more predatory forms of sectoral governance. The relationships between regime type, economic reform, and economic outcomes have received much attention in recent years. Some authors have argued that authoritarian regimes may be more successful at implementing reforms than democratic regimes because they are able to push through unpopular and ~ainfulstabilisation measures (Lipset 1959; O'Donnell 1973). Others have argued that such an 'authoritarian bargain' is ultimately unstable institutionally, and have questioned the ability of authoritarian states to deliver on the promise of accelerated development (Haggard and Kaufman 1995; O'Donnel and Schmitter 1986). The evidence presented here supports the latter position and attempts to enrich it by examining specific institutions of sectoral governance in detail. It challenges the argument that authoritarian regimes are good for producers, pointing to the decrease in incentives to innovate and increase efficiencythat sterns from the lack of control over sector-generated resources observed in the two more authoritarian of the three countries. It is argued that in more pluralistic settings-here, Thailand-control over sectoral governance tends to be distributed between the producers and the state in a way that prevents both extreme predation on the part of the state bureaucracy and excessive extraction of rents by the sector. By contrast, in more authoritarian settings-here, Indonesia and the Philippine+control over sectoral governance tends to be concentrated in the hands of executive power holders. As a result, agricultural policies suffer from state predation, while the limited access of sector insiders, here the producers and traders, to sectoral policymaking stunts the entrepreneurial drive for innovation, cost saving and diversification. The central contention, then, is that the divergent outcomes in the cases presented are best explained not by the scale of distortions, but by the spea9c institutiomL conjgurations of resource dzStribution between the sector and the economy as a whole. POLITICAL ECONOMY OF SECTORAL GOVERNANCE At the beginning of the studied period, institutions of sectoral governance in the three countries appear quite similar-a mix of protection for industry and sporadic efforts at modernisationof agricultural production. In all three cases agricultural production was the main source of budget revenue. The key difference lies in the configuration of power structures that influenced the modality and depth of state interventions in agricultural production and trade. In Indonesia and the Philippines, control over agricultural production and distribution has been ideologically framed as the main pillar of development-based legitimacy of the ruling authoritarian regimes. The 76 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH proclaimedimperative of rice self-sufficiencywas used to justii concentration of trade in an ever growing number of agricultural commodities under the auspices of non-transparent parastatals controlled by the immediate entourage of the president. In the name of rice self-sufficiency, governing bureaucrats imposed cross-subsidisationof some commodities at the expense of others, mandated production of a commodity (sugarcane in Indonesia), decreed export monopolies, and channeled exorbitant trade margins to a narrow group of preferred traders. The result was the continued exclusion of keysector actor-rowers and millers, as well as the majority of trader* from sector management, and the evolution of commodity-trading parastatals into rigid rent-generatinglideologicalstructures incapable of responding to the changing market environment. The state not only chose winners, but aha arbitrdrily created losers. In this unpredictable environment, incentivesfor ejTciency gains and divms$cation were limited and ahelopment of commodities considered previously k~ for the economies of both countries- sugar in particuhr-had been stunted. By contrast, in Thailand the institutions created for state-managed trade in agricultural commodities were much less invasive and encompassing than in the other two countries. The smaller scale of state intervention and the flexibility of institutional arrangements that organised agricultural production and trade in Thailand here are attributed to two factors. First, the traditional authority of the Thai monarchy required no legitimising. Thus, development projects did not carry the additional burden of ideological tools, and non-transparent state-run agricultural parastatals promising quick development and poverty alleviation at no-matter-what cost were never created. In this context, as the current chapter demonstrates, the regime underwent a process we call 'pluralisation by accrual' because sector insiders were granted increasing power in agricultural policymaking. Over time, producers found themselves operating in an increasingly transparent and predictable economic environment. Government interventions-for example, price supports-have not suppressed the entrepreneurial activity of producersand millers. ThisenvironmentharjGstered an increasingly eficient, increasingly productive, and increasingly diverse agn'cultural sector. Second, the dual track of economic policymaking described below constrained the ability of agricultural producers and traders to attract disproportionatelylarge flows fiom other sectors or commodities. This fiscal arrangement effectively constrained excessive allocation of public resources to a articularproduction or trading group. As an outcome, when agricultural development schemes proved too costly or otherwise ingective, they were allowed to fail and were rcplaced by arrangements more in tune with the changing economic environment. It may be argued that differences in the development of rice and sugar policiesare commodity-specificand critically depend on the roles ofThailand as a rice exporter of long standing and of Indonesia and the Philippines as rice importers.We contend that thisdoesnot adequatelyexplain the divergent dynamics of the incentives to produce observed in the three countries. In the beginning of the period studied, sugar was an import commodity for Thailand and an export commodity for the other two countries. By the end of the period, Thailand emerges as sugar exporter, Indonesia retreats to the position of importing sugar, and the Philippine position as sugar exporter is geatly undermined. Yet, the dynamics of institution building in sugar production and trade parallel the structures of rice production on a counuy- by-country basis-state monopolisation of trade in the Philippines and Indonesia and relaxed controls in Thailand. Thus, it is important to look at the regime dimension of these two stories and extract institutional lessons about the costs and benefits of an authoritarian bargain. DEVELOPMENT OF NATIONAL GOVERNANCE AND POLITY The pivotal events that determined the divergentdevelopment of governance structuresamong the three countriesvary. In Thailand,these were modernking projects pursued by the Thai monarchs. In Indonesia, the study starts with Suharto's accession to power and the evolution of a rice-based development project central to the legitimacy of his government. In the Philippines, we begin with the dissolution of the post-colonial democracy and establishment of Marcos' rule, which corresponded to the loss of power by sugar interests and the advent of centralised trading arrangements for sugar and rice. In each case, two dimensions of the political economy of sectoral governance are examined. The first is the form of legitimising state power, and the effectthese different forms have on the evolution of relatively more versus relatively less independent technocracies. The second is the relationship between the institutions of sectoral governance and the broader configuration of political power, specifically the inclusion versus exclusion of agricultural producers in decision making about the direction and shape of agricultural policies. 78 RURAL DEVELOPMENT AND AGRICUCTZTRAL GROWTH Thailand This section traces the successes of agricultural development, not to a well thought out developmental design, but to the mix of incentives and constraints provided by the evolution of both national polity and agricultural policies. In Thailand the development of a more inclusive political dialogue has led, in Philippe Schmitter's definition, to gadual democratic consolidation, and in terms of economic policies, to active substantive participation of an increasing number of interests in policy formulation (Schmitter 1992). Agricultural policies, for their part, have been shaped as a compromise between intra-sectoral interests, precluding both predation by non-sectoral actors and excessive rent seeking by the sector itself. Legitimacy of national governance. Uncharacteristically, the modernisation of Thailand was brought about by mobilisation around its traditional rulers. As Sar Desai (1997:139) observes, 'Thailand was undoubtedly the most fortunate of the Southeast Asian countries in having these two statesmen [Kings Mongkut (1851-68) and Chulalongkorn (1868-19lo)], who...introduced timely reforms on the domestic front, well calculated to preserve Thailand's identity and independence'. Modernisation was seen as means of preserving international competitiveness as well as securing independence in the face of European expansion in Asia (Wyatt 1984). By drawing up a constitution, the monarchy overcame the limitations of its role and emerged as embodiment and guardian of continuityand stability. However, the role of the monarch goes far beyond the limits prescribed by the constitution. As Samudavanija (1990:303) observes, '[tlhe most le itimate institution, which has geatly contributed to social and political g stability in Thailand, is the monarchy...While politicians, military leaders, and civilian prime ministers had come and gone, the king has remained the head of state, the focus of the people's loyalty and cohesion, the fount of legitimacy'. The monarch is removed from day-to-day political decision making but intervenes at times of national crisis, as occurred in 1973 and 1992, when political instability went beyond fluctuations deemed acceptable for national harmony. In fact, a chronology of political events in Thailand seems to describe anything but 'acceptable fluctuations'. Thai ~oliticalhistory of the past half century has been dramatically punctuated by nine successful and nine abortive coups &tat, 16 constitutions, 30 revisions of constitutions, and 87 g,eneral elections (Dixon 1999:260).However, most dose observers of Thai - politics warn against taking these numbers at face value. Assessments of the Thai coup d'etat vary from 'merely a way that Thailand changes its government' (Dixon 1999) to a 'struggle of rival bureaucratic leagues to gain control' (Riggs 1966) and a more specific 'government turnover' that 'meant little for the masses because it was essentially a regroupment of a narrow elite of about five thousand stru lin for reallocation of the spoils gg g of the office' (Sar Desai 1997). None of these assessments see coups d'ktat as events that seriously affect the fabric of Thai society (Muscat 1994). Pluralism by accrual. One consequence of the relative security of the kings authority is that increasing pluralism has not constituted a threat to regime stability or continuity. During the ~eriodstudied, civil society 'became more organised, more complex, and more vociferous in its demands' (Dixon 1999). New interests were graduallywoven into the political discourse and new civil, political, and economic institutions were created to accommodate these increasingly diverse interests. For example, in the 1970s, trade associations proliferated, becoming instruments of political pressure on the government. In the years of democratic rule from 1973 to 1977, farmer associations began to play a role as an organised force capable of affecting resource distribution. By the 1980s, businessmen were no longer lobbying the government, since they had become part of it-almost one-half of the Prem government came from business circles (Laothamatas 1992). The thickening of counterbalancing interests led to the gradual democratisation of both social and political institutions. According to Ockey (1996:348), '[dluring the last two decades, patron-client ties have become less certain and more instrumental, and the number of those seeking and expecting patronage from the government has risen with the increased participation by new political actors, such as provincial business leaders, NGOs, and students'. The fact that such a major upheaval as the 1997 financial crisis was not followed by a coup d'ktat can serve as an optimistic sign that 'pluralism by accrual' is taking hold and that the national political discourse is becoming flexible and pluralistic enough to withstand political and economic shocks. In the words of Samudavanija (1990:305), '[ilt is fair to conclude that a dynamic balance is currently maintained among various forces, each of which cannot possibly afford to dominate the political process on its own strength alone'. Dual economic policies. Like the national political structures, the economic policies of the Thai government have been characterised by a duality of stability and continuity at the macro level and fluctuations and brevity at the sectoral level. A distinguishing feature of Thai macroeconomic 80 RURA12 DEVELOPMENT AND AGRICULTURAL GROWTH policymakingis that the main fiscal and monetary agencies have been legally and institutionally set apart from the political process (Christensen 1993). Transparent and effective macroeconomic policies contrast, however, with the opaque sectoral policies controlled by political parties. As in the story about the elephant that is described differently depending on whether the blindfolded observer touches its trunk or its leg, the dual nature of Thai economic polices makes researchers draw different conclusions about their success depending on which subset of policies they examine. The World Bank, praising Thailand's consistently conservative fiscal and monetary policies and the institutions that sustain them, declares that 'a small technocratic core, insulated from politics, can set a positive tone for the entire economy' (World Bank 1993:172). Unger, who examines Thai social and economic interactions from a sector-specific perspective, states that '...administrationin Thailand sufferedfrom adisjuncturebetween planning and implementation' (Unger 1998:77), then concludes that this disjuncture was to the country's advantage, since no group had the power or the institutions to appropriate all benefits (TJnger 1998:78). Unger compares favourably the pluralistic pattern of power distribution in Thailand with the authoritarian pattern prevalent in other Southeast Asian countries: 'In Thailand on-going, sharp intra-Clite competition prevented the entrenchment of favoured interests to the degree evident in Indonesia, Malaysia and the Philippines' (Unger 1998:79). Christensen's comprehensive multi-level analysis of Thai political structures draws a picture of the rules of the game governing macro and micro policies. Thai macroeconomic policies are designed and managed by highly skilled and better paid technocrats employed by such agencies as the National Economic and Social Development Board (NESDB), the Bureau of the Budget, the Ministry of Finance and the Bank of Thailand (Christensen and Siamwalla 1993). By contrast, sectoral ministries, specifically the Ministries of Agriculture, Industry and Commerce, are managed by political appointees with shorter time horizons representing the dominant or a compromise interest within the sector.While in ofice they are expected to deliver patronage rents to their party's supporters at the local level andlor to specific interests within the sector. Commodity policies are managed by sector ministries, provided the commodity does not play a formative role in the national budget. Ministerial access to h d s for commodity-specific use is constrained by rigid allocation of resources by nationaleconomicagencies, but thesector ministrieshave broad discretion to distributethese funds to actors within the industry.' Hence, sector policies 'are not very transparent, since they are managed through administrative decisions largely in the bureaucratic arena; they lend themselves naturally to the kind of non-collective ad-hoc strategies the politicians are encouraged in the first place to pursue by their supporters' (Christensen 1993:23). Ministerial control over the use of resources generated within the sector precludes the diversion of sector-specific revenue to outside interests. At the same time, producers and traders within the sector are limited in their ability to attract budgetary resources to subsidise production and trade in a particular commodity. These two sets of constraints narrow the band of possible distortions in both the production and trade of a commodity. Indonesia Unlike in Thailand, political institutions in Indonesia have developed along a path of increasing autocratic centralisation. Rather than allowing a range of groups into the political process and sectoral policymaking, the tendency has increasingly been to fuse the political and economic functions of the government. As a consequence, the autonomy of macroeconomic agencies has been substantially undermined, and sectoral actors have been largely excluded from sectoral governance, which is dominated by non-sectoral bureaucracies with the power to divert resources from the sector for political or rent-seeking reasons. Legitirnisation of national governance. In contrast to the case of Thailand, which lacked a colonial history, in the Indonesian case a central problem faced by country's second-generation post-colonial ruler was one of regime legitimacy. In 1966, Suharto wrested national leadershipfrom his charismatic predecessor Sukarno. Suharto's inst:allation had very little in terms of legitimacy. This was not a hereditary power of the kind that made the supremacy of the kings of Thailand unquestionable, nor was it Sukarno's charismatic power as the leader of the anti-colonial struggle. Suharto's accession to power was preceded and accompanied by a wave of violence that began with an abortive military coup and escalated into one of the largest uninvestigated massacres of the twentieth century. Hundreds of thousands of people branded Communist syrnpathisers peri~hed.~At the end of the coup, the little known and personally uninspiring commander of the army's strategic reserves (KOSTRAD), General Suharto, was first promoted to Commander-in-Chiefand then made President of the Republic of Indonesia. General Suharto was sufficiently concerned with formal 82 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH legitimisation, however, to postpone his official seizure of power and allow Sukarno, from whom actual power was taken away in March 1966, to nominally remain president until March 1968, when the People's Assembly granted Suharto the title of President. The legitimacy of the new regime was based on the promise of development. 'Suharto and his men...quickly saw that solid political legitimacy could only be achieved by putting the country back on its feet' (Vatikiotis 1998:33). Suharto's government, reinforced by neoclassically trained economists, drew up an ambitious plan to transform the economy from command to market, its aim being to 'to remove numerous distortions from the domestic price structures and eliminate a major source of bureaucratic and political corruption' (Bresnan 1993:65). Clientelism in action. Despite high-level pledges to liberalise the Indonesian economy, political and ideological constraints prevented decisive marketisation, and led to forms of administration beneficial to a specific clientele. Politically, the market-oriented development program soon came to clash with the interests of the most powerful interest group in the countly-the military? The differencein the role of the militaryin Indonesia fiom the other two study countries is that in Indonesia the armed forces are officially part of the political process. It is not the case of civilian politicians supervising the army, but rather the army making sure that the politicians are operating within agreed-upon parameters. During the Sukarno rule, the influence of the military was counterbalanced by participation fiom other major forces in the political process, the Muslim parties and the Communist party. After the demise of the former and the physical annihilation of the latter, the military became more prominent and more involved both overtly and covertly in political and economic decision making.4 Direct participation in the political process allowed the military to create officialinstitutions that organised access to economic rents. Considering the agrarian nature of the society, it is not surprisingthat at the earlystages of the Suharto regime the rents of choice were extracted from distorted commodity made. One of the moststrikingexamples ofdevelopmentof clientelestructures of economic governance was the establishment of high-profile parastatals, which have rapidlygained power over manyaspectsof agriculturaldistribution. The traders that became exclusive operators for BULOG, the national commodit ~arastatal,were initially army quartermasters, the most y prominent of whom had ~rovisionedthe troops of Colonel Suharto. The former quartermastersthat constitutedthe exclusive club of BULOG's traders were predominantly Chinese (Vatikiotis 1998).Because they were excluded from conventional politics, their de endence on the new rulers was both p political and economic. In a patron-client relationship, they were an ideal client because their politically disadvantaged position, which Suharto made no effort to improve, precluded them from changing the relationship from hierarchical to more participatory. The political and economic incentives that the military had for interfering with the proclaimed market environment were reinforced by an organisational structure for carrying out these distortions. Technocrat markets vs. rent-seeker rents. The military-entrepreneurial rent- seekingsystem-with its direct access to power and well organised logistics- became a formidable opponent to the market-oriented form of development promoted by technocrats.Thus, 'there developed a permanent tension between the demands of economic planners on the one hand, and military officials in charge of the operationson the other, over the control of the state funds, and exercise of the economic power of the corporations' (Robinson 1986:228). Unlike that of their Thai counterparts, the institutional position of Indonesian technocrats was not well defined. In practice, they had neither a power base nor institutional independence and had to rely on Suharto's willingness to listen to their advice. The role of technocrats was visible and significant in the beginning of Suharto's rule, when the National Development and Planning Board (BAPPENAS) developed and implemented plans to extract the country from the post-Sukarno economic crisis. However, once macroeconomicstabilisation had been achieved, and particularlyafter increased oil revenues made fiscal prudence seem less vital, the technocrats lost Suharto's ear to two groups-the so-called technologist, nationalistically minded group that promoted import substitution, and bureaucratic business interests that counted on the national budget to finance their ventures (Cribb and Brown 1995; Bresnan 1993). The Bank of Indonesia did not have the independence of its Thai counterpart, and it became famous for financing the National Logistics Agency (BULOG) and infamous for financing Pertamina, the national oil company that fell into debt in the middle of an oil boom and had to be bailed out with $10 billion in the mid 1970s. By the 1990s, technocrats seemed to have been almost completely marginalised. In the words of one of the more prominent ministers of the New Order, '[uIp to 1976, [when ministers met with the president], he used to take notes; &er 1976 the ministers were made to sit and take notes'.5 84 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH The Philippines The limited democracy of the Philippine post-colonial period was followed by a consolidation of authoritarian rule that lasted until democracy was reinstated by the 'people power' revolution of 1986. The authoritarian government strengthened the hand of parastatals in sectoral governance and, as in Indonesia, effectively excluded actors within the sector. Although the people-power democracy widened the circle of interests included in economic discourse and decision making, the reinstatement of certain pre- Marcos power holders resulted in a democracy with unbalanced representation. At the time of the study, institutions of post-authoritarian agricultural governance had progressed enough to dismantle rent-seeking parastatals but not enough to include diverse peasant interests in political decision making or resolve acute outstanding issues of agriculturalinstitution building, including the critical issue of land reform. Transformations of national governance. During the post-colonial years, the Philippines has oscillated between authoritarian and ~luralisticpolitical interactions. Benedict Anderson famously referred to the government in power between 1954 and 1972 as 'a cacique democracy', after the group of mestizo Filipino-Chinese that acquired economic wealth under Spanish colonialism and consolidated political power by taking the lead in the two- party system created during the American rule (Anderson 1998). This American two-party model was adapted to clientelistic Philippine power structures, with representatives elected to satisfy the economic interests of their clientele and political leaders' failure to deliver leading to switching party allegiances. The circle of benefiting interest groups was narrow: 'A traditional landed elite dominated the Congress. A backward and inefficient but politically assertive sector of highly protected industrialists played a strong secondary political role' (Overholt 1987:90). The fluid nature of political alliances made the democratic arrangement - an inefficient and increasingly unpopular form of decision making, paving the way for the seizure of power by Marcos. Though Marcos himself was a civilian, the regime had all the trappings of a military dictatorship. Marcos's kleptocratic rule began with the suspension of political representation and introduction of martial law in 1972. Initially, the coup that brought Marcos to power met no active opposition.The nascent middle classes were tired of -- the corrupt two-party power of the cacique oligarchs, which had deprived them of a meanin@ political role. There were expectations that autocratic rule would bring about stability and economic gain. Technocrats in the state apparatus expected to be able to carry out unpopular macroeconomic reforms that could not be passed through the democratic Congre~s.~Peasants expected land reform to be completed. Export-oriented businesses expected a more advantageous exchange rate, increased political stability, and longer planning horizons. Such expectations were dashed for all concerned, with the possible exception of the military.In placeof a more stableand reformist environment, Filipinos found themselves in the hands of a small group of rent-seekers dose to the president. National GDP turned negative, national debt grew, real incomes fell, and income inequality increased (Ranis 1987; Balisacan 1992). Authoritarian legithisation and agricultural outcomes. In the context of these dashed expectations, the legacy of decades of democracy made legithising the new regime much more arduous than in Indonesia, and ultimately much less su~cessfd.~Efforts at legitimisation-enactment of a new constitution, the referendum on martial law, efforts to create a more manageable National Assembly-frequently had the effect of mobilising the opposition and exposing the coercive nature of the regime rather than increasing its legitimacy. The other legitimisation path was that of economic development. Here again success turned into failure much more quickly than in Indonesia. The technocrats leading the economic reforms and providing credibility to the reform program in the eyes of international donors soon came to realise that their plans for economic austeritywere secondaryto the financial claims of the Marcos-sponsoredinterests. Plans designed by the National Economic and Development Authority (NEDA) were not followed. Protests by NEDA technocrats failed to stop the creation of the Ministry of Public Works, which was a thinly veiled mechanism for 'financing a set of cronies that ran private monop~lies'.~Contrary to the protests and resignation of Gerardo Sicat, the first NEDA Director General, the hnctions of the National Food Authority (NFA) were expanded to include wholesale and retail trade in almost any commodity, to the h h e r detriment of both fiscal stability and the market position of small traders (Sicat 1986). Those technocrats who chose to stay, such as Prime Minister Virata, justified their participation in the government as a search for a window of opportunity to promote a reformist agenda, rather than a way of influencing economic policies through 86 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH institutionalised channels (Tolentino 1998). Thus, the technocrats continued to perform a legitimisation role in interactionswith donor agencies although they lacked the capacity to consistently implement or meaningfully influence macroeconomicand fiscal policies. Legitimisation by agricultural reforms at first appeared to bear fruit. The two main agrarian reforms that the regime promised to carry out better than a democracy were land reform and large-scaleinvestmentsin the Green Revolution. The populist goal of these reforms was to ensure national rice self-sufficiency.Success at land reform was at best marginal, and its scope was narrow because it excluded plantations and export crop lands (Department of Agricultural Reform 1983; Balisacan 1990). The investment in the Green Revolution also failed to deliver the expected returns to peasants. Masagana 99, a widely advertised nation-wide program of subsidised agricultural credit, successfdy delivered fertiliser and seed of high-yieldingvarieties to farmers, thus increasingyields. It was not successfd, however, in creating a sustainable system to finance lasting improvements to farmin practices (David 1984). g As a result, farmers found themselves in debt, and either stopped using high-cost inputs or used their input subsidies for other purpose^.^ Democratic contestation. In 1986, the bloodless people-power revolution swept away the Marcos dictatorship. Interpretations of Aquino's democracy run the spectrum fiom the pejorative'cacique democracy,' to the affirmative 'institutional democracy', successful in consolidating multiple interests (Kasuya 1995). For our inquiry, it is important to note that the first legal action of the new government was to return control over commodities to a ricultural producers-state monopolies in sugar, coconut, beef, wheat, g and fertiliser were abolished.'O The NFA lost its monopoly over a number of agricultural and non-agricultural commodities and was reduced to its pre-martial law functions of rice and corn price stabilisation. While Aquino's economic ministers could finally implement economic reforms, they were not adept at 'selling' their programs to the Congress. The above-mentioned reforms took place before the re-opening of Congress in 1987. It took years for the Congress and the government to learn to interact in a collegial and constructive manner." In order to involve ~roducersin ~oliticaldecision making in agriculture, a number of influential committees have been created which by law require privatesector representation.The most prominent are the PresidentialCouncil on Rural Development, the Coordinating Committee on the Uruguay Round, and the Philippine Exports Council. The history of slow progress DO DIFFERENT REGIMES DISTOR'r DIFFERENTLY! 87 in land reform proves that some interests continue to be more equal than others, yet the scope of groups involved in shaping policies has been widened to include a much broader range of interests in a meaningful way. COMMODITY-SPECIFIC POLICIES With the relevant political and economic frameworks in mind, we turn to the dynamics of rice and sugar policies in the three countries. The rise and fall in the production of a commodityis linked to the economicenvironment and production horizons allowed within the framework of different regimes. Specifically, we will discuss institutional arrangements that serve both as a framework and as channels for divergent outcomes. In Indonesia and the Philippines, trade monopolies for a number of commodities were officially tied to the food security argument, which was then elevated to the argument of national security (Simatupang 1999). Ideologically, then, self-sufficiencyand adequate access to staple foods were treated as a priority to be achieved at almost any cost. Institutions were created to redistribute resources to expand rice production. The political importance of the goal precluded inquiries into the financial and economic costs of the redistribution in Indonesia and the Philippines. With time, the ruling group realised what a potent instrument it had created-and lack of transparency allowed the resources to be redirected to causes remote from the core mandate of commodity-trading parastatals.The most consistently non-authoritarian country of the study, Thailand, did not develop food- trading parastatals. All the parastatals claimed to be improving conditions for farmers, but very rarely did they deliver on these promises. When they did, the beneficiaries were usually rice farmers, and the benefits usually came less from price support given by trading parastatals than from input provision delivered by other organisations. Meanwhile,producersof other commodities were disadvantaged, being officially exploited to subsidise the ideologically superior rice crop. Losses from large-scale distortions in commodty trading stifled production. Development for freedom, the concept underlying the authoritarian bargain, failed to materialise. Thailand Developments in the rice and sugar sectors in Thailand over the past 30 years can be summarised by a few basic trends. First, the dependence of the national budget on revenue from agriculture, most prominently the rice 88 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH trade, has decreased and commodity-specificpolicies have been passed down fiom national economic to sectoralministries. Second,at the sectoralministry level, interventions were increasingly used to support producers and/or commodity consumers. Because resources allocated to the sectoral ministries from the national budget were tightly controlled by national economic agenciesand few outside resourcescould be attractedto support a commodity, agriculturalsub-sectors learned to rely on resources generated by commodity trade, redistributed between consumers and producers. Third, and in crucial contrast to Indonesia and the Philippines, institutions that channelled these policies were allowed to fail when they outlived their usefulness or could not be sustained. Over time, as agricultural actors gained influence in agricultural decision making, the position of the state has evolved from supervisor and policy announcer to that of mediator and partner in policy design. Rice.This section traces two tendencies in Thai rice policies that are crucial to understanding broader developments in the sector. The first of these tendencies is the gradual phasing out of direct taxation of the rice sector; the second is the failure of rice procurement arrangements to evolve into national monopolies in the rice trade. In the first case, distortions generated by the state have been decreasing, while in the second case the distortions were never significant because neither of the rival ministries with a stake in procurementof rice controlled enough resources to substantiallyskew prices. Over time, the role of the market in rice pricing and distribution has become more important and procurement agencies have been used for small-scale patronage. The historyof rice policies in Thailand is marked by a progressivedecrease in taxation and a progressive increase in the influence of rice producers and traders over policies. Throughout the mid 1960s, rice policies were determined by the vital role of rice levies in the composition of national revenue. As budget needs began to be covered by other taxpayers, rice levies were reoriented to serve the interests of actors within the sector. A range of different agencies representingvarious groups in the sector-hmers, millers and trader- to have a voice in sector ~oliticsand policy design. The e opposing goals of these players prevented any given participant from distortin sectoral policies in their favour. Consequently, during the period g studied, no interest group was powerful enough to appropriateall the benefits of the rice trade to the detriment of any other groups, nor were any groups able to attract national budget financing beyond agreed-upon levels. Phasing out rice taxes. Except for a short ~eriodafter the Second World War, rice exports have not been monopolised by the state despite their importance as a source of budget revenue.'' A number of different levies were mounted on forei n and domestic rice sales. A rice premium was g collected by the Ministry of Commerce (MOC), which surrendered the revenue to the national budget. The premium was an easily adjustable fiscal instrument because adjustments did not require parliamentary approval. Other direct and indirect rice levies accruing to the national monetary and financial bodies included revenue from the dual exchange rate set higher for rice traders, collected by the Bank of Thailand (BOT), a key instrument for building up national reserves in the post-war years (Silcock 1967) an ad valorem duty of 5 per cent accruing to the Ministry of Finance (MOF). All these levies were deposited with national fiscal and monetary institutions and used in a non-targeted way to cover general budgetary expenditure. The MOC retained control over some rice collections. State reserve deposits established in 1960 were rice deposits that exporters had to make with the fund at a below-market price. The reserves were used for two purposes for government-to-government exports for consumer price subsidies in times of rapid rice price increases. While taxingagriculture, the state was increasingly engaged in agricultural investment projects. Over time, agriculture moved from being a net donor to the budget to a net recipient. Siamwallaand Setboonsarng (1989) estimate that net transfers from the budget to agriculture turned positive as early as 1965. The beginning of the 1970s was marked by the democratisation of Thai society, culminating in the 1973 student revolt and installation of a civilian government with broad-based democratic appeal. Decreasing the rice premium as a means to increase farmer revenue was recognised and written into the Third National Economic and Social Development Plan 1972-76 (Suvarnsit 1971), with the position being promoted by intellectuals and farm organisations, particularly the Peasant Federation of Thailand, the first national peasant organisationto become a vocal and politicallyinfluential part of the political process. Rice procurement arrangements. From 1962 onwards, as the urban population increased in size and political influence, reserve rice began to be 90 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH used as a consumption subsidy. Interventions were carried out by the Public Warehouse Organization (PWO), subordinateto the MOC. With its limited staff (400-500 employees) and limited size of rice deposits, the program was often derailed by world rice price fluctuations. Having largely &led as a price stabilisation program, it succeeded as a small-scale patronage device for groups of traders who benefited both from sale margins and the opportunity to resell the lower-priced rice on the open market. Over time, the role of the private sector in managing the PWO has increased, institutionalising its rent-generating role rather than its price-stabilising role. Since 1992, private businessmen closely affiliated with the ruling party have replaced the MOC minister as chairman of the PWO board. As household incomes in the cities grew, the pressure to subsidise city dwellers at the expense of rice farmers decreased. In 1974, rice premium revenue started to be deposited with the Farmers' Aid Fund and used for the needs of rice firmers and the rice industry as a whole (Siamwalla 1987). The fund was administered by the Farmer's Aid Committee, chaired by the Secretary of the Ministry of Agriculture and Cooperatives (MOAC). The Marketing Organization for Farmers was set up under the umbrella of the MOAC to carry out interventions in the rice market to support producer prices if they fell below the floor price, as well as for general agricultural development. The scale of intervention largely depended on the size of rice premium collections as well as on the political orientation of the party in power. Like PWO, the Marketing Organization for Farmers was a relatively small organisation with 800-900 employees, 8-10 times less than in BULOG in Indonesia or the NFA in the Philippines. It was also poorly funded compared to the parastatals in Indonesia and the Philippines with their virtually unlimited access to funding.The lack of funding made it impossible for the Marketing Organization for Farmers to successfully handle interventions and significantly influence the market as its mandate demanded. Over time, the Marketing Organization's interventions started to be used as patronage subsidies for political clients (Siamwalla 1987). Thus, the country ended up with two rice procurement agencies, one affiliated with the MOC (namely, PWO) and the other with MOAC (namely, the Marketing Organization for Farmers). Depending on the party in power, one or the other was tasked with sales interventions. In both cases, sales were too small to affect the rice market beyond gainingfavour with the electorate. Currently, rice price interventions tend to be targeted and regional.13 Price interventions by the Marketing Organization for Farmers were discontinued when because of its scale the operation started to pose a threat to national budget targets. In 1981-82, the Farmer's Aid Fund, dissatisfied with the amount of revenue generated by the decreasing rice premium, borrowed 4.5 billion baht (US$2OO million) from commercial banks, a loan it could not repay. Sensing a threat to the fiscal targets, NESDB and the Bureau of the Budget moved in. The repayment was rescheduled, and the program was curtailed and passed into the hands of the Ministry of Interior (MOI), which started to manage an $8 million revolving fund and use these limited resources for targeted local operations. These interventions may have had both political and economic motives, but in either case the scale of the price distortion the interventions caused in the market was contained by their own modest scale. In the 198Os, rice ~roducers,farmers and millers, by now a significant political force, saw further increase in profit retention. Compulsory rice reserve deposits were abolished in 1982. In 1986 the rice tax was abolished and the rice premium suspended. Rice price stabilisation in the 1990s was achieved through a rice-pledging program whereby farmers deposited their harvest with bonded warehouses or received credits from the BAAC to store their crop themselves. This arrangement is believed to be more transparent and less political than the ones managed by the ministries. The net result is a rice sector that is increasingly less distorted and increasingly relies on world prices both domestically and internationally. The procurement scheme is too small to affect incentives to produce and trade at the national level. Sugar.The most strikingaspect of sugar evolution in Thailand is that despite very different institutional arrangements, the broader context of sectoral governance was ultimately quite similar to that of rice in its readiness to substitute new strategies for ones that faded and to include the main sector actors in commodity-specific policymaking. Likewise, the sugar industry realised a remarkablegowth and increasingefficiency over the period studied. The difference between rice and sugar policies is that the production of sugar, a relatively new major export commodity, has grown at a time of higher national per capita income. Consequently, revenue from the sugar trade has never been treated as a generic contribution to the national budget. Most arrangements in the sugar trade have been based on intra-indusuy subsidies between domestic ttade and exports. Sugar profits were distributed among the main sector actors-growers, millers and traders-in a way that 92 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH protected domestic prices yet did not stunt incentives to innovate, increase efficiency and expand sugar exports (Rarnsay 1987). Anyone who studies the history of sugar production in Thailand over the past four decades has to puzzle over one paradox-the sugar trade arrangements between the late 1950s and early 1980s were extremelyvolatile and often altered.14These were not marginal adjustments but radical changes in the mechanisms of protection and taxation that could 'convince one that the Thai government has never had a clear picture of what kind of sugar industry it wants in the long run' (Jessadachatr 1977). These policies include cross-subsidisation of domestic production through taxation of sugar imports, cross-subsidisation of exports through taxation of domestic production, export bans, quantitative restrictions on the production of raw and refined sugar, establishment of a domestic price stabilisation fund, and quantitative controls over production of raw and refined sugar. All of these schemes collapsed because they failed to sustain the quantitative or price targets set for them. Yet, despite this institutional instability and highly volatile and unpredictable year-to-year pricing arrangements, sugar production in Thailand 'experienced the strongest growth rate on record in the world- 10 per cent per annum over 30 years [between 1960 and 19901' (du Genestoux 1991239). Furthermore,Thai sugar producers evolved fiom one of the highest-cost producers in the world to one of the most cost-effective (du Genestoux 1991). Viewed from the perspective of state-industry interactions, the institutional diversity of the first two decadesof sugar exportingdemonstrated not only confusion but also a creative search for sugar policies acceptable to producers, consumers and the state. The fact that the system was allowed to change as soon as it stopped working and that more often than not no effort was made to prolong the life of unworkable solutions suggests that there w2.s no dominant player, and that the state had no political stake in propping up failed programs. Here we argue that the swifi development of the sugar industry was prompted primarily by a secure understanding between producers and the state that the revenue generated by the industry would accrue to economic actors within the industryand not be expropriated by the state. From the mid 1960s onwards, a close collaborative relationship between producers and the state was established with the inclusion of a powerful millers' lobby in sugar policy design by the MOI (Rarnsay 1987). As of the mid 1970s,the group of actors with a voice in decision makingwas expanded to includethe canegrowers,who by that t i e weresufficientlywell organised to participate in collective bargaining for distribution of benefits within the industry (Siamwalla and Setboonsarng 1989). With the introduction of a third party into interactions between the state and millers, the role of the state has changed from that of supervisor and enforcer to that of arbiter in intra-industry negotiations. A detailed description of the rapidly changing institutional arrangements in the sugar trade is beyond the scope of this work. Instead, we will provide some illustrations of the dual role of state agencies in designing and monitoring sugar policies and comment on the evolving balance of power between producers and millers and between producers and the state. The concept of state-led development promoted by Prime Minister Phibun (1948-57) was manifested in the creation of the Thailand Sugar Corporation (TSC),a 60 per cent state-owned organisation (Hewison 1986; Siamwalla and Setboonsarng 1989). The TSC was authorised to monopolise sugar imports, exports and domestic trade. The margin the TSC received from domesticsales of imported sugar was used to subsidise domestic millers. Fuelled by protected domestic prices, however, sugar production rapidly increased, causing a fall in imports and a resultant loss of income from domestic sales for the TSC. By the late 1950sthe TSC did not have enough funds to support domestic producers. The limited budgetary support was insufficient to maintain the floor prices, and the TSC's monopoly in the domestic market collapsed, bankrupted by its own success. With Thailand's adoption of export-led growth policies in the 1960s, the government's role in economic activities was curtailed. In this context, the demands of the MOC and the MOI on state financing for their sugar- producing clients were seen as destabilising for the budget. Thus, it was decided that resources for development would be generated within the sector. In 1961, the first National Sugar Act authorised the collection of a - levy on the production of refined sugar to subsidise expom,15to be managed by the Sugar Industrial Aid Fund. Imports of refined sugar were banned (Hewison 1986). As with the TSC-based system, the increase in production, combined with a slump in world prices, placed the subsidy targets beyond the fund's financial capacity. With the sugar industry heading into another crisis, the sector ministries (the MOI and MOC), pressured by millers and exporters, demanded a budget-financed bailout. Sensing a threat to national fiscal 94 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH targets, the National Economic Development Board (NEDB) stepped in. In contrast to developments in Indonesia and the Philippines, the NEDB's recommendations overrode sectoral rent-seeking considerations In 1966, on the recommendation of the NEDB, the sugar levy and export subsidy were abolished. The NEDB recommended that the technological capacity of mills be increased in order to bring production costs down and lessen the.need for subsidies. In a rare case of direct state support, the government allocated resources to R&D in the sugar industry. The new stabilisationscheme focused on protection of ceiling prices through consumer pricing control. Again, the fund collapsed shortly after being established because the resources committed to its goal proved insufficient. The new quantity restrictions on sugar production unravelled with the sharp increase in world prices in the early 1970s. The international sugar boom induced millers to smuggle sugar out of the country, triggeringsharp increases in domestic prices. As a more democratic government came to power in 1973, its main objective became consumer price subsidies. Thus, the premium on sugar exports was used to subsidise domestic sales and, unlike rice revenueof the 1970s,was channelled to sugar sector development rather than being transferred to the general national budget. With the increase in world prices in the 1970s, world exports became profitablewithout subsidies.The monopoly power of the?SC was challenged by powerful groups of millers,who claimed that theTSC did not adequately represent their interests. With the installation of a more democratic government in 1973, the export monopoly of the TSC was broken. New regulations allowed other parties to take part in sugar export. Under these new regulations, TSC and the Thai Sugar Trading Corporation (TSTC), representing a rival millers' group, came to dominate exports. Lnthe late 1970s,as in the rice sector, growers in the sugar sector emerged as an organised interest group. This affected the balance of power within the sugar industry and subsidy arrangements. Increasingly, the MOI found itself mitigating disputes between millers and growers.'G The lasting legacy of the National Sugar Act is the codification of the mediating role of the MOI in the so-called 70130 revenue-sharing agreement. Under this agreement, the state determines average recoverable sugar content by sampling.The monitoring bodies are tripartite committees comprised of representatives of the state, growers, and millers. The committees are present at big mills at the time of crushing. Efficient mills with a higher extraction content keep the difference because there is no averaging at the mill level. Each mill receives quotas for protected domestic contracts and exports under long-term contracts. Any production above the established quotas can be sold at the world market. The scheme is self- regulating and contains built-in incentives both for increased production efficiency and export expansion carried out by private exporters (ACIAR 1994).For the scheme to succeed, both growers and millers must trust the state to intermediate fairly within the agreed-upon framework. The system has remained in operation for 17 years. The development of the Thai sugar industry can be considered a spectacular success. Increases in both production volume and eff~ciency have occurred in an environment of steep border price fluctuations, unprotected by an exclusive foreign quota arrangement like that which defined sugar trading in the Philippines. As has been demonstrated, the main characteristic of Thai state intervention in the sugar sector was its pro- producer nature and its willingness to allow trade and other policy implementingagencies to fiil once they outlived their usefulness.This created space for institutional experiments and a search for solutions acceptable to all parties. Increasingly, the state came to act as a mediating rather than a prescriptive force. The state was ready to subsidise the development of the sugar industry through high consumer prices, but not through budgetary allocations. However, at no juncture has the state tried to expropriate revenue from the industry or take advantage of its privileged position as an informed arbiter. The long-term predictability of the state position in an otherwise unstable sector has served as a guarantor for the continuity of the rules of the game, a factor that has proved important to the success of the industry. Indonesia The evolution of commodity policies in Indonesia has two main features. First, the effort to increase rice production and the well-being of rice farmers was central to Indonesia's development strategy. Second, this strategy was pursued with iron-fisted top-down control. The rigid extra-sectoral control over thesestructuresensured that theydid not evolvein tune with the changing needs of h e r s , but collapsed at times when economic crisis coincided with the political crisis of interests that supported the government as a whole. The institution that played the dominant role in rice and sugar policies was the massive parastatalBLTLOG (Bddtn UmanLogktik, or State LogisticsAgency). 96 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH BULOG. It is safe to say that BULOG is one of the highest-profile food procurement agencies in the world. It owes its reputation to 'success in storage, handling, and movement of large quantitiesof riceover the extensive geographical area of Indonesia' (Ellis 1990:48). While BULOG's rice management practices are internationally well known, widely cited by policymakers and frequently reported in professional journals by Western advisors to BULOG, the full range of BLILOG's activities is yet to receive scrutiny in the West (Timmer 1996, 1997). BULOG was created when Suharto became president of Indonesia. In the apt expressionof one prominent member of the New Order government, BULOG grew with Suharto.I7During Suharto's time in office, BULOG's institutional prominence and the number of commodities it controlled both increased. When the regime ended, BULOG's economic power imploded. To understand these developments and their impact on Indonesian agriculture, we examine the legal and institutional arrangements that secured BULOG's role in the procurementand distribution of a number of staple commodities, as well as the sources used to finance BULOG. BULOG has its roots in military procurement. Its predecessor was KOLOGNAS (KomanhLogistik Nasioml),a military logistics agency whose last commander under the Old Order was Suharto. Although KOLOGNAS was renamed BULOG in order to make its name better suit its new civilian role,I8 the agency nonetheless continued to be run by the military and headed by a military officer right up until 1995. KOLOGNAS, and later BULOG, had close working ties with a group of traders made up of former army quartermasters. As mentioned earlier, this exclusive group of predominantly Chinese traders was granted preferential access to the commodity trade, whiie the military officials of BULOG who secured this access were generously compensated by their clients." BLILOG has served intertwined ideological and economic functions. The agency's main official mandate was to stabilise the price of rice. Rice has been vital to the political success of the regime (Arifin 1995). In his 1969 annual address to the nation, President Suharto announced support for rice farmers and stable and affordable consumer rice prices as cornerstones of his develo ment policy (Mears and Afiff 1996). Nation-wide investment p programs for irrigation and agricultural inputs were launched. While these programs were coordinated by the Department of Agriculture, BULOG's price stabilisiq policies were designed separately and there was little coordination between the two agencies. It is interesting to note that as Suharto's national leadership became better established and oil crises demanded the curtailment of public expenditures, public investments in agriculture diminished, while the scope of BULOG activities expanded (Kasrynoet d.1999). As a consequence,the bulk of financingfor agriculture shifted from a mechanism based on transparent budget allocations, which could not be easily diverted, to BULOG, where transactions were secretive and not transparent. Unlike Thailand, Indonesia has responded to crises in the sector by strengthening rather than weakening the hand of procurement agencies. BLTLOG's role in the sector was elevated to new prominence after the 1973 rice crisis, when rice shortages posed a threat to the survival of the regime." At that time BULOG's military staff was reinforced with a team of Western advisors, who designed a mechanism for well timed and effective rice market interventions.The combination of technical savvy and operational discipline helped produce an image of technical efficiency and professionalism, traits that set BULOG apart from the majority of public agencies in a developing context. Considering the fact that BULOG reached the heights of its international acclaim without ever reporting on the cost of its operations or the use of its profits, it may be usell to take a further look at BULOG's operations and financing. Expansion of BULOG. In 1969, a presidential decree2' made important changes to BULOG's system of governance and operational scope. The agency was made directly accountable to the President, and in addition to rice procurement for government employees was given the tasks of state warehouse construction and rice quality control. While BULOG was to advise the government on price stabilisation and trade in nine staple commodities, it was not to monopolistically control domestic or foreign trade in these commodities. An official report made in 1970 highLghted BUI,OG's corrupt practices. To streamline control over agricultural policies, the report recommended drastic curtailment of BCTLOG's operations as well as its re-subordination to the Department of Agriculture (Bresnan 1993). The report was never officiallydiscussed in the government. In fact, subsequent developments moved in the opposite direction. In 1971, the responsibilities of the agency were dramatically expanded as BULOG became the sole importer and distributor of sugar, the sole importer of wheat, and the sole distributor of wheat flour." In the following years, other commodity operations were 98 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH added to this list, including meat supply to Jakarta (1974), control of soya imports (1977),a price support for corn (1978),and price support programs for soya, mungbeans and groundnuts (1979). A 1979 decree23expanded BLrLOG's operations yet further, making the agency directly responsible for the control and stabilisation of both consumer and producer prices for rice, wheat and other staples. The oil crisis of 1982 underscored the need to curtail budgetary and extra-budgetary expenditures and develop non-oil sources of foreign exchange.Technocratsin the governmentstarted to protest against BULOG's monopolistic position. Minister of Trade Sumitro and Economic Coordination Ministers Widjojo and Wardhana pushed for a reorientation of Indonesian agricultural policies from commodity control to trade liberalisati~n.~~The technocrats found allies in international development agencies, includingthe World Bank, which in 1986 questioned the efficiency of BULOG rice commodity operations and its monopolisation of the wheat, sugar and soya trades (World Bank 1996). None of the advice to curtail BLrLOG's power coming from within or from outside the country was adopted by the government. Eventually the pro-liberalisation Minister ofTrade was removed from officeand the Ministry of Trade was merged with the Ministry of Industry, which was oriented toward state-led import substitution. In 1993, BULOG's mandate was expanded yet further to include maintenance of reserves, and distribution and marketingofcattlefeed.25It took thefinancialcrisis of 1997and Suhards downfall to wrestle the trade in agricultural commodities away from BULOG. A 1998 presidential decree narrowed BULOG's responsibilities to management of price, quality and distribution of rice. Food imports were opened to private traders holding general import licenses. Management of BULOG was passed from a Suharto appointee to the Minister of Industry and Trade (Economist Intelligence Unit 1998a). Thus, over 30 years of Suharto's rule, BULOG came full circle. During these three decades, BULOG expanded massively, so that by 1997 it had grown into a public trading structure that owned or controlled 1,500 warehouses and employed a staff of 7,000.26 The official goals of BULOG expansion were 'improved standard of living for the hmers and through increased farm productivity and farm income' and 'promotin increased consumption of non-rice staples (especially of g the domestically grown "palawija")...thereby bringing a more balanced diet to the people' (Amat 1982:187). These stated goals, however, did nothing to explain the monopolisation of trade in wheat, for example, an imported commodity whose control could not benefit domestic farmers. Neither could centralised importation be viewed as beneficial to consumers, since domestic prices of wheat and wheat flour were 2-2.5 times higher than world prices (Bank of Indonesia, various issues). In 1998, BULOG failed to deliver on its main function, national food security. In the wake of the economic crisis, panic food hoar*, betraying a lack of ~ublicconfidence in BULOG's ability to control prices, led to a drastic increase in the prices of staple foods, particularly rice. BULOG's ceiling prices were broken. According to official estimates, 40 million people (20 per cent of the population) were suffering from hunger (Economist Intelligence Unit 1998b). In 1997198 BULOG operated at a loss (Garcia- Garcia 1998).The machine created to prevent just such a disaster could do nothing to mitigate or alleviate the problem. On the contrary, the expectation that BULOG would be capable of handling food pressures precluded other safety net mechanisms from being developed. Financing of BULOG. A less publicly discussed reason for placing non-rice commodities under BULOG's control was cross-subsidisation-the use of 'quota premiums' from wheat, sugar and soya imports to support rice stabilisation expenses (Pearson 1998). Commoditycross-subsidisation is not the only aspect that makes BULOG financing not transparent. Legally, BULOG financing is the least precise part of otherwisevery detailed documents that govern BULOG's organisation and responsibilities. For example, a 1993 decree on the operations and organisationalstructure of BCTLOG contains 41 articles detailing its structure and organisation, but only one one-line article dedicated to financing: 'The financing of BULOG shall come from the Government, bank loans, and other financial resources pursuant to prevailing stipulation^'.^' The financial accountability reports of BULOG are described as 'composed separately', without any clarification on auditing arrangements. Consequently, BULOG's access to commercial or public resources was unregulated and could take virtuallyany form. The size and use of BULOG's profits have never been officially revealed. The known sources of BULOG financing fall into three categories: state payments for rice sales to civil servants and the armed forces, reportedly at above-market prices; Bank of Indonesia credits to cover the operational costs of BULOG, usually with low interest rates and poorly enforced repayment; and profits derived by BULOG from its operation. 100 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH BULOG's revenue. As noted by Ellis (1990:48), 'BULOG's technical capabilities need to be set apart from the cost of running BUI'OG, about which reiativelylittle is known. The problem with the latter is that BULOG's three-fold food security role has never been differentiated for accounting or financing purposes'. Calculations by Garcia-Garcia (1998) demonstrate that, in the four years preceding the 1998 crisis, the profitability of BULOG stood at about 20 per cent, an improbably high figure for a private trader but believed to be standard for BULOG. According to Garcia-Garcia's calculations, BULOG revenue constituted about 0.4 per cent of the national GDPZ8 By the early-mid 1990s, the cost of rice price stabilisation was thought to have risen substantially to $100-150 million (Pearson et al. 1997). According to the calculations of Garcia-Garcia,this sum constitutes between one-third and one-fifth of BULOG's profits, even if profits are calculated without accounting for preferential credits from the Bank of Indonesia (Garcia-Garcia 1998:13). BULOG sold all imported wheat, all domestically produced and imported sugar, all imported rice, and an estimated 5-10 percent share of domestically produced rice to a group of lar e-scalelicensed traders (Garcia-Garcia 1998). g Traders belonging to an association of BULOG distributorsreceived a quota from BULOG which they then traded among themselves, an indication of quota rents. The post-Suharto scandals revealed massive siphoning of BULOG resources for personal and political gain by the highest officials of the New Order, Habibie and even Megawati governments (Economist Intelligence Unit 2OOOb). The lure of transactions via BULOG that are not transparent will remain as long as the agency's role and new accounting practices have not been codified and the political situation remains unstable. Assessment of BULOG. Literature that examines BULOG's successes primarily concentrates on its rice price stabilisation function. Assessments range from the view that price stabilisation puts rice producers at risk to the view that stability in intra-year and inter-year rice prices is key to farmers' decision to produce (Jones 1995; Dawe 1996). The current assessment of many observers is that BULOG's rice stabilisation effort has been well designed and executed, and that in the 1970s it played a positive role as a complement to rice investment programs. However, most agree that, with time, the share of rice in private consumption has decreased and the readiness of farmers to diversify into other crops has increased, as have the cost of the program and the rate of protection contained in BULOG-supported rice prices. Today, both designers and long-time supporters of rice stabilisation programs suggest a re-examination of BULOG's role in rice policies, including turning to tariff-based controls (Tabor and Meijerink 1998; Timmer 1993, 1996; Warr 1999). BULOG's control over other commodities did not have the ideological - or political forceof the rice mandate, nor the justification of superior technical competency. No food security argument could justify the monopoly in imports of sugar, wheat or soybeans, nor was there an economic justification for the persistent rigidityof the trading regime for these commodities (Mears 1982). In any case, as we will show below, it was not just these "econdaq' commodities that suffered from excessive control, but rice itself, the main justification for BULOG's existence. Rice. Development of the rice sector was not just one development goal among many. It was central to the legitimacy of the regime and its leader. 'Rice self-sufficiency was President Suharto's personal. obsession and that was the main reason why it had always been a top priority of agricultural development during the New Order regime' (Simatupang 1999:144). Support for rice production at almost any price was the outgrowth of this position. Domestically, BULOG intervenes in rice markets at harvest time and buys at a pre-announced floor price. It also imports rice both for the intra- year buffer stock and the inter-yex buffer hnd. It stands ready to protect the announced ceiling prices and sells when the price rises above the announced limit. When imports are required, designated rice importers come from the same group of appointed traders that is involved in exports and domestic distribution. After the public announcement of rice self- ~ ~ c i e n iny1984, the government adopted a policy of exporting excess c rice in years of surplus and importingin years of shortfall. Both arrangements have been coordinated by BULOG, and both have proved problematic. Ideologically, afier self-sufficiency had been announced, the government was reluctant to import, not wanting to tarnish the imageof the achievement. Aftera bad domesticharvestin 1986, ricewas imported too lateand BULOG was forced to break the ceiling price. Only afier the face-saving formula of 'self-sufficiency on trend' had been created were scheduled imports resumed. In 1987, the year of a big rice harvest, the government authorised and subsidised exports of rice by BULOG. However, BULOG had neither su6cient experience to promote rice exports nor sufficient interest in it 102 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH since domestic costs and prices were both higher than world costs and prices at that time. BULOG did not invest in cross-subsidsing rice exports by domestic revenue, the path taken byThai sugar exporters, because revenues were channelled for uses outside the sector. Thus, BULOG's structure remained unchanged despite a dramatic change in its functions. Rice distribution or political control? Despite a fairly small market share in rice, BULOG has the ability to control not just domestic prices, but also channels of domestic rice sales. BULOG makes 80 per cent of its domestic rice purchases from KUDs (producer cooperatives) at prices slightly higher than the domestic market price. Producer cooperatives are government- created entities headed by appointed leader^.'^ There are numerous reports of economic inefficiencies and corruption in KUD operations and of farmer resistance to KUD prescriptions for the production of commodities and use of inputs. However, KUDs acted as conduits of government policies at the villagelevel and as organising units for GOLKAR, the governingpolitical party. Thus, assertions of free domestic trade in rice would seem to underestimate BULOG's control of this key distribution and marketing channel. The benefits to rice farmers from BULOG's price stabilising operations are a matter of much debate. Above-cost domestic procurement and diminished price fluctuations combined with the government irrigation program and input subsidies have allowed rice production to expand. At the same time, as returns from the Green Revolution have diminished, farmers have been willing to move to other more lucrative cash crops. Such diversification, however, is reported to be hampered by KUD coercion of farmers to continue with rice produ~tion.~~ In the late 1980s, the advice of international donors to curtail trade intervention in agricultural commodities was not followed, whereas advice to curtail state investment in rice production was. One interpretation of this selectivity is that the legitimising role of rice subsidies has diminished, while the redistribution role of BULOG has not. As subsidies decreased, so did the incentives for farmers to concentrate on rice. However, farmers' ability to move to other crops was constrained by continuing ad hoc subsidies of rice and uncertainty about state controls on other c~mmodities.~' In the post-Suharto years, support for blanket consumer subsidy schemes was no longer an option. On IMF insistence, the program of rice subsides for the poor has replaced the blanket ceiling price support scheme. The ideology behind the rice quest has been dismantled (Economist Intelligence DO DIFFERENT REGIMES D~STORTDIFFERENTLY? 103 Unit 1999b),and production of rice has been disrupted by the inability of post-Suharto BULOG to povide input credits through KUDs and the vagueness of possible private arrangements (Economist Intelligence Unit 1999a). Restructuring rice production into an efficient productive sector will take much more than a change in tradin arrangements. There is a g chance, however, that rice producers will have much more say in the design of the new generation of Indonesian a ricultural policies. g Sugar. The top-down system of sugar-marketing controls includes disincentives for increased and more efficient domestic production-both growingand processing.Until recently, the majorsector playerswere removed from decision making and could not participate in the redesign of the institutions governing sugar production and sales. The spectacular failure of sugar policies is sometimes explained, if not justified, by the role sugar revenue played in subsidising rice production. This justification appears weak, however, since the bulk of sugar revenue diverted from the sector benefited an exclusive group of traders much more than it did sustainable investment policies in the rice sub-sector. Development and disincentives. Under Dutch rule, Indonesia was the world's second largest sugar exporter. Since then, however, sugar production has suffered from a number of afflictions, the most important 'being rigid production and milling arrangements. The system of forced production introduced by the Dutch has in principle remained unchanged. A revenue- sharing arrangement deprived both growers and millers of incentives to invest in increased yields or extraction. With the sugar trade removed from the key actors in the sub-sector, Indonesian sugar became less competitive. During the period studied, Indonesia became a major sugar importer. Growing. From a legal standpoint, the main difference between the colonial arrangement and the Cane Intensification Program (TRI) launched in 1975 was that under the former farmers were forced to lease land to sugar mills, - whereas under the latter theywere forced to cultivatecane to ensure adequate supplies to sugar mills (Mubyarto 1977). The TRI program also included a package of credits and inputs for farmers, who were required to use the subsidised inputs for sugarcane production and sell the output to designated mills. Payment was determined by the rate of extraction and paid in cash minus the input advances. Before 1980, farmers had the choice of being paid in either cash or sugar; because they thought the sugar price was too low, they tended to prefer in-lund payments. Since 1981, BULOG has monopolised the domestic sugar trade, and farmers have been deprived of 104 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH the in-kind payment option; consequently, they have not been able to use sales price differentials to enhance their revenue (Amin and Rahman 1982). Between 1988 and 1997, the price received by farmers decreased from 66.4 per cent to 61.8 per cent of the retail sugar price, a very low share relative to international practices (Rusastra et al. 1999). Farmers suffered losses from sugarcane growing and would likely have switched to other crops if not for restrictions. Despite the average increase in sugarcane area between 1970 and 1996 of 15 per cent per year, cane and sugar productivity per hectare has been decreasing.The fall in yield has not been a function of the move to non-irrigated areas alone. For example, between 1988 and 1997, yields have fallen in both irrigated and non- irrigated areas (Rusastra et al. 1999). MiUmg. Whereas in colonial times millers had established a collective sugar marketing structure to secure collective bargaining in the overseas markets and coordinate self-regulation of the sector, the TRI arrangement effectively removed millers from decision making in the sugar sector (Nelson and Panggebean 1991). The price of sugar was decided on a cost-plus basis by BULOG and the Ministries of Agriculture,Trade and Finance. This pricing arrangement contained no incentive either to increase the efficiency of sugar extraction or to increase overall production. The competition-free environment allowed state-owned mills, notorious for outdated equipment and low productivity, to survive in parallel with private mills, which produced almost twice as much sugar as the state-owned mills (Rusastra et al. 1999). The restructuring plan adopted after the collapse of the Suharto regime calls for the closure of 26 out of 57 aged mills operating in Java (Economist Intelligence Unit 2000a). Tradmg. As described above, BULOG sells sugar to an appointed group of traders that trade sugar allocations among themselves.The list of BULOG sugar distributors includes 400 traders, in practice controlled by a much smaller number of ~li~opolistictraders who collude to increase sales margins (University of Indonesia 1998). Between 1988 and 1997, the domestic price share of traders increased from 17.2 per cent to 23.1 per cent (Rusastra et al. 1999).Accordin to Garcia-Garcia's calculations (1998),in the 1990s g before the 1997sugar crisis, imports were BULOG's main source of income, generating one-half of the gross margins of the agency's operations, although it did succeed in inter-year sugar price stabilisation. Considering that domestic prices have been stabilised at a level 2-2.5 times higher than that of world prices, however, this is hardly a meaningful achievement. For BULOG, rents from monopoly sugar imports were higher than from domestic procurement, while transaction costs were much lower, giving the agency no incentive to enhance domestic production. It is not surprising, then, that sugar imports increased from 6 per cent of consumption in 1985 to 38 per cent in 1997. In any list of achievements of the Suharto regime, rice self-sufficiencyis considered one of its proudest moments. The examination of collateral damage, in this case to the sugar industry makes this achievement seem far less impressive. The Philippines The Philippines presents the most graphic picture of the negative effects of a state monopoly of commodity trading for both commodity producers and more broadly for the national economy. In the case of sugar, the Marcos regime's undermining of the sector's economic power was not just a side effect of cross-subsidisation or rent seeking, but rather an explicit political objective, intended to undermine the powerful sugar interests. In the case of rice, impressive research and development breakthroughs were thwarted at the implementation srage by poorly coordinated input distribution policies and unfinishedland reforms. For sugar, theend of the Marcos regime brought trade liberalisation and higher levels of protection, whereas for rice the post- Marcos years saw the continuation of unbalanced investment programs and poorly orchestrated state procurement. Here we discuss the institutional arrangements that have led to these grim outcomes and relate them to the effective exclusion of producers from sector-specific decision making. The National Food Authority. State procurement arrangements in the Philippines look l i e a poorly executed version of BULOG trading practices, with the waste and diversion of resources apparently even higher and the welfare effects even lower. History of expansion. In 1962, the Rice and Corn Administration (RCA) was created under the ofice of the Pre~ident.~'The body's mandate was to distribute rice imports in urban areas, set floor and ceiling prices for rice and corn, and accumulate stocks for the national reserve. To achieve this goal, it was authorised to procure rice and corn directly from producers. Rice imports, however, were not monopolised. Beginning in 1966, the RCA was allowed to import rice and corn through private traders after public bidding in cases of national emergency or a shortage of domestic production. 106 RURAL. DEVELOPMENT AND AGRICULTURAL GROWTH The RCA was controlled by a Board of Administrators composed of a chairman and four members, two of whom represented consumers and two producers. A fimd was allocated by the national treasury to cover operating expenses. In subsequent years, a series of consolidations centralised and expanded the role of the parastatal. With the onset of martial law in 1972, the RCA was placed under the supervision of the National Grains Authority (NGA), controlled by the Ministry of Agriculture."The mandate of the NGA was expanded. In line with the development fashion of the time, the NGA was supposed to implement an integrated approach to crop production, effectively meaning ill-defined coordination among growers, millers and traders. Under new arrangement, the NGA was to be financed by the National Grains lndustry Development Fund, which had accumulated proceeds from the additional tax on rice and corn mills, surcharges on warehouses, import licensing fees, and funding from international donors. The centralisation of commodity trading in the late 1970s,when Marcos's rule began, went beyond any possible integrated development recommendations. Like BULOG, the NGA was placed under the direct control of the office of the President.34The NGA's scope was expanded to include wheat,soya, mongo and cassava,and the licensingof grain processors, wholesalers and retailers. The composition of the NGA Council was reconstituted, and representativesof rice producersand traderswere removed. The NGA was granted fill control of both import and export of all grains. The final dramaticexpansion took place in 1981. The NGA was renamed the National Food Authority (NFA). The NFA was authorised to control not only grains but 'basic food commodities' in general, including meat and fish products, vegetables, fruit, drugs and an array of consumer goods. The NFA was given the power to regulate food producers, processors and traders, and manufactured and processed food products. It controlled a network of retail stores and appropriated a food terminal. The Minster of Human Settlements, Imelda Marcos, was appointed chairman of its council. NFA financing under Marcos. Under Marcos, the NFA enjoyed access to a vast range of &ding sources. In addition to the authorised capital stock of 5 billion pesos (approximately 2 per cent of GNP in 1981) fully underwritten by the government, the government could make additional equity investments. Further, the NFA could issue its own tax-exempt, m y guaranteed bonds, obtain guaranteed credits from abroad, and purchase land. The government could subsidise operations of the NFA, and the body also received preferential interest rates from the Central Bank.These subsidies were grantedto the NFA on top of its own revenue, which officially doubled after lucrative wheat imports were added to the list of controlled commodities. A World Bank report estimated that between 1976 and 1982 the NFA made $310 million in net profits from these imports alone (World Bank 1984). From an organisation that employed 1,000 people in 1966-67, the NFA had grown into a corporation of 11,207 employees by 1984. NFA in a pluralistic setting. These fantastic provisions for financing the NFA proved excessive even for a regime h o u s for its excess. Eventually, pressure began to mount for some change to the structure of NFA. Gerardo Sicat, the director general of the National Economic and Development Authority, resigned, and in 1985, under pressure from the IMF and World Bank, the IVFA's control over milled rice and non-grain commodities was terminated.At the same time, import of feed grains and wheat was liberalised. However, it was not until the Aquino administration came into power that the hTFA lost control over the commodities outside its original core competencies.In 1986, the secondexecutiveorder of theAquino government removed the NFA from presidential control and placed it under the Department of Agriculture. With the removal of profit-generating commodities from the NFA's jurisdiction, the NFA had to depend heavily on state subsidies. The size of the subsidy was set at about 1 billion pesos per year, a sum insufficient to cover its sales operations, high overhead, and interest payments on guaranteed state borrowing. Changes in the management of the NFA were more difficult to achieve. An institution designed as an authoritarian rent-seeking machine was slow to adapt to a more transparent and legally binding setting. Recent reviews of the NFA proved that its rice price stabilisation role was very costly- delivering one peso of price support to producers in the late 1980s cost more than one peso.35At the same time, the scale of the NFA operations, which fluctuated from 7.0 to 0.4 per cent of marketable surplus, was too small to affect a large percentage of rice farmers. The abolition of the NFA, however, proved to be politically difficult, pady because consumers learned to regard rice price stabilisation as a public good, even though the rice share of household expenditures had decreased, particularly in urban areas. 108 RURAL DEVELOPiMENT AND AGRICULlWRAL GROWTH Nonetheless, the operational differences of the NFA in an authoritarian, as opposed to democratic, setting are dramatic. Cross-subsidisation that affected the production of other commodities has been discontinued. Moreover, financing sources of a parastatal in a democratic setting have become a matter of public knowledge. Unlike BULOG, NFA has allowed its income statements and balance sheets to be published and scrutinised (Clarete et d.1993). In its heyday beginning in the mid 1970s,the NFA's mandatewas broader than that of BULOG. Unlike BULOG, it was much less successful in fulfillingits main function-stabilising the rice price-and its investment failures were much more visible. One of the main official mandates of a grain parastatalis to support grain farmers. Has the NFA's policy fulfilled this goal? According to Intal and Power's (1990) study of the effects of trade and agricultural pricing policies in the Philippines, in the 1960s direct output price interventions in agriculture resulted in positive transfers to the producers. In the 1970s these transfers turned negative for rice, sugar and coconut farmers. Overall, nominal and real incomes of sugar and coconut farmers were negatively affected by direct price interventions. The income of corn farmers improved, as did that of rice farmers, if the restrictive output effect of price changes is disregarded. With the output effect factored in, the revenue of rice farmers turns negative. Furthermore, both instantaneous and short- run direct income effects turned negative &om 1980 to 1984 when input subsidies decreased. Rice. Growth in rice yields generated by Green Revolution breakthroughs was first realised in the Philippines in the late 1960s. The successful launch of the Green Revolution was one of the reasons Marcos was able to muster enough popularity to consolidate power in the early 1970s without much protest. In the late 1970s, credits granted by the Masagana 99 agricultural investment program were not repaid, and the program was curtailed. Producer prices received by Philippine rice farmers were close to world prices (World Bank 1984),yet the profits for both firmers and millers were reduced by the NFAls inefficient handling of rice exports. As in Indonesia, the NFA's rice-exporting mandate produced a conflict of interest. Exports required initial expenditures and knowledge of markets that the NFA lacked and, as a national monopolist, had no incentive to acquire. In the late 1970s, the NFA accumulated stock that it was reluctant to sell. When it was pressured to sell in the early 1980s,the world price was low, thus its losses were high (World Bank 1984). Loss of income through inefficient sales created disincentives for both fsrmers and millers to expand or improve production (David 1996). From the second half of the 1980s onward, nominal protection rates in rice increased somewhat, but they still remained fairly low, at below 20 per cent. It was argued that such a low level of protection served to o&et the effectsof peso overvaluation. The increase in protection rates, particularly as of 1995, was a testimony to the increasing power of the rice lobby. Growth in rice production slowed from the mid 1980s onward, in part for technolo ical reasons that are beyond the scope of this analysis. g Sugar. Sugar has been one of the main Philippine exports since the end of the 19th century. During the last century, Philippinesugar exports depended on the preferential US quota. After the Second World War and up until 1974, the United States allocated the Philippines a quota of approximately 1 million tons. One effect of this preferential pricing was expanded sugar production without increased efficiency (Nelson 1988).Securesugar markets also assured secure prosperityfor large sugar growers, whose market position was reinforced by political leverage. In pre-Marcos times the government did not actively intervene in self- policing of the sugar sector. The Sugar Quota Authority allocated the US and domestic quotas through consultations with the sector. The quotas were divided into three parts: the domestic quota, sold at prices lower than US quota prices but higher than world prices; the lucrative US quota; and the quota for other exports. Each mill was individually responsible for the sale of the quotas. Sugar under Marcos. Major changes in the management and control of the sugar industry occurred in 1974. The US quota was removed, and internally the government assumed a much more controlling and much less collegial attitude toward sugar traders. World sugar prices were at a record high, so windfall revenue could have eased the adjustment of the Philippine sugar industry to post-US quota market conditions. The new state-controlled structures, however, turned this opportune situation into a crisis. In 1974, a Philippine exchange company (Philex), the agency of the state-owned Philippine National Bank, was designated as the sole buyer of sugar mill output and the sole exporter of sugar. Philex, later renamed NASUTRA, boughtsugar at composite prices and divided it between quotas for sale without producer participation. Transactions were not transparent, 110 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH and no information was made available to the public. Philex was inexperienced in international trade and held the sugar, expecting prices to rise, but was later forced to sell at a loss (Pabuayon and Catelo 1985). The Marcos government took control of both domestic and international trade with two objectives in mind. The official justification for the government takeover of sugar marketing was the same as in Indonesia- protection of consumers and producers from fluctuating world prices. The unstated goal was to reduce the financialand political power of the powerfd sugar lobby.The takeover of the sugar sector gave greatest control to Marcos's closest associates, who were able to realise the greatest benefits for themselves. Thus, the unofficial goal of political consolidation and weakening of the sugar lobby was achieved. In 1979, the National Sugar Trading Corporation (NASUTRA), headed by the infamous Marcos associate Roberto Benedicto, started to operate through a group of licensed traders, effectively increasing the chain of intermediaries and creating rents for this exclusive group. NASUTRA published no information on its sales prices or the margins it received for its services. Like Philex, NASUTRA was grossly incompetent and sold at low prices (Larkin 1993). As with other state agencies, traders' margins increased from 8 per cent before the NASUTRA monopoly to 11- 21 per cent for NASUTRA's trading (Intal and Power 1990). These direct interventions cost producers dearly-the nominal protection rate due to direct price interventions fell from 101 per cent in 1960-71 to -16 per cent in 1972-8 1(Intaland Power 1990).While the composite price was higher than the world price in 1980181, the producer price was kept 20 per cent lower than the world price. Even at these low rates, producer payments were delayed, causing sugar growers and millers to accumulate debt. Many plantations and smaller growers were effectively made bankrupt. Sugar production fell from 2.5 million tonnes in 1973 to 1.7 million tonnes in 1985. In 1983 and 1984, the Philippines imported sugar for the first time since the pre-war years (Pabuayon and Catelo 1985). The sugar sector found itself in a deep crisis, with NASUTRA effectively bankrupt. On the insistence of the IMF, its monopoly was broken, with millers being allowed to sell directlyto domestic traders. The government, however, retained a monopoly over exports. Sugar in the post-Marcos years. Under the Aquino government, the sugar trade was returned to the private sector. Over time, sugar producers and millers agreed that to be competitive in the world market, their industry had to be made more efficient.36 Multiple ineff~cienciesand grower-miller feuds still plague the sugar sector. However, production in the period 1992-96 was 45 per cent greater than in 1986187. Investment in the sector remains low. One important reason for low investment is outside the sector's control-the unresolved issue of land reform, which impedes long-term planning. Producers have been trying to ensure protection for the sector through import controls as a trade-off for agreeing to GATT regulations. The democratic process is slow and has its pitfalls, but it is now more transparent and the relationship between the sector and the government is more equal than before. The Philippine sugar sector realises that it has lost its international competitiveness and is currently trying to reinvent itself. The National Council of Sugar Producers has emerged as an important player in negotiations with the government. The council sponsors the Sugar Research Foundation, whose objective is to conduct research on ways of increasing efficiency and improvingsugar marketing. In 1989190, the National Council of Sugar Producers requested that the Sugar Regulatory Administration impose a special lien on the purchasing agreements of all sugar producers to create a Sugar Market Development Fund 'to pursue programs which will benefit sugar industry'. Millers' and planters' associations are currently negotiating new rules of engagement that would increase the competitiveness of Philippine sugar.37 CONCLUSIONS AND POLICY IMPLICATIONS This paper has examined the effects of different regimes on different agricultural sectors and incentives for key actors to enhance production and innovate. Changes in agricultural policies for two key commodities-rice and sugar-in Thailand, Indonesia and the Philippines were examined. Among the regimes, there were specific differences in the distribution of benefits and production incentives built into rice and sugar policies. The main differences were not in the inherent ability of a regime to design a better or worse agricultural policy, but rather in its ability to redesign, in a timely manner, a policy that had outlived its usefulness and in a way that benefited key actors in the sector. In a pluralistic setting where a policy was not the ideological or legitimising quest of an autocrat, such timely replacement of a bad policy was possible. Another major difference was the composition of policy designers. In a pluralistic arrangement, producers, not bureaucrats, were the judges of policy effectiveness. They were also among the designers of new arrangements. Where producers did not have 112 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH access to policy design and were aware of the government's ability to unilaterally prolong or replace a certain policy, their ability to plan was limited. Rice policies in Indonesia and the Philippines may not have been bad at the time they were conceived; they did, however, become bad when the international and domestic economic climate changed but they remained the same. They turned worse when the institutions created to implement them began to be used as safe havens for the reallocation of resources and graft. Producers were excluded from access to knowledge of how resources were used, and often were not among the beneficiaries of resource exploitation. Producers did not have consistent access to public resources generated outside the sector, thus they were deprived of incentives to innovate and increase efficiency. When producers can operate in a 'developmental band' that allows for innovative policy design, such policies can translate into sectoral innovation, which, along with limited access to public resources, is a prescription for expansion and more efficient growth. Policy implications The differences between the effects of autocratic and more democratic institutions on the welfare of producers have not been well studied. The lack of consensus on the long and short-term effects of various levels of authoritarianism on economic outcomes leaves development practitioners on shaky groundswhen consideringadvice that is socially and economically beneficial in an authoritarian context.The lack of clarity about these central development issues is unsettling primarily because it deprives development practitioners of guidelines for a sequence of desirable and mutually re- enforcing changes in political and economic realities-changes that would make betterment of the human condition more plausible Time and a p n , the World Bank and other donor agencies havesupported large-scale economic projects started by authoritarian regimes, yet leaders like Suharto and Marcos have failed as both political regimes and conduits of development policies. Often the argument was that the stability such strong leaders ensured was more conducive to economic reform than the instability created by weak democracies. Thls study of interactions between commodity producers and the state in three Southeast Asian countries shows that failures of coordination and resource distribution are not accidental, but are built into the logic of a regime that does not need to build consensus to achieve its Such regimes appear strong because they can relocate large shares of national resources in a virtually uncontrolled manner. However, they have a crucial weakness: the need to legitimise their actions. In the context of the two authoritarian nations, the promise to feed the country through centralisationserved as a legitimisingdevice. Such abuse of national resources and agricultural products is the reason agricultural and food policies are more likely to fall prey to ideological imperatives than almost any other policies. It is crucial for development practitioners to always work against such political use of economic policies. Before development schemes are approved, the political economy of a sector should be better understood and more carefully considered. More country-based and commodiq-based studies in combination with cross- country empirical work will improve development advice and help determine the chances for such advice to be heard and followed by more developmental rather than predatory power holders. ACKNOWLEDGMENTS I would like to thank Steven Collier, Leah Cohen, and the World Bank residence missions in Indonesia, the Philippines, and Thailand for very helpful suggestions and support. All errors and omissions are mine. NOTES Interviewwith Mr Sampop,advisor to the Prime Minster, fall 1998. For an account of the coup and bibliography,see Legge (1972). For a discussion of possible participation byvariousnationaland internationalforces in thecoup, seeVatikiotis(1998). For adiscussionof the roleof the military,seevatikiotis(1998). For thelatestdevelopments, see Shiraishi(1999). * John Bresnancontraststheprimacyof themilitaryin Indonesiawith their moreconstrained positions in the Philippinesandin Thailand. InThailand,'the civilserviceremainedseparate and independent (from the military), owing pcrhaps to its history in the service of the monarchy', which allowed it 'to sustain flexibilityin Thai politics'. In the Philippines,'the army was led by men trained to serve as apolitical officers' (Bresnan 1993:273-4). It is important to note that in the Philippines, martial law had to be declared to make the militarypoliticallyprominent,whereasin Indonesiaparticipationof themilitaryis not only traditional but constitutional. 114 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Interview with Mr Frans Seda, former Minister of Finance and Minister of Agriculture of Indonesia, Jakarta, October 1998. Interview with Mr Lim, Under-secretaryof Agriculture, 1980-86, Anuila, October 1998. Much of this discussion of legitimisation is based on Wurfel (1988). Interview with Marinela Castillo, Executive Director, Agricultural Credit Policy Council, Department of Agriculture, Manila, October 1998. Interview with Raul Montemayor, Business Manager, Associations of Farmers, Manila, October 1998. l oUnless specified otherwise, discussion of the post-Marcos economic polides is based on Tolentino (1994). " Interview with Marinela Castillo, Executive Director, Agricultural Credit Policy Council, Departmenc of Agriculture, Manila, October 1998. ' In the 1960s,the share of the rice premium--a fee paid for export licenses-in national revenue ranged from 12.1 per cent in 1961 to 6.8 per cent in 1969 (Siamwalla1975). l 3Interview with Dr RungruangIsankura Na Ayutya, Deputy PermanentSecretary,Ministry of Agriculture and Cooperatives, Bangkok, October 1998. l 4Unless specified otherwise, the description of institutional arrangements in sugar trade from the late 1950s to the mid 1970s is based on Jessadachatr (1977). l 5Sugar IndustryAct B.E. 2504,23 November 1961. l GFor a description of the MOI's shififrom the roleof asupervisorof millers in the 1960s to thatofamediacorbetweenmillersand growersinthe 1980s,seeSiamwallaandSetboonsarng (1989)and Ramsay (1987). l 7InterviewwithMrSeda,former Directorofthe DepartmentofAgriculture,former Director of the Department of Finance, Jakarta, 21 October 1998. Interview with BustanilArifin, Chairman of BULOG,1972-93,Jakarta, October 1998. ' For more details on the nature of the military-trader alliance,see Robinson (1986). Interview with BustanilArifin, Chairman of BULOG, 1972-93, October 1998. Presidential Decree No.1111969, 1969. 22 The First Long-Term Development Plan for BULOG (BULOG dalam PJPTl), Jakarta, 1977. 2 3Presidential Decree No. 3911978, 1978. 24 Interviewwith Mr Sudaryanto,Deputyhead of BPN (NationalLandAgency),21 October 1998. 25 Presidential Decree No. 10311993, November 1993. 26 The Fist Long-Term Development Plan for BULOG (BULOG dalarn PJPTl), Jakarca, 1977. 27 Presidential Decree No. 10311993, November 1993, artide 37. 28 Calculatedfrom data in Garcia-Garcia (1998). 2' Interview with Dr Sayogyo, Institute of Agriculture (IPB), Bogor, October 1998. 30 Interview with Dr Sayogyo, Institute of Agriculture (IPB), Bogor, October 1998. 3' Interviewwith Dr Sayogyo, Institute of Agriculture (IPB), Bogor, October 1998. 32 RepublicAct No. 3452, 14 June 1962. 3 3Presidential Decree No. 4, 26 September 1972. 34 Presidential Decree No. 1485, 1l June 1978. 3 5Unless noted othenvise, information on the NFA and its operationscome from Clareteet al. (1993),Asian Development Bank (1998),and Intal and Power (1990). 3"nte~iew with Jose Maria Zabaleta, Executive Director of the Philippine Sugar Millers Association, Manila, October 1998. 37 Report of the Presidential Task Force on the Sugar Industry, Ofice of the President, Manila, 1997. REFERENCES Australian Centre for InternationalAgricultural Research, 1994. Philippine Sugar: an i n d u sfinding its feet, ACIAR Project Report, Centre for ~ International Economics, ACIAR, Canberra and Sydney. Amat, Soegang, 1982. Indonesia: problems and issues of national foodgrain agencies in the decade of the eighties, Paper presented at the FA0 WorkshoponTechnical Cooperationamong NationalFoodgrainAgencies in Asia and the Pacific, Jakarta, 7-10 December. Amii, M. and Rahman, Anas, 1982. The Prospects of Sugar Cane and its Alternative Sources, Paper presented at the Badan Penerapan dan Pengembangan Teknologi [Agency for the Assessment and Application of Technology] Symposium, Jakarta, 28 April. Anderson, B., 1998a. The Spectre of Comparkons:mtiomlimz, Southemt Asia, and the world, Verso Books, New York. Arifin, Bustanil, 1995. Food crisis management, Jakarta (unpublished). Asian Development Bank, 1998. Grains Sector Development Progmm, Technical Assistance Report PHI 30087-01, Asian Development Bank, Manila. Balisacan,A., 1990. Why Do Govmments Do WhatThqDo?Ap&n Re@m in the Philippines, Worlung Paper Series No. 90-03, University of the Philippines at Los Banos, Laguna. -,1992. Rural Deuelopment in the Philippines: patterns, cons~aints,and respomes, Institute of Development Economics, Tokyo. Bank of Indonesia, various issues (1987-97). Weekly reports, Bank of Indonesia, Jakarta. 116 RURAL DEVELOPMENT AND AGRICIJLTURAL GROWTH Booth, A., 1999. 'Initial conditionsand miraculous growth: why is Southeast Asia different from Taiwan and South Korea?', World Development, 27(2):301-21. Borell, B., 1994. Philippine Sugar: an industryfinding its feet, Centre for International Economics, Canberra. Bresnan, J., 1993. Managing Indonesia: the modern political economy, Columbia University Press, New York. Christensen, S., 1993. Democracy without equity?The institutions and political consequences of Bangkok-based development, Paper presented at the Thailand Development Research Institute (TDRI) Year-End ConferenceWho Gets What And How?Challengesfor the Future, Chon Buri, Thailand, 10 December. -andSiamwalla,A.,1993.WEJoGetsWEJatandHow?Challengesforthe Future, Thailand Development Research Institute, Bangkok. Clarete, R., Fabre, R., Balbosa, C. and Castro, E., 1993. Securing food security: performance assessment and future directions of the NFA, Interim report prepared for the Department of Agriculture and the Government Corporations Monitoring Coordinating Council, Department of Agriculture, Manila (unpublished). Cribb, R and Brown, C., 1995. Modern Indonesia: a histoy since 1945, Longman Group, New York. David, C., 1984. Economic Policies and PhilippineAgriculture,Working Paper Series 83-02, Philippine Institute for Development Studies, Manila. -, 1996. Food PoliLy:its role in price stability andfood security,Discussion Paper Series 97-11, Philippine Institute for Development Studies, Manila. Dawe, D., 1996. 'Rice Price Stabilization in Indonesia: a comment', Bulletin of Indonesian Economic Studies, 32(2):105-14. Department of AgriculturalReform, 1983. Opuation Land Tamfe,Program Accomplishment, March, Manila. Dixon,C., 1999. The ThaiEconomy:unevendevelopmentandintemtionalimz, Routledge, New York and London. du Genestoux,I?,1991. Evolution and Present Status of the Sugar Industries of Asia, Paper presented in workshop on the Asian Sugarcane Industry, United Nations Industrial Development Organization (UNIDO),May. Economist Intelligence Unit, 1998a.Country Report: Indonesia,4 December, Economist Intelligence Unit, London. -, 1998b. Country Rtport: Indone&, 11 August, Economist Intelligence Unit, London. -, 1999a. Country Report: Indonesia, 31 May, Economist Intelligence Unit, London. -, 1999b. Country Report: Indonesia, 28 May, Economist Intelligence Unit, London. --,2000a.CountryReport:Indonesia,15November,Economist Intelligence Unit, London. -, 2000b. Country Report: Indonesia, 31 May, Economist Intelligence Unit, London. Ellis, F., 1990. 'The rice market and its management in Indonesia', IDS Bulletin, 21(3):44-5 1. Evans, P., 1995. Embe&d Autonomy: states and industrial transformation, Princeton University Press, Princeton. -- 1998.'Transferablelessons?Re-examiningtheinstitutional prerequisites of East Asian economic policies', Journal of Developmental Studies, 34(6):66-86. Garcia-Garcia, J., 1998. BULOG Operations, Public Expenditure, and Income Distribution, World Bank, Washington, DC (unpublished). Haggard, S. and Kaufman, R, 1995. The Political Economy of Democratzi Zamitiom, Princeton University Press, Princeton. Hewison, K., 1986. 'Capital in the Thai countryside: the sugar industry', Journal of Contemporary Asia, 16(1):3-17. Intal, I? and Power, J., 1990. Trade, &change Rate, and Agricultural Pricing Policies in the Philippines, World Bank, Washington, DC. Jessadachatr, Phitsanes,1977. A History of Sugar Policies in Thailand, 1937- 75. MA thesis, Faculty of Economics, Thammasat University, Bangkok. Jones, C., 1995. 'Rice price stabilization in Indonesia: an economic assessment of the changes in risk bearing', Bulletin of Idnesian Economic Studies, 31(1):108-28. Kasryno, Faisal, Nataaunadja, Hidayat and Rachman, B., 1999. 'Agriculture development in Indonesia: entering the 2lSrcentury', in R. Stringer, S. Bahri, M.S. Pasaribu and Pantjar Simatupang (eds), Indonesia2 Economic Crisis: efects on agrt'culmre and policy responses, Centre for International Economic Studies, Adelaide University, Adelaide:403-24. Kasuya, Y., 1995. The 'Fdilure'of Agarian worm in TrdmitionulDemocrq: Philippines, 1986-92, Working Paper No. 194, Institute of Social Studies, The Hague. Laothamatas,Anek, 1992. BusinessAssociationsand theNew PoliticalEconomy of Thailand.fiom bureaucraticpolity to liberal corporatism,Westview Press, Boulder. 118 RURAL DEVELOPMENT AND AGRICUITURAL GROWTH Larkin,J., 1993. Sugarandthe OriginsofModernPhilippineSociety,University of California Press, Berkeley. Legge, J.D., 1972. Sukamo: a political biopphy, Praeger Publishers, Washington, DC. Lipset, S.M., 1959. 'Some social requisites of democracy: economic development and political legitimacy', American Political Science Review, 53(1):69-105. Mears, L., 1982. Experience in Food Grain Management in Asian Countries, Paper presented at the FA0 workshop on Technical Cooperation among National Food Grain Agencies in Asia and the Pacific, Jakarta, 7-10 December. -andMff,Saleh,1996.'Anoperationalricepolicyrevisited',inFachry Ali, Fachry Ali, Imam Ahmad, Abdul Hamid and I Ketut Mardjana (eds), Beras, KoperasiDan Politik Or& Baru: BustanilArifin70 tahun [in honour of former Minister of Cooperativesand Head ofthe State LogisticsAgency Bustanil Arifin, b. 19251, Pustaka Sinar Harapan, Jakarta:399-413. Mubyarto, 1977. 'The sugar industry: from estate to smallholder cane production', Bulletin of Indonesian Economic Studies, 13(2):29-44. Muscat, R., 1994. The Fzfi Tiger: a study of Thai dmelopment policy, UN University Press, Toyko. Nelson, G.C. and Panggebe+n, M., 1991. 'The costs of Indonesian sugar policy: a policy analysis matrix approach', American-JournalofAgricultural Economics, 73(3):703-12. Nelson, G., 1988. 'Sugar in the Philippines: policy responses to a distorted world market', Food Policy, 13(3):283-92. O'Domell, G., 1973.Modernization and Bureaumtic-Authoritariankm:studies in South American politics, Institute of International Studies, Berkeley. ---- and Schmitter, P, 1986. TrdnsitionsfiomAuthoritarian Rule,John Hopkins University Press, Baltimore. Ockey, J., 1996. 'Thai society and patterns of ~oliticalleadership', Asian Survg, 36(4):345-60. Overholt, W;, 1987. 'Pressures and policies: prospects for Cory Aquino's Philippines', in C.H. Lande (ed.), Rebuilding a Nation: Philippine challengs and Amm'can policy, Washington Institute Press, Washington, DC:89-110. Pabuayon, Isabelita and Catelo, Salvador, 1985. 'Policy issues in the Philippine sugar industry', in UPLB Agricultural Policy Working Group (ed.), Policy Issues on the Phil;Ppine Rice Economy and Agricultural lid, Center for Policyand Development Studies, University of the Philippines at Los Banos College, Laguna:247-79. Pearson, S., 1998. An Assessment of Rice Policy in Indonesia, USAID draft report No. PO. 497-0-00-99-00006-00, Jakarta. -, Monke, E. and Baulch, R., 1997. 'The cost of rice stabilization under self-sufficiency', in BULOG: 30 years, Government of Indonesia, Jakarta:353-66. Ramsay, A., 1987. 'The political economy of sugar in Thailand', Pacifc Afairs, 60(2):248-70. Ranis, G., 1987. 'Marcos's economiclegacy: problems, policies, and prospects for President Aquino and how the United States can help', in C.H. Lande (ed.), Rebuilding a Nation: Philippine challenges and American policy, Washington Institute Press, Washington, DC:111-22. Riggs, F., 1966. Thailand the modernization of a bureaucraticpolity, East- West Press, Honolulu. Robinson, R, 1986. Idnesia: the rise of capital, Asian Studies Association of Australia, Sydney. Rusastra, Wayan, Suprihatini, Rohayati and Iqbal, Muhamad, 1999. 'The anticipative Sugar Development Strategy facing economic crisis and competitive market', in R. Stringer, S. Bahri, M.S. Pasaribu and I? Simatupang (eds), Indonesia?Economic Crisis: gets on agriculture and policy responses, Centre for International Economic Studies, Adelaide University, Adelaide:255-82. Samudavanija, Chai-Anan, 1990. 'Thailand: a stable semi-democracy', in L. Diamond, J.J. Linz and S.M. Lipset (eds), Politics in Developing Countries, Comparing Experiences with Democracy, Lynne Reinner Publishers, Boulder, Colorado:271-313. Sar Desai, D.R, 1997. Southeast Asia, Past and Presmt, 4th edn, Westview Press, Boulder, Colorado. Shiraishi, T., 1999. 'The Indonesian military in politics', in A. Schwartz andJ. Paris (eds), B e PoliticsofPost-SuhartoIndonesia,Councilon Foreign Relation Press, New York. Schmitter, I?, 1992. 'The consolidations of democracy and representation of social groups', in L. Diamond and G. Marks (eds), Comparative PmpectivesonDemomcy:essaysinhonorofSglmourMartinLipset,American Behavioral Scientist, 35(4-5):422-49. Siamwalla, Ammar, 1975. 'A history of rice policies in Thailand', Food Research Institute Studies, 14(3):23349. 120 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH ----, 1987.The Farmer'sAid FundActof 1974: thegenesis andaftermath, Paper prepared for EDIIKDI seminar on Econo,pic Policy Change and Governmental Process, Seoul. -andSetboonsarng,Suthad,1989.Pade,ExchangeRate,andAgrzgncultural Pricing Policies in Thaihnd, World Bank, Washington, DC. Sicat, G., 1986. A Historical and Current Pmpective on Philippine Economic Problems, Philippine Institute for Development Studies, Manila. Silcock, T.H. (ed.), 1967. Thailand: Social and Economic Studies in Development, Australian National University Press, Canberra in association with Duke University Press, Durham. Simatupang, Panjar, 1999. 'Toward sustainable food security: the need for a new paradigm', in R. Stringer, S. Bahri, M.S. Pasaribu and P. Simatupang (eds), Indonesia1 Economic Crkis: efects on agriculture and poliy responses, Centre for International Economic Studies, Adelaide University, Adelaide:141-68. Suvarnsit, Renoo, 1971. The Summary of the Third National Economic and Social Development Plan, Plenary Session of the National Economic Development Board, Bangkok, 17 September. Tabor, S. and Meijerink, Gerdien, 1998. 'Food security and trade liberalization: can opposites attract?', in BULOG 30 years, Government of Indonesia, Jakarta:367-98. Tirnrner, I?, 1993. 'Rural bias in the East and Southeast Asian rice economy: Indonesia in comparative perspective', Journal of Development Studies, 29(4):147-76. -, 1996. 'Does BULOG stabilize rice prices in Indonesia?Should it try?', Bulletin of Indonesian Economic Studies, 32(2):45-74. ----, 1997. 'Building efficiencyin agricultural marketing:the long-run role of BULOG in the Indonesian food economy', J m m l ofIntermtioml Development, 9(1):13345. Tolentino, B., 1994. Institutions vs. Implementation of Philippine Economic Reforms Under Aquino, 198692, Discussion Paper Series No. 94-01, Philippine Institute of Development Studies, Manila. -, 1998. Capacity Building and AgrzgrzculturePoliy and Phnning, Policy Paper, Department of Agriculture, Asian Development Bank, Manila. Unger, D., 1998. Building Social Capital in Thaihnd: fibers, finance and injastrueture, Cambridge University Press, Cambridge. University of Indonesia, 1998. Revaluation of Indonesian Sugar Policies, Working Paper, LPEM, University of Indonesia, Jakarta. Vatikiotis, M., 1998. Indonesian Politics Under Suharto: orb, development, and pressure for change, Routledge, London. Warr, l?, 1999. 'Indonesia's crisis and agricultural sector', in R Stringer, S. Bahri, M.S. Pasaribu and l?Simatupang (eds), Indonesid?Eonomic Crisis: #em on agricultureandpolicyresponses,Centrefor InternationalEwnornic Studies, Adelaide University, Adelaide:23-5 1. World Bank, 1984. The Phil+pines Agriculture:issues inpricingpolicy, World Bank Country Sector Report, World Bank, Washington, DC. -, 1993. TheEatAsian Miracle:a@cy researchrtport,Oxford University Press, New York. ----, 1996. Indonesia: a review of public agricultural expenditures, Draft, 21 November, Agriculture Operations Division, Country Department 3, East Asia and Pacific Region, World Bank, Washington, DC. Wurfel, D., 1988. Filipino Politicr: developmentand decay, Cornell University Press, Ithaca, NY. Wyatt, D., 1984. Thailand:a short history,Yale University Press, New Haven. QUOTED INTERVIEWS Indonesia Mr Bustanil Arifin, Chairman of BULOG, 1972-93 Mr Frans Seda, former Minister of Finance and Minister of Agriculture Dr Sayogyo, senior researcher in the Institute of Agriculture (IPB), Bogor Mr Sudaryanto, Deputy Director of BPN (National Land Agency), 1998 The Philippines Ms Marinela Castillo,Executive Director, &idturd Credit Policy Council, Department of Agriculture Mr Lim, Under-secretary of Agriculture, 1980-86 Mr k u l Montemayor, Business Manager, Associations of Farmers Mr Jose Maria Zabaleta, Executive Director of the Philippine Sugar Miller's Association, Manila, October, 1998 Thailand Mr Sampop, advisor to the Prime Minster Dr Rungruang Isankura Na Ayutya, Deputy Permanent Secretary, Ministry of Agriculture and Cooperatives 122 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH QUOTED LEGISLATION Indonesia Presidential Decree No. 11411967, 1967 Presidential Decree No. 1111969, 1969 Presidential Decree No. 3911978, 1978 The First Long-Term Development Plan for BULOG (BULOG dalam PJPTI), Jakarta, 1977 Presidential Decree No. 10311993, November 1993 The Philippines Republic Act No. 3452, 14 June 1962 Presidential Decree No. 4, 26 September 1972 Presidential Decree No. 1485, 11 June 1978 Thailand Sugar Industry Act B.E. 2504, 23 November 1961 Sugar Industry Act No. 2 B. E. 2508, 30 April 1965 Sugar Industry Act B.E 2511, 24 December 1968 RURAL ORGANISATIONS AND DEVELOPMENT: THE SOCIAL BACKGROUND FOR COLLECTIVE ACTION Shinichi Shigetomi When development of the rural sector persistently lags far behind that of the urban sector, deliberate actions can be taken to enhance rural sector development. In a broad sense, such actions can be considered 'rural development policy'. An important strategy in the effective and efficient implementation of rural development policies is mobilisation of the rural population in rural organisations. Rural organisations can make significant contributions to rural development in at least four ways.' They can act as receivers and distributors of resources and information from outside agencies. It is quite difficult for a government to individually contact rural dwellers. They can summarise the needs of rural people and deliver such information to outside agencies in the form of policy menus. They can mobilise resources held individually and communally by the rural population, lessening the burden on outside supporting agencies and increasing the self-reliance of rural development projects. They can serveas tools that empower the rural population. By organising themselves, people may gain the power to change their conditions and enhance their well-being 124 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Thus, rural organisations can be used to identify effective rural development policies, help eficient policy implementation, and help to make the development process persistent (Oakley et al. 1991). In spite of continuous government efforts, however, organisation of rural people has not always been successful (Esman and Uphoff 1984; Cernea 1985; Yogo 2000). An example of failed rural policy can be seen in the formation of rural cooperatives. Although the idea of cooperatives was introduced to developing countries nearly a century ago, they are not well regarded as a means to alleviate rural poverty (Food and Agriculture Organization 1979). It is not true that rural people are unorganised. Rather, there are many traditional groupings, including those formed for economic purposes such as mutual labour exchange and mutual financing. In the process of economic development, however, people have been required to form more institutionalised organisations for their economic survival. The organisation of rural populations faces two major difficulties. One comes from the organisational method itself. People form an organisation becausetheysee an economicopportunity,but, unlike in market transactions, people can only benefit as expected when they cooperate with one another (March and Simon 1993). Uncertain about the potential benefits of such organisation, poor people may be inclined to invest less energy and resources than necessary in organisational activities. A second dificulty lies in organising small-scale entrepreneurs such as farmers. Even after an organisation is successfully established, organisers mayface conditionsthat can negativelyaffecttheir efforts. Unlikerelationships within corporations, relationships in rural organisations are horizontal, with management based on mutual agreement among members. Members can survive even if they withdraw from the organisation, however, because they have their own enterprises (for example, small farms or workshops). The organisation cannot rely heavily on coercion to maintain its membership. Economic opportunities and regulations within a rural organisation are not sufficient conditions for its success. In order to guide the members of an organisation, it is necessary to mobilise existing institutions in the society to which the members of the organisation belong. Since 'institutions are the rules of the game in a society or, more formally, are the humanly devised constraints that shape human interaction' (North 1990:3), they reduce the uncertainty of cooperation among an organisation's members. The institution, however, is just a tool used to guide to people. Organisations are never formed without organisers,and those who promote RURAL ORGANISATIONS AND DEVELOPMENT 125 rural organisation should recognise the group of local ~eoplewho may give birth to formal organisations for development activities (Yogo 1985). In any discussion about organising ~eople,two types of organisations emerge. One is the formal organisation that has specitied purposes for rural development, a 'development organisation'; microfinance organisations, joint marketing groups and cooperatives are examples of this type. The other type is the organisation that is not necessarily for development activities but functions as a body that can give birth to formal organisations.We call this a 'communal organisation'; examples include consanguineous groups, neighbourhood organisations and peer groups. The communal organisation also guides the behaviour of members in development organisations because members of the latter are usually also members of the former. It is essential to recognise communal organisations as an underpinning for successful organisation of rural people. This chapter uses microfinance organisations in Thailand, the Philippines and Indonesia as examples to identif) the hnctions of communal organisations and their policy implications for the establishment of development organisations. FRAMEWORK FOR ANALYSIS Definition and analytical framework In this paper, we define 'organisation' as 'a system of consciouslycoordinated activities of two or more persons' (Barnard 1956). We assume two types of communal organisation, the first of which emerges endogenously through long-term and frequent social exchange among a closely related group of people (Blau and Scott 1962). This type has rules and norms that guide member behaviour.The other type is the local administrative body that is exogenously and deliberately formed by the national and local government. It installs institutions to control local people, and such institutions can serve as tools for self-administration. These two types of communal organisation and the relationship between them are unique in each locality Past studies of rural organisations have focused on endogenous rather than exogenous institutions (Cernea 1987; Kent 1981; Uphoff 1985; O'uchi and Yogo 1985). Because many government-sponsored rural organisation progrm have failed, scholars often give poor marks to official bodies of local administration (OaMey and Marsden 1984; Food and Agricultural Organization 1979; Chopra et al. 1990). It is unrealistic, however, for endogenous institutions and exogenous local institutions to 126 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH exist without having anything to do with each other (Midgleyet al. 1986; Uphoff and Esman 1974; Uphoff c. 1986), since the state is always an important actor in rural development, and local administrative bodies affect the success of almost all rural development programs. On the other hand, we choose microfinance organisations as an example of development organisations, not only because they are common in developing countries (Ledgerwood 1999) but also because there have been successll cases in all three countries, varying in both form and origin. We assume that such variation occurs because communal organisations differ from country to counuy. Rural organisations for development can be sustainable when there is both objective feasibility and organisational capability. 'Objective feasibility' includes economic and political opportunities, and implies that organisers can expect economic benefits through organising. Organisational capability implies that the state authority will not prohibit the organisational activity of local people. We assume that objective feasibility alone is not sufficient, and therefore try to identify local people's organisational capability by focusing on communal organisations. Limitations of analysis This chapter focuses on microfinance as a proxy to study the role of communal organisations. Our intention is not to describe microfinance organisations in general or discuss their effectiveness in diierent countries, but rather to focus on particular microfinance organisations formed for rural development projects in each of the three countries and compare them with those in the other two countries. There are of course considerable differences among regional social and economic conditions within each nation, but because this study does not investigate rural communal organisations as a whole in each country, it is enough to understand how the communal organisations in question define microfinance organisations under given regional conditions. If we write 'in Thailand's case', we do not mean to represent the entire country, but rather to distinguish the case from the other two countries. The analysis is based on a field survey conducted in 1999 and 2000. The chapter therefore does not take into account of important changes that occurred in microfinance organisations and local administrations after the survey period. RURAL ORGANISATIONS AND DEVELOPMENT Typology of microfinance organisations Two criteria can be used to categorisemicrofinanceorganisations-the source of funds and the institution used to ensure that members repay their loans. When a group of local people receives money from outside and makes loans to its members, it collects ~ r i n c i ~and a l interest from the members and repays the fund ~roviders,an ~r~anisationaltype we can call 'receivers'. A second source of hnds can be the members themselves, who invest money in an institution they own. The organisation pools the money and lends it to members, in a system we will call 'pool-and-distribution'. The third source of hnds is the organisation itself, which we can call the 'internal common resource'. Of course many microfinance organisations secure their funds from two or more sources, and the characteristics of each nnay depend on the composition of its funding sources. Organisations must ensure that members repay their loans. One approach is to rely on personal social relationships among members. Members may follow the rules of the organisation simply because they do not want to damage relationships with their neighbows, relatives or friends. Each 'peer goup' should be small enough to maintain close relationships. If the organisation has a larger membership, it may be difficult for it to control members by relying only on direct personal relationships. This 'cooperative group' may need collective consent among its members about regulations and their enforcement. If the group is too big, it may not be able to rely on social relationships or a sense of unity among its members. The organisation may have a professionalstaff that deals with member transactions. Borrowers may consider themselves customers of the fund provider rather than members of a rural organisation. This type of transaction is similar to that which takes place in the market, where debt repayment can be secured by the law andlor any other state authority. We have described characteristics that can be portrayed as a matrix of nine boxes (Figure 5.1). One organisation may use two or more institutions to control member behaviour.When such a matrix is applied to microfinance organisations in the three countries,we can summarisethe situationas follows. The dominant type in rural Thailand is a savings group, a pool-and- distribution' and 'cooperative' organisation. There are quite a few 'receiver' organisations.The organiser of savings groups, especially in the Northeast region, is usually the leadership of an administrative village. 128 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH By contrast, in the Philippines (Central and Southern Luzon), NGOs tend to apply Grameen Bank replicas: typically 'receivers' and 'peer group' types, for their microfinanceprojects.Credit cooperatives,another popular form of microfinance project in rural areas, also rely on outside finding sources. There may be consent among their members, but since most have fewer than 50 members, the members are also united through dose personal relationships. In Indonesia (Central and East Java), NGOs promote 'self-help groups' with memberships smaller than the Thai savings groups but larger than the Philippine Grameen-type groups. These are the 'receiver' and 'cooperative' types of organisation. At the same time, there is a villa e-level bank called Bank Desa in Central and East Java that relies g on its own internal funds. When people organise themselves for a microfinance project, why do such clear differences emerge?We will discuss this issue by focusing on the characteristicsof the communal organisations in each of the three countries' rural society. Figure5.1 Makix for categorising microfinance organisations Major institutions for controlling member behaviour I Law and/or I Collective I Personal social 1 1 stateauthority / consent (relationships 1 Type Organisation Bank Desa Internal $ [Indl common Z, resource M .9 Each member Savings group Pool-and- B tTh1 distribution .g Outside agencies Self-help Grameen bank Receiver 2 groups [Ind] group [Ph] Rural coops [Ph] $pe Bank Cooperative Peer group Notes: Ind = Indonesia; Ph = Philippines;Th Thailand. = Source: Prepared by the author. RURAL ORGANISATIONS AND DEVELOPMENT 129 THAILAND In rural Thailand, the savings group is the dominant form of microfinance organisation. This is a typical pool-and-distribution organisation based on the collective consent of its members. The organiser is usually a village community, especially in Northeast Thailand. L o d administrative system in rural areas Figure 5.2 approximates the structure of local administration in Thailand. Administrative levels down to the district are part of the central government administration, while sub-districts (tambon) and administrative villages (muban) come under the administration of village representatives. The average size of an administrative village was 144 households or 746 persons in 1990, whiie a sub-district had about 1,300 households or 6,700 residents (National Statistical Office 1991, 1992). The administrative village exhibits two major characteristics. The first is that priority is given to demarcation of administrative villages according to the indigenous residential pattern of local people. Since introducing this local administrative unit in the early twentieth century, the government has spontaneously united settlements into administrative villages wherever possible (Tej 1977). The size of these villages is based on the area over Figure5.2 Local administrative hierarchy in rural areas of Thailand National government Government agencies District (ampboe) ........................ I I Sub-district (tambon) Administered by representatives of Administrative village local people (muban) - Source: Prepared by the author. 130 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH which people can maintain acquaintances. The second major characteristic is that village headmen have long been elected by the villagers, thus they reflect relationships among the villagers-without a certain degree of influence among villagers, few have been able to become headmen. Administrative villages in Thailand have thus been formed mainly from considerations of unity and social relationships among residents. Especially in the Northeast and Upper North where homes tend to be built in clusters, settlements have often been automatically made into administrative villages. In contrast,sub-districts were allowed to exercisevery little self-governance up until 1994.On average, each sub-districthas more than 1,000 households and 6,000 people, making it difficult for residents to have face-to-face relationships outside village boundaries. As an administrative unit, a sub- district had a council consisting of a chief, assistants nominated by the chief,and village headmen, but no regular budget for administration. Because the administrative villages were the official inits of local administration under the sub-districts, it was they that implemented most government projects. As a result, sub-districts served as a 'pipeline' through which government resources flowed to villages. Since 1994, however, many sub- district councils have been converted to Tambon Administrative Organizations (TAOs) with their own budget, assembly and staff. Structure and functions of village communities3 Endogenous village communities in Northeastern Thailand have a number of defining characteristics. Nearly every village settlement has its own shrine for a guardian spirit to protect the entire settlement from the evil spirits believed to inhabit the surrounding forest. A worshipping ceremony for the guardian spirits is performed each ye-. Unhappy events that afflict the entire village, such as sickness or drought, are often attributed to the withdrawal of protection by the guardian spirit due to acts of sacrilege by villagers. To organise collective actions to protect the village as a whole, the villagers must accept their common responsibility as residents and define the range of people who should enjoy collective protection,a demonstration of the close and integrated relationshipsat the village level. This indigenous village is distinct fiom the administrative village (mubdn) mentioned above. Of the collective actions undertaken at the level of the indigenous village, the most frequent relates to the Buddhist temple.* In order to construct and maintain temple facilitiesand hold festivals, at certain intervals villagers have to organise to gather and manage resources (materials, money and RURAL ORGANISATIONS AND DEVELOPMENT 131 labour) (Tambiah 1970). They also sometimes organise collective activities to manage common natural resources such as swamps. In Northeastern and Northern Thailand, the administrative village and the indigenous village tend to share boundaries. In the indigenous village, people have a feelin of unity and have accumulated the experience of g collective activities through temple affairs. The administrative village, on the other hand, is a formal institution for collective decision making and receipt of outside resources. Villagers may use the institution of the administrative village to organise activities related to the indigenous unit, such as temple activities, while the administrative unit can mobilise the sense of unity at the indigenous level. If an outside agency proposes a development project, the administrative village can accept it, start organisinglocal people, and supervise management of the organisation. Characteristics of microfinance organisations in rural Thailand The first microfinance organisation in Thailand was a credit cooperative started in 1917. Until the 1960s, each cooperative was formed by a small group of neighbours, with average membership not exceeding twenty (Shigetomi 1998b). The funds for each cooperative came from the government; when the number of cooperatives increased, most failed to secure loan re~ayments.~The government then scrapped credit cooperatives and built agricultural cooperatives (Sahakon 1963; Ekathat 1964). Each cooperative covers one district and secures funds fiom the government and banks, making them semi-governmental organisations rather than local people's organisations. Thus, 'receivers' and 'peer group' microfinance organisations were not successful in rural Thailand.6 In the mid 1970s,savings groups emerged. In this type of organisation, members pool their savings regularly, usually once a month. On entry, each member promises to pay a certain amount of money each month (compulsory savings). Deposits are loaned among members, in most cases at an interest rate of 2per cent per month. Profits are shared among members according to the size of their savings. Most savings groups can earn more than the fixed deposit rate at commercial banks. Some savings groups were established to provide special loans for medical emergencies and treatment, and to make donations for member funerals.These loans require guarantors rather than collateral, but the loan amount often exceeds the deposit amount of the borrower and his or her guarantors. Some groups require collateral 132 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH when a large amount is borrowed. Members are limited to those living in the same locality, mostly in the same village, for convenience and control. Management is provided by savings groups leaders. Savings groups have been promoted by both the government and non- governmental ~r~anisations,and their numbers have increased constantly (Table 5.1). In 1997, nearly 20 per cent of villages had savings groups, while the figures in the Northeast are even higher (Table 5.2). The savings group emerged out of a demand for loans not covered by other financial Table 5.1 Development of savings groups in Thailand Provinces Deposits1 with savings Savings Total Deposits Members1 group groups groups members ('000 baht) group ('000 baht) 2 2 394 137 197 69 20 34 2,283 667 67 20 23 54 2,411 1,173 45 22 26 213 9,136 2,487 43 12 61 639 21,459 4,788 34 7 69 1,093 64,614 21,130 59 19 71 1,345 67,783 33,302 50 25 71 1,584 81,591 44,088 52 28 71 1,960 110,357 73,001 56 37 72 2,821 160,055 113,864 57 40 3,411 196,404 142,830 58 42 4,319 278,799 200,133 65 46 6,048 352,206 254,560 58 42 7,167 408,646 317,372 57 44 7,670 448,827 341,343 59 45 8,156 487,601 403,527 60 49 9,099 588,193 610,434 65 67 9,177 595,890 721,412 65 79 9,868 672,950 915,667 68 93 9,949 765,168 1,326,696 77 133 10,284 799,269 1,069,663 78 104 11,248 884,437 1,837,690 79 163 11,881 937,085 2,157,963 79 182 11,863 971,743 2,391,397 82 202 aAll data are for year's end except 1997, which is for May. Source: Thailand, Ministry of the Interior, 1997. Klurn omsapphua kanphalit:khothet ching, khornun, kho sananae [Saving Groups for Production: the facts, data and policy proposal], Community Development Department (CDD) (Krom kan phatthana chumchon), Government of Thailand, Bangkok. RURAL ORGANISATIONS AND DEVELOPMENT 133 institutions. It provides small loans at lower interest rates than, informal loan sources, and in most cases it does not require collateral. The procedure for borrowing money is very simple compared to that used by financial institutions. It is convenient for villagers to save their small sums because they can make deposits within the village, and the interest rate is higher than that of commercial banks. Why savings groups? After the failure of cooperatives as a microfinanceinstitution,the government established the Bank of Agriculture and Agricultural Cooperatives (BAAC), which provided small and short-term loans directly to farmers, without collateral and at a low interest rate (around 12 per cent per year). However, loans from BAAC were limited to agriculturalpurposes,and the complicated application procedure prevented many farmers from borrowing small amounts from it, a situation that encouraged moneylenders. Moneylenders Table 5.2 Number and size of savings groups in Thailand by region (includesonly groups under the supervisionof the CommunityDevelopmentDepartment, c. 1997") Villages with Number of groups groups (percent of total villages) Lower North 824 15.9 Central 1,035 16.6 East 665 15.5 West 542 11.9 Lower Northeast 3,136 21.6 Upper Northeast 2,747 25.1 Upper North 940 16.1 Upper South 673 17.4 Lower South 615 15.8 Whole Kingdom 11,177 18.9 aThe collection date for these data is unknown, but it is thought to be around 1996 or 1997. Source: Thailand, Minisuy of the Interior,1997. Mum omappbua kanpbalkbo tbet cbing, kbomun, kbo samnae [SavingGroups for Production: the facts, data and policy proposa]], Community Development Deparunent (CDD) (Krom kan phatthana chumchon), Government of Thailand, Bangkok:5-8. 134 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH did not require collateral, but their interest rate could be as high as 4-5 per cent per month. This situation was conducive to the emergence of savings groups. Most members of savings groups in rural Thailand are poor, hence their deposits are small. The membership of savings groups, however, must be large enough to support an effective organisation. As a result, the average membership of savings groups exceeds 80, making them larger than credit cooperatives (Table 5.2). Villagers are the organisers. Even though most of the groups are formed by persuasion from governmental and non-governmental agencies, a savings group cannot be established without villagers' motivation. Afier a group is founded, local leaders become group managers, taking responsibility for all aspects of the operation, including receiving deposits, accounting, lending, collecting debts, and distributing profits. Outside agencies can assist but only to give management advice. Members of savings groups, especially in the Upper Northern and Northeastern regions, tend to live in administrative villages, most of which consist of just 100-200 households, assuring face-to-face relationships among residents. Because the administrative village usually coincides with the indi enous village, there is a sense of unity among people at this level, g making it natural for them to accept a savings group as a village project and to start recruiting members in the same village. Outside agencies also know that the administrative village is the suitable unit for organising people, thus even NGOs usuallyfirst contact the village head and attempt to persuade him or her to organise a group within the village. In this way, villagers in Northeast Thailand combine the benefits of both the administrative and indigenous villages to organise and manage savings groups, their chosen form of microfinance organisation. Savings groups can run into obstacles. When members are recruited from beyond the village boundary, the sense of unity among the people may be weakened and the group's stability threatened. This is the dilemma of size and stability (Shigetomi 1998~).It occurs when the sub-district has not hnctioned or organised local people beyond the administrative village. Since 1994 when the local administrative system was reformed, however, people have had access to collective decision making, common resources, and staff at the sub-district level. Thus, new opportunities exist to change sub-district bodies into communal organisations. RURAL ORGANISATIONS AND DEVELOPMENT 135 PHILIPPINES One of the popular forms of microfinance organisation in the rural Philippines, especially in central and southern Luzon, is the Grameen-type group, a typical 'receiver' and 'peer group' type of organisation. There are also credit cooperatives, but most are very smd and rely on outside funding. Those who try to organise people, such as NGOs, tend to bypass the barangay (the administrative village) and contact local people directly. Barangay as an administrative village The barangay is the local administrative unit that directly governs rural people in the Philippines (Figure 5.3). The barangay as a social unit existed even before the Spanish colonial period. Most were small gatherings of kinspeople with 15-100 households in the inland area of Luzon Uocano 1998; 'Corpuz1997). Outside of individual barangays, there was no political power that integrated barangays (Jocano 1998, 1975; Corpuz 1997). Under the Spanish colonial regime, people in rural areas were forced to gather into pueblo, a unit of local administration consisting of about 500 households (Corpuz 1997). The living quarters of former rural residents were called Figure5.3 Localadministration ladder in the Philippines (barangayand above) President Province Highly urbanised city Municipality Municipal trial court Barangay Barangay BarangayJustice Committee Source:Prepared by the author based on Ocarnpo,Romeo B. and Panganiban, Elena M., 1987. The Local Political System in Asia,Seoul University Press,Seoul. 136 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH barrio (Romani and Thomas 1954). Each barrio had a barrio lieutenant, assigned by the provincial governor. Later, barrio residents gradually moved away from the township and formed a kind of 'breakaway barangay'. Local priests could not neglect such new settlements and took to visiting them regularly. Later, a sinall chapel was built in each of the settlements (Corpuz 1997).Thus, the barrio became an administrative unit of its own. Now the pueblo is called a municipality, while the barrio has been renamed barangay. Although this system survived in the post-independence period, the role of the barangay has changed. The government has increased resources allocated to the barangay, considering it the body most suited to implementing rural development policy (Po 1980).' Barangays now have more authority to govern their residents. Each barangay has a decision- making council; residents directly elect council members and the village headman, the barangay captain. Administrative procedures have been highly formalised. The barangay council enacts ordinances to control or coordinate the conduct of residents. For example, the law mandatesthat a barangayordinanceshould be discussed at three readings of the barangay council, and stipulates what should be done at each readin . Several thick manuals on procedures for barangay g administration are published for village leaders (Ortiz 1996; Ayson and Abelets 1985; Flores and Abeletez 1995). Structure of rural communities In 1995, the average population of a barangay in the Central Luzon and SouthernTagalog regions was about 2,000 people,or 320 households (NSCB 1997).This was more than twice thesize of the administrativevillage (muban) in Thailand. Once established, a barangay tends to keep its boundary regardlessof populationincreases. From 1980 to 1996,the numberof barangays increased by only 5 per cent (NCSO 1981; NSCB 1997). As a result, it is not unusual for barangays in Central and Southern Luzon to consist of thousands of household^.^ In the larger barangays,social relationships among residents have eroded. The same formal system is applied to every barangay, regardless of its size. In contrast, the Thai administrative village tends to be divided if the population becomes too large to keep close contact among villagers. From 1981 to 1996, the number of administrative villages in Thailand increased by 22 per cent (NationalStatistical Office c.1981, 1996). Each barangay has a chapel, which people recognise as the village chapel and where they collectively organise an annual festival to worship the patron RURAL ORGANISATIONS AND DEVELOPMENT 137 saint of the village. Leaders take this opportunity to collect donations for chapel maintenance. Barangay chapels thus resemble temples in rural Thailand. Chapel priests, however, do not live in the village, and villagers' tend to make contributions to their chapel just once a year.9Although a patron saint is assumed to protect the entire village, there is no ritual that connects the existence of a patron saint to the fate of an entire village. Today, the sense of being protected by a patronsaint is weak among villagers. A barangay has fav communal resources. Its popular assets are a barangay hall, a healthcare centre, a daycare centre and a basketball court, none of which requires extensive maintenance. A sense of community possession of these assets is also weak. Thus, communal spirit in the barangay is weaker than in the Thai village. This does not mean that Filipino villagers lack a cooperative spirit, only that they express it differently. Valsan (1970) and Hayami and IGkuchi (2000) describe cooperation between dyadically related people and within small groups. Abueva describes such Philippine communities as 'individualistic and unorganised', and declares that 'their primary attachmentsand loyaltiesare to their nuclear family, their kin and neighbours' (1969:470). Jocano writes that 'outside of the family, the neighbourhood is the only larger social unit which provides the venue for local affairs' (1988:11,93). According to Jocano, 'neighbourhood' in this sense does not necessarily imply a geographical sphere but rather closeness in terms of human relationships. Thanks to the close dyadic relationships common in the rural areas of the Philippines, cooperative actions tend to be organised and carried out smoothly. Characteristics of microfinance organisations in rural Philippines At present, four institutions provide rural microfinance in the Philippines (Agabin 1998). The central bank supervises rural banks and other lenders such as investors and pawnshops. It offers loans to individual customers; clients form no rural organisation. The Cooperative Development Authority (CDA) supervises credit cooperatives and multipurpose cooperatives. NGOs can carry out various kinds of projects fbr rural microfinance. Informal lenders are mostly moneylenders. We focus on cooperatives and NGO programs, since they are the institutions that form people's organisations for project implementation. 138 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Development of cooperatives in the Philippines. In 1952, the Philippine government began in earnest to promote a type of rural financing cooperative called Farmers Cooperative Marketing Associations (Facoma). Because one Facoma covered one municipality and had an average membership of 582 in 1957 (Rocamora and Panganiban 1975), no system of mutual control could operate among members. Many Facoma were used as tools for local politicians. Borrowers regarded loans as subsidies from politicians. Four- fifths of the Facoma operated at a loss and two-thirds were in arrears in 1959 (Po 1980). During the Marcos regime, another type of cooperative was introduced. The Samahang Nayon (SN) was formed at the barangay level with a membership not exceeding 200 (Po 1980). Members created the Samahang Nayon's fund through contributions, making it a 'pool-and-distribution' type of organisation, in which social relationships were expected to control the organisational behaviour of members. One important mission of the Samahang Nayon was to provide loans for farmers to purchase land during the Marcos land reform program, but most were unable to become self- reliant cooperatives (Po 1980). After financial markets were liberalised in the late 1980s, the national government stopped intervening in the financing activities of the private sector, including those of cooperatives (Llanto 1990). The major actors in microfinance are now credit cooperatives and multipurpose cooperatives. In the Luzon area, of 10,328 cooperatives that have savings andlor engage in lending activities, 85 per cent have 50 or fewer members (Table 5.3).1° Philippine cooperatives are generally small. Microfinance organisations under NGO projects. In 1997, the Philippine Coalition for Microfinance Standards (PCMS) surveyed 223 of about 300 NGOs running microfinance activities (Dingcong and Joyas 1998; Agabin 1998).11The survey showed that most NGOs let people organise small groups of fewer than 10 members (Table 5.4). It is difficult for such a small organisation to rely on its own funds for lending activities. Among the 36 active NGOs surveyed in detail by PCMC, 18 use a group lending method and 13 apply a Grameen Bank-like system (Agabin 1998). These data indicate that most people's organisations supported by NGOs in the Philippines are 'receiver' and 'peer group' types. Like the Grarneen Bank in Bangladesh, these NGOs encourage local people to form groups of about five members. The NGOs train members, RURAL ORGANISATIONS AND DEVELOPMENT 139 then loan money to one of each group's members under the condition that the whole group takes responsibility for repayment. The borrower repays the principal with interest according to a weekly plan, and an NGO worker collects the weekly payments and closely monitors the performance of the group. Although voluntary savings are also promoted, the Grameen-type microfinance organisation fiuldarnentally relies on outside money resources. The mutual control that operates in small groups, combined with the close sdpervision of an outside agency, is crucial for success. Table 5.3 Size of cooperatives in Luzon region, June 2000 (active, morethanone member, savings and/or financing activities) Number of Cumulative Size membets per cent 1-30 6,209 60.1 31-50 2,575 85.1 51-100 1,175 96.4 101-1,000 366 100.0 1001-5,000 2 100.0 5001-10,000 0 100.0 10,001 + 1 100.0 Total 10,328 100.0 Source: Calculatedfrom CDA database. Table 5.4 Distribution of differentsized groups under NGO microfinance programs Group size Number of groups Rural-based groupsa 1-9 38 23 10-24 24 19 2549 15 12 50+ 3 3 Total 80 57 aNot less than one-half of activityarea is rural. Source: Questionnaireof inventoryof NGOs promotingsavings and credit.The form is in Dingcong,Clarence G. and Joyas, Lalaine M., 1998. ProfileofPhihppineMimojnance NGOs:a nationwidemrvey, Coalition for Microfinance Standards,Makati. 140 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Why 'receiver' organisations? Even before liberalisation of the financial markets, informal lenders dominated Philippine rural financial markets, meeting an estimated 71 per cent of total demand in rural areas (AgriculturalCredit Policy Council 1991). These lenders usually charged 10 per cent per month for a two- month loan. Although rural banks used to be the main player, after financial liberalisation they became less active in providing credit without collateral. From the late 1980s onwards, then, no agent provided small loans to poor villagers without collateral or a high interest rate. Thus, there was a void to be filled by Grameen-type banks and cooperatives formed by local people. Cooperatives are small. The reason most cooperativesare so small can be found in the source of the funds they draw on for their lending business, as illustrated by some of the cooperatives supported by SACOP (Social Action Center of Panpangas), an NGO in Panpangas.'2 The largest cooperative, Palat Cooperative, has 86 members, but most have fewer than 50 members. Even Palat Cooperative reduced its membership from 100 to 86. The major source of hnds is outside agencies, such as NGOs and politicians. Members have no incentive to increase membership because they fear their own share may diminish. Another successll cooperative (Sta. Monica Farmers' Cooperative)has also reduced its membership, from 72 to 51. According to the leaders, this organisation used to be a farmers' association but changed its status to a cooperative in order to receive more assistance from the government agency, SACOE and local politicians. Although each of these cooperatives covers an entire barangay, they do not expect support from the barangay administration. Larger membership makes management more difficult and reduces the potential share for each member. Dominance of the Grameen type. The popular form of an NGO-supported microfinance organisation modelled on the Grameen Bank was introduced to the Philippines after it had had unsuccessfd experiences with other types of microfinance. For example, CARD (Center for Agricultural and Rural Development), before becoming the initiator of a Grameen-typesystem, formed an association of about 45 landless workers in each target barangay and provided the associationwith a loan. Within eight months of the project's first implementation, however, CARD faced delinquency problems arising from the fact that the group was too big for its leaders to control the members.13 It then applied the Grameen system, and this has been a success because of the smaller size of its groups, which make them easier to control. RURAL ORGANISATIONS AND DEVELOPMENT 141 Such ex erience has important implications. It suggests that a small p group with five members can use social relationships to control one another. In the Grameen system, local ~eoplechoose neighbours as members of the same group. As mentioned above, in Philippine rural society cooperation is more common between people engaged in dyadic relationships than among large numbers of peo le. If a group has five or so members, all may be p connected to one another by cordial dyadic relationships. NGOs understand how to use such social relationships among rural people in order to establish successful microfinance organisations. When NGOs try to organise people, they tend to bypass barangays. In the beginning, they send their workers to the barangay captain and ask him or her to call a meeting about their project. The NGO explains its project and persuades people to form a group.When the NGO has identified those who intend to join the program, it investigateswhether each applicant satisfies the conditions of the program. After a training course, the NGO formally admits the membership and lets them submit a loan application. The NGO provides loans only after the application is approved. In this process, the role of barangay captain is just to call a village meeting. The NGO, not the barangay captain, is the organiser of the local people. This method contrasts with the Thai case, in which NGOs and government agencies expect the vhge headman or other village leaders to play the role of organiser.14 Barangays cannot be used to organise people. Many barangays in Central and Southern Luzon include more than 400 households, hence social relationshipsamong the villagers tend not to be dose. No sense of unity has formed at the village level. Historically, the barangay as a local administrative body was formed in order to meet the needs of rulers rather than out of local people's need for self-governance. Barangay chapels, unlike temples in Thailand, do not induce a sense of affiliation or encourage collective action among the villagers. Rather, the barangay has been used by politicians to theirown politicalends,for example togather votes or legitirnisetheirauthority. INDONESIA In Central and East Java, we find two types of microfinance organisation unlike any to be found in the other two countries. One is the self-help group of 20-30 members, smaller than the Thai savings group but larger than the Philippine Grameen-type group. The other is a bank based on and managed by the administrative village, or desd. 142 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH The desd as administrative village15 Figure 5.4 illustrates the hierarchy of local administrationin rural Indonesia. The desa and units under the desa are administered by representatives of local people, while the other administrative bodies are agencies of national and regional government. Because the local government screens candidates for desa chief before elections take place (Schiller 1996),local figures whose attitude towards the government is unfavourable cannot become chief.16 The chief of the dwun is nominated by the desa chief and designated by the district chief. It is impossible, however, to ignore local people when appointing a dusun chief, since his or her duties are closely related to the daily life of local people.'' The chiefs of units under the dusun, the mkun warga and mkun tetanggd, are decided by consultation among residents. Thus, although the desa is regarded as the main local government unit in rural areas, the government maintains a broader system of intervention through assignment of desa chiefs, who are in a strong position to determine the assignment of leaders of lower administrative units. Desas in Java are quite large. In 1995, the average desa had 1,100 households and 4,600 residents (BPS 1995, 1998),and desas with more than 2,000 households were not rare. Even a dusun generally has more than 200 households, and dusuns with more than 400 households are not unusual. The mkun warp is often the same size as the dusun. To compare Indonesia and Thailand, the Indonesian desa tends to be the same size as the Thai tambon (sub-district), however the dusun is still larger than the muban (administrative village). On the other hand, the mkun tetanggd is smaller than the muban, usually with about 40 households. Different local administrative units have different functions. The desa represents the local people to the outside and contains institutions for decision making-the Lembaga Musyawarah Desa (LMD, Village Consultative Council)-and policy implementation-the desa secretariat and the Lembaga Ketdhanan Mayarakat Desa (LKMD, Village Council for Development Planning and Guidance). The Lembaga Musyawarah Desa includes dusun chiefs and the leaders of other local social organisations. Under the Suharto regime, because the members of both the Lembaga MusyawarahDesa and the Lmbaga KetdhdnanMayarakat Desawere assigned by the desa chief, ordinary local people had little opportunity to express their opinions about village administration. Below the level of the desa, no formal system of consultation and decision making exists for residents. If a chief feels the need to consult with residents, he or she invites them to an RURAL ORGANISATIONS AND DEVELOPMENT 143 Figure5.4 Structureof localadministrationfor rural finance organisationsin Indonesia National government I Province (Propinsz] I Regency (Kzpupaten) I District (Kechamatan) I rukun tetangga (RT) Source: Prepared by the author with referencesfrom Morita, A., 1998. 'Indoneshia' (Indonesia), in A. Morita (ed.), Ajia no chiho seiab,Universityof Tokyo Press, Tokyo:167- 93; and Kosuke, M., 1998. 'Indoneshia ni okemsonraku gyosei soshikito jumin soshik- Nishi jawa puriangan kochi noson no jirei' (Rural administrativeorganisations and people's organisations in Indonesia:a case study from the Priangan Highlands,West Java), in H. Kano (ed.), Tonan djianoson hattenn no shutaito soshiki-Kinhi nihon tono hikaku kara, Institute of Developing Economies, Tokyo:222-56. informal meeting. In some villages, there is a tradition at the dusun level to have annual meetings and meetings every 35 days. Each desa has its own budget. In the provinces of Java, desas secure 70 per cent of their total revenue from internal sources. Most of this revenue comes from communal land and various kinds of fees (Kano 1987). In addition to land, each desa has healthcare centres, a desa hall and other resources. Thus, a desa is not just an agent of central government but also an autonomous body that holds its own resources. Administrative units under the desa have no such autonomous budget or assets apart fiom the communal cemetery.18Although ofElcially a desa has no judicial authority 144 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH it is not rare for villagers to sanction unofficially those who violate the peace and order of the 10cality.'~ To sum up, administrative units in rural Indonesia have the following characteristics. The desa has its own resources and a formal institution for decision making. The desa has the state-backed authority to regulate its residents; however, it is not well suited to consultation with residents. The desa is too big for all its residents to maintain face-to-face relationships with one another. Units under the desa have few resources or institutions for collective decision making. Some dusuns and ruhn warga are still too big for residents to have a sense of unity; at the rukun tetangga level, however, close social relationships do seem to exist among residents. Endogenous social organisations in rural Java According to Kishi (1967), in colonial Java the desa was an indigenous settlement unit, a community that worshiped a village spirit and had communal lands. During the late nineteenth and early twentieth centuries, however, the Dutch consolidated these indigenous desas into new and larger administrative villages (also known as desas), and in 1906 passed a law that transferred the communal properties of the indigenous villages to these new administrative villages. Thus, the present desa did not originate in the indigenous culture. Kano (1990),comparing the historical records of Central Java in 1846 with the results a field survey he carried out 1987, finds that villages with fewer than 100 households in 1846 were five times larger in 1987. He also shows that all land in two villages included in a Central Java land title survey carried out in the 1860s was communal (Kano 1990). In earlier times, each village had a shrine for worshipping the village ancestor, held village festivals, and experienced a sense of unity among residents oay 1969; Boomgaard 1991). Leaders tended to be descended from the earliest settlers of the area, and descent usually legitimised their governing position in the village (Kano 1990; Boomgaard 1991). These communities' strong internal bonds, communal resources and system of self-governance were all integrated into the administrative village, the present desa. Kano assumes the dusun to be equivalent to the indigenous desa, however the community's dramatic increase in size has destroyed any social unity that might once RURAL ORGANISATIONS AND DEVELOPMENT 145 have existed, and there is now no formal unit that coincides with the indi enous village community. g Besides these local administrative units, people in rural Java form many other sorts of organisation, such as groups for literacy education, Koran reading circles, cemetery maintenance teams, singing groups, dance clubs, sports groups, organisations related to mosques, zakat (donation) organisations, and savings groups.'O Nearly every village has arisan (a kind of rotating savings and credit association, or ROSCAS), not only to deal with money, but also to procure goods, such as chickens for a party. People in rural Java emphasise the social aspects of arisan." Another type of traditionalmicrofinanceactivity is simpan- pinjam,a kind of savingsgroup.This also has an aspect of asocial organisation because deposits are distributed for feasting and celebration at the end of the fasting month of Ramadan. Such group activities have been reported by many researchers, including Soemardjan (1963), Kawagoe et al. (1992), and Shimagami (2001). Characteristics of rnicrofrnance organisations in rural Indonesia The government of Indonesia has promoted cooperatives and established several microfinance institutions. Under the Suharto regime, village unit cooperatives (Kopuasi Unit Desa, KCTD) were introduced as a district-level financial institution for farmers (Thorbecke and van der Pluijm 1993). However, the KUD did not succeed in coordinating with grassroots community organisations,and by 1983 about half of all their loans were in arrears (Soemardjan and Breazeale 1993; Kern 1986). Small group formation by NGOs. NGOs working in rural Indonesia have emphasised s m d microfinance organisations. In most cases, a group has 20-30 members, making it larger than the Philippine Grarneen-type group but smaller than the Thai savings group. For example, Bina Swadaya, one of the largest rural development NGOs in Indonesia, has been helping local people to form self-help groups for microfinance since 1988 (Seibel and Parhusip 1992). In this project, Bina Swadaya borrows money from commercial banks and lends it to self-help groups. The commercial banks feel secure about repayment because the NGO takes responsibility for delinquency According to Bina Swadaya, many microfinance groups have been formed endogenously by local people. Bina Swadaya assesses such groups and 'shortlists' them for loans, with groups of 15-25 member$ usually receiving priority. According to other researchers (Geertz 1962;1 146 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Kawagoe et al. 1992), most arisdns have 10-30 members, suggesting that Bina Swadayafollows an organisationalsize that local peoplefeel is adequate. Bina Desa, another Indonesian rural development NGO, began setting up microfinance organisations in the 1970s. During the 1970s and 1980s, Bina Desa encouraged villagers to form self-help groups of 20-25 members and provided these groups with funds. It organised local people not through the desa administration, but by contactinginformal local leaders, who talked about the economic problems of villagers during a mdjeIis takIim (a meeting for Islamic religious activities) and persuaded villagers to form groups for economic activity. Self-help groups tend not to be large in size. Accordingto the community organiser of Bina Desa in Desa Jambangan, Ngawi District (Central Java), there are 10 self-help groups in the desa, the largest of which has 64 members. self-help groups in Batang District, Central Java, also tend to have less than 50 members.A self-help group in Desa Ujung Negoro of the same district had 104 members in 1996, but a conflict among members that occurred in the following year caused the group to split in half. These self-help groups strictly limit their size in order to maintain individual benefit shares in resources given by outside agencies and prevent group instability.This tendencyis also seen in cooperativesin the Philippines. In Java, however, the Grameen type is not popular at all. Desa banks. Another microfinance organisation unique to Central and East Java is the Bank Desa (Kern 1986). The Bank Desa was established in 1908 with initial capital from Iumbung desa (village rice bank) 22 or BRI (Bank Rakyat Indonesia, a government bank focusing on rural finance). Bank Desa usually gave loans to villagers without taking collateral. Four villagers were assigned to a management commission (with the chief commissioner being the desa chief), while accounting was done by a specialist who supervised several Bank Desa. Most Bank Desa are now under the supervision of Bank Rakyat Ind~nesia;~and new Bank Desa generally receive initial funds from Bank Rakyat Indonesia. Table 5.5 shows that Bank Desa and Iumbung desa ate concentrated in East and Central Java. According to the database of Badan Pusat Statistik (BPS: Central Board of Statistics), Bank Desa exist in more than 20 per cent of all desas in Central and East Java (Table 5.6). Some Iumbung desd now also function as Bank Desa. Up to the present, the Bank Desa has been based on the unit of the desa, with thedesachiefservingas the Bank Desa's firstcommissionerand assigning RURAL ORGANISATIONS AND DEVELOPMENT 147 two other commissioners from among the villagers. The Bank Desa has a professional accountant, recruited by the local branch of Bank Rakyat Indone~ia.~~The salaries of the commissioners and accountant are calculated from the outstanding loans of the Bank Desa and paid with the profits. The interest rate for loans is 3-4 per cent per month with a repayment period of 3-4 months.25There is no need for customers to prepare collateral or guarantors. Every desa resident can apply for a loan at the Bank Desa. The commissioners know the customers, so no complicated forms need to be fded in to borrow money. The simplicity of the procedure is one of the elements that make it most attractive to local people. However, although the beneficiaries are all desa residents, this is a bank-type organisation and local people are its clients rather than its members. Table 5.5 Distribution of Bank Desa and lurnbung desa in Indonesia Bank Desa Lumbung desa West Java 480 540 Central Java 1,886 1,155 Yogyakarta 119 75 East Java 1,853 1,516 Java, total (except Jakarta) 4,338 3,286 Othet provinces 108 684 Total Indonesia 4,446 3,970 Source: Badan h a t Statistik, 1996. Survey on Klkzge Potency, Badan Pusat Statistik, Jakarta. Table 5.6 Number of desa with Bank Desa and lurnbungdesa in Java, 1996 No. of desa with Bank No. of Desa or Per cent No. of Per cent desa with Per cent lurnbung of total desa with of total lurnbung of total desa desa Bank Desa desa desa desa West Java 631 Central Java 2,354 Yogyakarta 119 East Java 2,379 Source: Badan Pusat Statisti, 1996. Survey on Klkzge Potency, Badan Pusat Statistik, Jakarta. 148 RURAL. DEVELOPMENT AND AGRICULTURAL GROWTH According to Bank Rakyat Indonesia statistics, more than 60 per cent of loans are repaid without delay (Table 5.7). If the amount of loans 'still possible to repay' is included, more than 70 per cent are expected to be collected. Economic opportunities for microfinance organisations The BankAct of 1967 prohibits banks fromlendingmoneywithout collateral (Mizuno 1999). During the 1970s and 1980s, when the government promoted the BIMAS Program, Bank Rakyat Indonesia provided soft loans without collateral through the BRI Unit Desa. However this program ended in the late 1980s because of delinquency problems. Kredit UmumPedesaan (KUPDES),a general rural credit program started in 1984 by Bank Rakyat Indonesia, provides low interest loans but requires collateral (Patten and Rosengard 1991). A group liability system also exists, but in most cases the group tends to dissolve itself if a member does not meet repayment obligations (Patten and Rosengard1991). Land ownership is usually regarded as collateral, but most rural land in Java has no title (Mizuno 1997).26 Another option is moneylenders, who do not require collateral. According to Mizuno (1999), moneylenders charge interest rates as high as 25 per cent per month or 300 per cent per year. We can conclude then that most of the rural population has no access to low-interest loans, and that this may be one of the reasons people have formed traditional mutual loan and savings systems such as arkan and simpan-pinjam. However, these organisations do not seek continuity or an economic scale. In the case of arisdn, members cannot predict when they might receive a loan. Consequently, there exists a void that is not filled by Table 5.7 Repayment of loans from Bank Desa inJava, December 1999 (per cent of total loans) Delayed, but still Repayment No delay possible to repay doubtful In arrears West Java 31.3 24.I 10.6 34.0 Central Java 59.7 11.8 6.3 22.2 Yokyakarta 65.3 6.1 16.7 11.9 East Java 71.1 9.6 3.9 15.3 Total 65.8 11.0 5.1 18.1 Source: Bank Rakyat Indonesia. RURAL ORGANISATIONS AND DEVELOPMENT 149 governmental financial institutions, private moneylenders or traditional organisations.Thus, the NGOs fd this void and try to organise local people. Why two contrasting organisations-desa banks and small membership groups? There are two types of microfinance organisation unique to rural Central and East Java; the Bank Desa, a bank-type organisation based on internal common resources, and the small organisation of 20-30 members that is financiallysu orted by NGOs and in most cases based on social groups of pp neighbourn. The latter type is formed and managed without relation to the desa or any lower administrative unit. Whyis the Bank Desa successful?The reason it can recoup more than 70 per cent of its loans is that the commissioners, especially the desa chief, have the motivation and authority to secure repayment. Since the Bank Desa loan fund is the common property of the desa, its chief has more incentive to collect it than if it originated elsewhere. Moreover, 20 per cent of the Bank Desa's profit goes into to the desa's general budget. In regards to authority, that of the desa chief is also backed by the state. The accountant also plays a crucial role in the sustainability of the Bank Desa, because he or she has a solid educational background and training in professional accounting. The accountant is not hired by the desa, making him or her independent from the desa chief. These conditions allow the accountant to maintain high standards. Why are small-scale (20-30 member) organisations popular in NGO microfinance projects? In this regard, it is important to take note of the strong tradition of group formation in rural Javanese society. There already exist many endogenous organisations (for example, arisan and simpan- pinjam) that are founded on the principles of collective consent and clear membership. Many such organisations are based on locality and have 20- 30 members. Moreover, throughout rural Java there are many informal leaders, such as Islamic religious leaders. The fact that Bina Swadaya had to 'shortlist' endogenous groups to form self-help groups is evidence of the richness of rural organisations in rural Java. These small organisations do not intend to expand their membership because there is no economic incentive to do so. On the other hand, Grameen-type organisations have not become popular in rutal Java because conditions in the area facilitate the formation of larger organisations, making five-member groups unnecessary. 150 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH These two types of microfinance organisation can serve to illustrate a salient feature of communal organisations in rural Java. There are various kinds of small groups based on closesocial ties among neighbours, suggesting that rural Javanese are experienced in organising themselves. On the other hand, the desa is an example of people's capacity to manage bank-type organisations; however, it has not succeeded as an organiser of local people. Other administrative units (for example, dwun and rukun tetanga) have also failed to make the most of existing social organisations and local people's organisational capacity. Thus, there seems to be a mismatch between endogenous and exogenous systems of organisation in rural Java. CONCLUSION Differences among microfinance organisations In this chapter, we have seen that each of the three countries has some popular and successful types of microfinance organisation. Interestingly, they show considerable differences in their formation and organisation (Figure 5.1). In Thailand, a popular form of microfinance organisation is the savings group, a 'pool-and-distribution' type of organisation with on average about 80 members, whose behaviour is guided by their collective consent to certain rules. The administrative village is usually the organiser of the savings group in Thailand. In the Philippines, a Grameen-type organisation is the most popular form. These are typical 'receiver' and 'peer group' types of organisation, and each group is very small, with about five members. Personal social relationships are the key element that secures repayment, but the strong initiative and supervision provided by NGOs is also indispensable to these organisations' success. Most rural cooperatives limit their membership to no more than 50 in order to ensure continuing close relationshi s among members. In Indonesia, self-help groups are a popular p form of microfinance organisation supported by NGOs. This is a 'receiver' type of organisation with 20-30 members. Self-help groups are not facilitated by the desa administration, but the desa itself manages the Bank Desa, a bank-type microfinance organisation. This summary does not mean that there are no other types of microfinance organisation in the three countries, but that the organisations described here are conspicuous and well suited to comparative study. For RURAL ORGANISATIONS AND DEVELOPMENT 151 example, Grameen-type organisations are hardly seen in rural Thailand, while savings groups are nor popular in Central and Southern Luzon in the Philippines, and villa e banks like the Indonesian Bank Desa are rare in g Thailand and the Philippines. Why do such differences emerge?To answer this question, because the economic opportunities offered by microfinance organisations in the three countries are not very different, we have focused on differencesamong the communal organisationsthat facilitateorganisation among rural people in the three countries. Differences in communal organisations In Thailand, savings groups with many members guided by collective consent can be formed because the administrative village coincides with the indigenous village community. In this setting, village leaders can mobilise a sense of unity among community members and at the same time use the decision-makin institutions of the administrative village. In the past, there g were no communal organisations beyond the administrative village in rural Thailand, thus the economic scope of savings groups was limited. Today, however, the situation is changing, since the government has introduced a new local administrative organisation at the sub-district level. The Grameen-type organisation is dominant in the Philippines because in this country people can only rely on dyadic human relationships for social organisation, and a small number of people in a close relationship is appropriate for the Grameen-type organisation. As an administrative village, the barangay has no indigenous community as a base, thus NGOs tend to bypass it and organise local people directly. In rural Java, people have a long tradition of organising themselves in various kinds of groups, thus they are well prepared to participate in the establishment of new microfinance organisations promoted by NGOs. However, the formal administrative units of the desa and its lower units are not suited to the role of facilitator. Although the desa has its own resources and institutions to control local people, it is too large to mobilise a sense of unity among them. The dusun and ruhn tetangga may be able to mobilise such a sense of unity, since the former can be regarded as the present-day equivalent of the indigenousvillagecommunity and the latter is small enough to maintain face-to-face relationships among residents. What they lack, however, are the resources and institutions needed for communal decision making, since people in rural Java tend to rely on informal leaders and the 152 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH informal process of self-organising rather than formal systems of local administration. This does not mean that the desa has no use as a local administrative body; rather, it shows its capacity to control local people and ensure repayment of loans from the Bank Desa. Policy implications for rural organisation In this chapter our discussion has been confined to microfinance organisations; however, policy implications for the formation of rural development organisations can also be drawn from it. One such policy implication should be of most relevance to practitioners and designers of rural development, especiallythose who want people to organise themselves. In order to ensure the achievement of policy goals, a form of development organisation that suits indigenous organisational patterns must be chosen. While savings groups operate successhlly in Thailand, they should not be introduced to an area that has no local community to act as a facilitator (such as the Philippines). Similarly, Grarneen-type organisations would not operate efficiently in areas where larger collectively bound groups ate easily formed (such as Java and Northeast Thailand). Methods of organising local people in different ateas should differ according the to the features of the ateas' local communal organisations. How then can we identify adequate organisations? For a communal organisation to be suitable as a foundation for a development organisation to build upon, it must have threecrucial elements to guide its members' conduct. The first element is a sense of unity or affiliation to the organisation among local people. Such a shared sense makes people inclined to cooperate with one another (Axelrod 1984). The second element is an institution of decision making for cooperative action-individual needs coordinated towards a common target. Even if people share a sense of unity, they are unlikely to cooperate productively without such an institution. The third element is an institution of enforcement. People cannot have confidence in the organisation unless it has a system of policing its members (North 1990). Organisers of rural people should identify which groups have these elements. In this chapter we have investigated two organisational types- social organisations and local administrative bodies-in relation to their roles in the establishment and management of microfinance organisations. RURAL ORGANISATIONS AND DEVELOPMENT 153 The shape and sphere of social organisations define which members of a society share a feeling of unity, upon which informal rules ancl regulations to control member conduct can be based. Some social organisations also have institutionsfor decision making. On the other hand, local administrative bodies generally have institutions to make common consent among local people and the authority to control people's behaviour. The relationship between local administrative bodies and indigenous social organisations is also important in determining the characteristics of communal organisations. This relationship decides whether or not local people and development organisations can make complementary use of the functions of social organisations and local administrative bodies. If there is a mismatch between the functions of an indigenous social organisation and the relevant local administrative body, there could be a need to reform the local administrative system. This is another policy implication,of particular interest to designers of local administrativesystems. Tne current situation in rural Java implies the need for reformation of local administrative systems. As seen in the reformation of the tambon-level administration in Thailand, a new system can help people organise beyond the administrative village. This should be one of the key goals of any rural development policy-that is, reformation of local administrative systems geared towards helping local people organise themselves. ACKNOWLEDGMENTS I am grateful for help with my rural surveys from Mr. Irdam Ahmad in Indonesia and Mr. Nimrod Dela Pena in the Philippines. NOTES ' 'This summaryfollowsthe discussionof Uphoffand Esman (1974)on the hnctions of rural organisations. The Grameen Bank is a microfinance program that originated in Bangladesh to provide small loans to the local poor without taking any collateral or charging high interest ratq. Five-membergroups have joint responsibilityto repaya loan provided byan outsidesource. This system has realiseda high rate of repayment and has been disseminated to manyother developing countries. Thii sectioil extensively relies on Shigetomi (1998a:Chapter3). * Ninety-five per cent ofThailand'spopulation is Buddhist.The templesare a common sight in rural areas of the country. 154 RURAL DEVELOPMENT AND AGRICULTURtlL GROWTH The repayment rate for loans from the Cooperative Bank during 195660 was just 9 per cent. Only one-half of this was repaid without delay (Thi~~arnondolet al. 1965). The Bank ofAgriculture and Agricultural Cooperatives (BAAC) has formed a small group of neighbours to take responsibilityfor loan repayment. This group resemblesa Grarneen- type group, but the members do not have a sense of unity and actually have no group activityexcept deliveringinformation from BAAC. ' During the Marcos regime, the governmentbroughtmore resources to barangaysthan ever before (Wurfel 1988).Barangays receivedone-quarterof the Special Highway Fund, while the nationalgovernmentprovided20 per cent of the National InternalRevenueTax to the local government. A Barangay Development Fund was also established (1'0 1980). Forexample,BarangayLusacanin Quaon Provincetodayhas morethan3,000households, but it still occupiesthe same territoryit occupied in 1936. In the same province, Barangay Pinagdanlayan (about 400 households) has not changed its boundaryfor 150 years. In contrast, monks live in the compounds of Thai temples, and villagers mobilise their resources daily to ensure the monks' survival. l aAuthor'scalculationsbasedon thelist ofcooperativesinthedatabaseoftheCDA(Cooperative Development Authority, master list of registeredlconfirmedcooperatives, based on the Cooperative Information System (CRIS) database, as of June 2000). " Author's aggregated figures based on the original answer sheets and made with the cooperation of PCMS. l 2SACOP comes under the protection of the regional diocese. It was establishedin 1972 to support the victims of a natural disaster. In about 1986, it began to establish cooperatives to assist poor people. In 1991 Mt Pinatubo erupted and SACOP became an agent that received donationsand distribuced them to victims. As part of its program of assistance, SACOP persuadedlocal peopleto establishcooperativesand otherorganisations (interview with SACOP, Panpangas, September 2000). l 3Interview with villagers of BarangaySantaCruz,LagunaProvince, where the initialproject was implemented. l 4Bypassing the barangay administration is not only the scrategy for NGOs focusing on microfinance activities, but also for rural development NGOs like PRRM (Philippine Rural ReconstructionMovement). Usuallythe PRRM worker stays in the target village to try to gec to know the community's unofficial leaders and learn about the allocation of community resources, then he or she helps people to identify problems needing to be tackled by people working to ether. Thus, an issue-specific group is formed, its main g activitiesable to changeaccordingto local people's needs. This is theway PRRM organises local people and expands village-levelactivities.One exception is the Ihgao ethnicgroup's RURAL ORGANISATIONS AND DEVELOPMENT 155 village, which is characterised by a strong sense of unity focused on the village's leaders. Here, PRRM is reluctant to rely on the barangay administration, since its formal leaders maychange in the next election of barangaycaptain in threeyears. PRRM alsofears that its activity may be misunderstood and associated with the ~oliticalactivities of certain local leaders with connections to local and/or national politicians. ' The descriptiongiven here is based on the situation that existed before reformation of the local administrativesystem began in 2001. l 6Schiller (1996) surveyed 40 desas during 1979 and 1981, and found that 40 per cent of candidates had been disqualified. Even after having been elected, some chiefs whom the government regarded as uncooperative were dismissed from their posts (Schiller1996). l 7In WestJavanese villages,somedusun chiefs wereelected directly by residents. Up until the VillageAdministration Act of 1979, the chiefofadusun was calledpunduhand regardedas the informal leader of the local people (Mizuno 1998:229). I sSometimes local people gather money for communal purposes on their initiative. For example, one dusun in Desa Ngeposari (YogyakattaProvince) bought chairs, glasses, and sound systems using compulsory donations from residents and lent these assets to the villagers. However, these assets are managed by informal community leaders, not by the dusun chief,the formal leader of the dusun. l 9For example,Mizuno reportsthat villagers beat a man who had allegedlyhad asexual &air witha marriedwoman,and that thewomanand herhusband wereforced toleavetheir desa (Mizuno 1998). 20 Based on observations made by the author and by Mizuno (1998). Everyparticipantoffers thesamestake throughoutone rotation, and peopleenjoy drawing lots to decide who will receive the stake (Boumanand Moll 1992; Mizuno 1999). 22 Lumbungdesa was originallya traditionalinstitution formed by localpeople. In 1898, ED. De WOEvan Westerrode, Assistant Resident in Punvokerto, Central Java, set it up as a formalinstitution (Suharto1991). The initialcapitalwasvillagers' paddy.In the beginning, the paddywaslent to thosewho providedthe initialsupply. Later,everyvillagerwasallowed to borrow. Repayment was made with paddy. 2 3Accordingto interviewees inJava,some Bank Desa have been established by the provincial government. 2 4Theauthority of assignmentand dismissal, however,is given to the bupati,or regencychief. Z5 Interviewat Ngawi District, East Java, October 2000. 2 6In a village survey in West Java, only 14 per cent of households had land titles in 1986 (Mizuno 1999),thus many rural populationsare excluded from the formal loan market. 156 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH REFERENCES Abueva, J.V., 1969. 'Local government, community development, and politicalstability', in J.V. Abueva and RE de Guzman (eds), Foundations and Dynamics of Filipino Government and Politics, Bookmark, Manila:468-75. Agabin, M.H., 1998. Philippine Microfinance NGOs: at the crossroads? Developing Standards for Microfinance Project, Makati. Agricultural Credit Policy Council, 1991. Rural Finance Development Phn: an agenda for action 1992-95, Agricultural Credit Policy Council, Manila. Axelrod, R, 1984. The Evolution of Co-operation,Basic Books, New York. Ayson, F. and Abletez, J., 1985. Barangay: its operation and organization, National Book Store, Navotas. Badan Pusat Statistik (BPS), 1995. Statistical Earbook of Indonesia 1995, Badan Pusat Statistik, Jakarta. ----, 1996. Survey on Elhge Potency, Badan Pusat Statistik,Jakarta. -, 1998. Statistical Earbook of Indonesia 1998, Badan Pusat Statistik, Jakarta. Barnard, C.I., 1956. The hnctz'onsof the Exemtive, Harvard University Press, Cambridge, Massachusetts. Blau, PM. and Scott, W.R., 1962. Formal Organizations: a comparative approach, Routledge & Kegan Paul, London. Boomgaard, E, 1991. 'The Javanese village as a Cheshire Cat: the Java debate against a European and Latin American background', TheJouml of Peasant Studies, 18(2):288-304. Bouman, F.JA. and Moll, HA.]., 1992. 'Informal finance in Indonesia', in D.W. Adarns and D.A. Fitchett (eds), Informal finance in Low-income Countries, Westview Press, Boulder, San Francisco and Oxford:209-23. Cernea, M.M., 1987. 'Farmer organizations and institution building for sustainable development', Regional Development Dialogue, 8(2):1-19. -, 1985. 'Sociological knowledge for development projects', in M.M. Cernea (ed.), PuttingPeople First:sociologcalvariablesin ruraldeuelopmmt, Oxford University Press, New York:3-21. Chopra, K., Kaidekodi, G.K. and Murty, M.N., 1990. Participatory Development: people and common properg, resources, Sage Publications, New Delhi. RURAL ORGANISATIONS AND DEVELOPMENT 157 Corpuz, O.D., 1997. An economic history of the Philippines, University of the Philippines, Quezon City. Dingcong, C.G. and Joyas, L.M., 1998. Profile of Philippine Microfinance NGOs:a rutionwidesurvq, Coalitionfor MicrofinanceStandards,Makati. Ekathat, Sawat, 1964. Lak sahakonpraphet tdng tang lae wikbro kan borihan ngan sahakon ha thun ruiprathet thai (phim khrang thi 2) [Principles of Various CooperativesandAnalysisof Management of CreditCooperatives in Thailand (2nd. Edition)], Sarnnak phim sarnakhom sangkhomsat haeng prathet thai, Phranakhon. Esman, M.J. and UphofF, N.T., 1984. Local Organizations: intermedidries in rural development, Cornell University Press, Ithaca. Flores, C.O. and Abletez, J.P., 1995. Barangay: its government and management, revised edition, Manila. Food and Agricultural Organization of the United Nations, 1979. Participation of the Poor in Rural Organization, Food and Agricultural Organization of the United Nations, Rome. Geertz, C., 1962. 'The rotating credit association: a "Middle Rung" in development', Economic Deuelopment and Cultural Change, 10(3):241- 63. Hayami, Y. and Kikuchi, M., 2000. A Rice Ellage Saga: three decddes of Green~ o l u t i oin the Philippines,MacMillan Pressand Barnes & Noble, n London and Boston. Jay, R., 1969. Javanese Ellagers: social relations in ntral Mo4okut0, MIT Press, Cambridge, Massachusetts. Jocano, EL., 1998. FilipinoPrehistory:redis~oven'n~precolonidl heritage,Punlad Research House, Manila. -, 1988. Social Organization in Three Philippine Ellages: an exploration in ntral anthropology, Centro Escolar University, Manila. -, 1975. Philippine Prehistory:an anthropologtcal overviewofthe beginnings of Filipino society and culture, University of the Philippines, Quezon City. Kano, H., 1990. 'Jawa sonraku shi no kensho' [Continuity and change of Javanese villages: the case of Ungaran], Tqo bunka kenbujo kiyo, 11133- 129. -1987.'Indoneshianoseijitaiseitogyoseikiko'[Thepoliticaland , administrativesystem of Indonesia], in Y. Hagiwara and E. Murashima (eds),ASEW shokoku no seiji tizisei, Insititute of Developing Economies, Tokyo: 23-52. 158 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Kawagoe, T., Ohkama, K and Bagyo, A1 Sri, 1992. 'Collective actions and rural organizations in a peasant economy in Indonesia', The Developing Economies, 30(3):215-35. Kent, G., 1981. 'Community-based development planning', The World Phnning Review, 3(3):313-26. Kern, J.R., 1986. 'The growth of decentralized rural credit institutions in Indonesia', in C. MacAndrews (ed.), Central Government and Local Development in Indonesia, Oxford University Press, Singapore:101-3 I. Kishi, K., 1967. 'Jawa no sonraku soshiki ni tsuite no oboegaki-dessa to karurahan ni tsuite' [A note on village organisationsin Java: with a focus on desa and kelurahan], Tyo Bunka, 43:23-48. Ledgemood,J., 1999. Microjinance Handbook:an institutionaladjinanciaL penpective, World Bank, Washington, DC. Llanto, G.M., 1990. 'Agrculd credit under financial liberalization: the case of the Philippines', Studies in Rural finance and Development, l(1):l-27. March,J.G. and Simon, H.A. (with the collaborationof Harold Guetzkow), 1993. Organizations, 2nd ed., Blackwell, Cambridge, Massachusetts. Midgley, J., Hall, A., Hardiman, M., and Dhanpaul Narine, 1986. Community Participation, Social Development and the State, Methuen, New York. Mizuno, K., 1999. Indoneshia no jiba sangyo-Ajia keizai saisei no michi towa naniku? [Community-based Industries in Indonesia: the perspective of revitalising the Asian economy], Kyoto University Press, Kyoto. -, 1998. 'Indoneshia ni okeru sonraku gyosei soshiki to jumin soshiki- Nishi jawa puriangan kochi noson no jirei' [Rural administrative organisations and people's organisations in Indonesia: a case study from the Priangan Highlands, West Java], in H. Kano (ed.), Tnan djid noson hattenn no shutai to soshiki-Kinahi nihon tono hikaku kara, Institute of Developing Economies, Tokyo:222-56. -, 1997. 'Indoneshia ni okeru tochiken tenkan mondai-Shokuminchi ki no kindaiho tochiken no tenkan mondai wo chushin ni' [Conversion of land titles in Indonesia: with special reference to land titles registered in the colonial era], in K. Mizuno (ed.), Tonan djid no keizai kaihatsu to tochishoyu seido, Institute of Developing Economies, Tokyo:115-54. Morita, A., 1998. Indoneshia [Indonesia], in A. Morita (ed.), Ajid no chiho seido, University of Tokyo Press, Tokyo:167-93. RURAL ORGANISATIONS AND DEVELOPMENT 159 National Statistical Office, 1996. Statistical Earbook Thaihnd 1996, National Statistical Office, Bangkok. -, 1992. Statistical Yedrbook Thaihnd 1992, National Statistical Office, Bangkok. -, 1991. Statistical Yearbook Thaihnd 1990, National Statistical Office, Bangkok. -, c. 1981. Statistical Yearbook Thaihnd 197C80, National Statistical Office, Bangkok. NCSO (National Census and Statistical Ofice), 1981. Philippine krbook 1981, National Census and Statistical Office, Manila. NSCB (National Statistical Coordination Board), 1997. 1997 Philippine Statistical krbook, National Statistical Coordination Board, Makati. North, D.C., 1990. Institutions, Institutional Change and Economic Pe$omzance, Cambridge University Press, Cambridge. Oakley, I? et al., 1991. Projects with People: the practice of participation in rural development, International Labour Office, Geneva. Oakley, E and Marsden, D., 1984. Approaches to Participation in Rural Development, International Labour Office, Geneva. Ocampo, R.B. and Panganiban, E.M., 1987. The Local Political System in Asia, Seoul University Press, Seoul. Ortiz, J.S., 1996. The Barangay of the Philippines, 1996 edition, Hiyas Press, Quezon City. O'uchi, M. and Yogo, T., 1985. 'The role of social organizational resources in local-level development and management: synthesis and conclusions', Regional Development Dialogue, 6(1):156-79. I'atten, RH. and Rosengard, J.K., 1991. Props with ProJits: the development of rural banking in Idnesia, ICS Press, San Fransisco. Po, B., 1980. 'Rural organizations and rural development: a documentary study', in Marie S. Fernanda (ed.), h r a l Organizationsin the Philippines, Institute of Philippine Culture, Ateneo de Manila University, Mani1a:l- 123. Rocamora, J.E. and Panganiban, C.C., 1975. Rural Development Strateges: the Philippine case, Ateneo de Manila University, Manila. Romani, J.H. and Thomas, M.L., 1954. A Survey of Local Governmentin the Philippines, Institute of Public Administration, University of the Philippines, Manila. 160 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Sahakon, Phraprakat (ed.), 1963. Lak khong kan sahakon he kan chat sahakon nai prathet thai [Principles of Cooperatives and the Cooperative Movement in Thailand], Minisuy of Cooperatives, Bangkok. Seibel, H.D. and Parhusip, U., 1992. 'Linking formal and informal finance: an Indonesian example', in D.W Adarns and DA. Fitchett (eds),In$ml Finance in Low-Income Countries, Westview Press, Boulder San Francisco and Oxford:23948. Schiller,J., 1996. DeuelopingJepara: state and society in New Or& Indonesia, Monash Asia Institute, Clayton. Shigetomi, S., 1998a. Co-operation and Community in Rural Thaihnd an organizational analysis of participatory rural development, Institute of Developing Economies, Tokyo. ---, 1998b.'Nosonkyodokumiainosonritsu joken:Shinyokyodososhiki ni miru tai to nihon no keiken' [How can rural cooperatives develop beyond the village community?A comparative analysis of Thailand and Japan], in H. Kano (ed.), Tonan ajia noson hattenn no shutai to soshiki- Kinhi nihon tono hikaku kara, Institute of Developing Economies, Tokyo:179-219. -, 1998c. 'Community development and local administrative hierarchy: how can a community-based development break the fetters of community?', in Pmceedings: International Sminar of Swtainable Humn Resource Deuelopment and Community Deuelopment in Asia and the Pac;J;c, 13-15 May 1998, Kasetsart University, Bangkok:42-7. Shimag--mi,M., 2001. 'Jawa noson ni okeru jumin soshiki no inboryushon: Suharuto seiken ka no "sonraku kaihatsu" no ichi sokumen' [Organizational involution in rural Java: a charateristic of 'village development' under the New Order], Tonan Ajid Kenkyu, 38(4):512- 51. Soemardjan, Selo, 1963. The Dpmics of Community Dmelopmmt in Rural Central and WestJava:a comparative report,Monograph Series of Modern Indonesia Project, Department of Asian Studies, Cornell University, Ithaca. -andBreazeale,K.,1993.CulturalChangeinRuralIndonesia:impact of village akuelopment, Sebelas Maret University Press, Surakarta. Suharto, Pandu, 1991. Small or people's credit institutions in Indonesia, Jakarta (unpublished). RURAL ORGANISATIONS AND DEVELOPMENT 161 Tambiah, S.J., 1970. Bdhism and the Spirit Cults in Northeast Thailand, Cambridge University Press, Cambridge. Tej, Bunnag, 1977. The Provincial Administration of Siam 1892-1915, Oxford University Press, Kuala Lumpur. Thailand, Ministry of the Interior, 1997. Hum omsappbuu kan phalit:kho thet ching, khomun, kho sananae [Saving Groups for Production: the hcts, data and policy proposal], Community Development Department (CDD) (Krom kan phanhana chumchon), Government of Thailand, Bangkok. Thisyamondol, Panthum, Arromdee, Virach and Long, M.F., 1965. Agrt'cultural C ~ d iin Thailand the09 data,poliy, Kasetsart University, t Bangkok. Thorbecke, E. and van der Pluijirn, T, 1993. Rural Indonesia: socio-economic development in a changing environment, New York University Press, New York. Uphoff, N., c. 1986. Local Institutiod Development: an amlyticalsourcebook with cases, Kurnarian Press, West Hartford. -, 1985. 'The role of vdage management traditions and institutions in local development', Regional Development Dialope, 6(1):ix-mix. ---- and Esman, M.J., 1974. Local Governmentfor Rural Development: analysis of Asian experience, Rural Development Committee, Cornell University, Ithaca. Valsan, E.H., 1970. Community Development Programs and Rural Local Developmt: comparative case studies of Indid and the Philippines, Praeger Publications, New York. Wurfel, D., 1988. Filipino Politics: developmentand ahay, Cornell University Press, Ithaca. Yogo, T, 2000. 'Local communities in development process: an analytical framework', in Asian Productivity Organization (ed.), Local Social Development, Asian Productivity Organization, Tokyo:17-30. -, 1985. 'An overview of regional development approaches and a hypothesis on the formation of endogenous receiving mechanisms in the local community: a conceptual framework', Regional Development Dialogue, 6(1):1-23. THE POLITICAL ECONOMY OF RURAL DEVELOPMENT IN INDONESIA Toshihiko Kawagoe Despite rapid economic growth over the past three decades, rural development in Indonesia has received both criticism and praise. The rice sector attained self-sufficiency, but only at great cost and after many government failures. While rice production expanded impressively, export crops were almost completely ne lected.This paper seeks to identify political g reasons for these contradictions through an examination of the political economy of rural development and rural development policy in Indonesia, and a review of the development process since the colonial period, a process that has left many cultural, economic and social legacies. We suggest that the ideas of Indonesian policymakers during the early stages of the New Order rule were influenced by the following cultural, historical and ideological factors the idea of nual pessimism in academic thought, represented by Boeke (1953) and Geertz (1963a) anti-colonial sentiments the state ideology of pancaila, with its stress on national unification. Three more ideological and social factors were proclaimed in the 1945 constitution of Indonesia, these were state nationalism as the ideology of an independent Indonesia, rationalising extensive state intervention in the economy a strong belief in the potential for cooperatives to benefit economically weak groups RURAL DEVELOPMENT IN INDONESIA 163 the importance of farming families. Indonesia's rural development polices in general have been characterised by extensive state intervention, ignorance of traditional sectors, and distrust of markets. The rice sector was an exception, since in addition to the hctors noted above it was also characterised by a heavy reliance on cooperatives and distrust of the ability of farmers. In the 1950s and 1960s, academics and policymakers were pessimistic about the rural sector. Industrialism was widely supported while agricultural exports were ne lected, and agriculture was not regarded as a leading sector g for growth. Memories of colonialism were reflected in rural development policies, which tended to regard colonial-era approaches such as free trade, foreign investments and foreign enterprises as evil. Similarly, because they were the dominant industry during the colonial era, export crops were ignored. Rice was an exception. Indonesia attained self-sufficiencyin rice in the mid 1980s, an achievement that allowed the government to save valuable foreign currency, though this was not the sole reason for its attention to the rice sector. For many Indonesians, rice carries great emotional and symbolic weight, being associated with the rural family, whose importance is proclaimed in Indonesia's constitution. Former President Suharto also often made reference to his peasant origins. Although rice has been produced since the colonial era, it was not grown on foreign-owned plantations. Although the rice sector was successful, institutional problems were reflected in policy, and regional diversity was ignored through both bureaucraticformalities and a political focus on the unitary development of the state. Indonesia's constitution states that economically weak groups should organise according to family principles. This ideology, coupled with a distrust of both markets and farmer ability, sets the framework for an Indonesian style of rural development, in which cooperatives play a key role. Though village cooperatives were assigned a key role in the implementation of rice policies, the relevant bureaucracy &led to function properly. Indonesia is a large and highly diversified country, and industry and agriculture differ significantly across regions. Regional differences in income are still large, but are trending toward convergence. An overview of the regional diversity of industrial and agrarian structures is presented in Annex 1 of this chapter. 164 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH INDONESIA'S ECONOMIC PERFORMANCE OVER THE PAST THREE DECADES Over the past three decades, the Indonesian economy has performed impressively. In 1996, per capita income in Indonesia reached US$1,100, substantially higher than in Sri Lanka ($750) and Kenya ($320), and almost the same as in the Philippines ($1,160) and Egypt ($1,08O), though still lower than in Thailand ($2,930).' Indonesia's economic growth was not as robust as Thailand's, but exceeded that of the Philippines and was much better than that of many developing countries in South Asia and Africa. In 1969, per capita GNP in Indonesia had been only $70, substantially lower than in Sri Lanka ($180), Kenya ($130), the Philippines ($250), Egypt ($210) and Thailand ($200). By the mid 1980s, the Indonesian economy had nearly reached the level of the Newly Industrialised Economies, though its impressive development was abruptly interrupted by the financial crisis of 1997. A N OVERVIEW OF ECONOMIC GROWTH2 Macroeconomic growth Beginning in 1966, the Indonesian economy recovered and accelerated its growth in a period known as the New Order, when Suharto was the second president of the country. In a seminar report on the 'East Asian miracle' (World Bank 1994), the World Bank recognised Indonesia as one of the High Performing Asian Economies (HPAE). In the 1970s, Indonesia's real GDP grew at an annual rate of 7.8 per cent and per capita GDP at an annual rate of 5.3 per cent. Although softened world oil markets slowed GDP growth to about 4 per cent in the early 1980s,the economy recovered to an annual growth rate of 6-7 per cent in the late 1980s. This rapid growth lasted for a decade, until the unexpected onset of the financial crisis in July 1997. Before 1997, per capita GDP increased from $190 in 1960 to $707 in 1996 ('In constant 1987 US dollars) (Table 6.1). Indonesia is now categorised as a lower- middle income count$ This is a marvellous achievement if we compare it to other economies in South Asia and Sub-Saharan Africa, but Indonesia's GDP is still mediocre for an HPAE (though it is greater than the GDP of the Philippines, it is lower than that of other HPAEs such as Thailand, Malaysia and Singapore). RURAL DEVELOPMENT IN INDONESIA 165 We should keep in mind that not all regions enjoyed the fruits of economic gowth equally. Indonesia is a large country of great geographical, social and ethnic diversity, and its economic characteristics differ greatly from region to region. In 1995,for example, per capita GDP for the nation as a whole was Rp 2.1 million. Per capita GDP in the poorest region of East - Table 6.3 Changes in economic indicators in Indonesia, 1961-96" Economic GNP per capita (constant 1987 US$) 184b 274 411 707 GNP per capita (annual percentage) -1.2' 5.0 3.3 5.7 Gross domestic investment (% of GDP) 9.2 7.5 23.7 26.7 31.4 Inflation, consumer prices (annual percentage) 72.6 516.4 16.7 6.6 8.7 Value added by sector (% of GDP) Agriculture 51.2 52.7 29.8 23.6 16.7 Industry 14.4 12.4 33.9 35.4 42.1 Services, etc. 34.4 34.8 36.3 41.0 41.2 Trade Exports of goods and services (% of GDP) 10.5 9.4 24.5 22.7 26.3 Merchandise exports by sector (% of total) Agricultural raw materials 31.9 14.2 7.7 6.2 Fuel exports 34.6 71.0 56.8 25.6 Manufactured goods 3.4 1.4 19.1 51.0 Labour force Economically active population (million)* 53.1 57.8 72.7 94.1 123.8 Labour force by sector (% of economicallyactive p~pulation)~ Agriculture 74.4 70.2 61.7 56.4 44.0f Industry 7.6 8.9 11.2 12.8 18.4f Services, etc. 17.9 20.9 27.2 30.8 37.6' " Three-year avemges for the years 1961436. 1996 data are 1995-96 196667 average 1967 growth rate Population aged 1544 1961 and 1966 figures refer to 1960 and 1965 data, respectively.For 1976 and 1986, the figuresare interpolated from the data in census years 1970, 1980 and 1990 'Share of total employment in 1995 Source: World Bank 1998. WorldDmelopmmt Indicators,World Bank, Washington, DC; Manning,C., 1998. IndonesianLabour in Transition:an EastAsian mccessstory?,Cambridge University Press, Cambridge; BPS, Population Census 1990, quoted in Manning (1998:93, Table 4.5). 166 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Nusa Tenggara, however, was only Rp 0.8 million, while the richest region was Jakarta with Rp 7.6 milli~n.~Obviously, national aggregated data hide regional differences, making investigations with a regional perspective important in any analysis of this vast country. Regional issues are discussed in detail in a later section of this chapter. While the major reason for Indonesia's relatively successhl increase in per capita income was the rapid growth of GDE demographic factors also contributed. In the mid 1960s,Indonesia had an annual population growth rate of 2.3 per cent (Table 6.2). By world standards, the fertility rate was high, and the government of the time rejected population control (Hull and Jones 1994:124). From the mid 1970s on, however, demographic factors changed quickly. Birth and death rates declined, and there was a Table 6.2 Changes in demographic and socialindicators in Indonesia, 1961-96" Demographic factors Birth rate, crude (per 1,000 people)b Death rate, crude (per 1,000 people)" Life expectancy at birth, total (years)b Population growth (annual percentage)' Population, total ('000) Urbanisation Urban population (% of total) Social indicators School enrolment, primary (gross per cent)d illiteracy rate, adult total (% of people 15+)' Safe water access, total (%of population)f Safe water access, rural (% of rural population)f aThreeyear averages for the years 1961 to 1986. 1996 data are 1995-96 average Data refer to 1962, 1967, 1977, 1987 and 1996 196142 average Data refer to 1960, 1965, 1975, 1985 and 1994, respectively Data refer to 1961, 1965, 1985 and 1995, except for 1976 which is the average of 1971 and 1980 data 'Data refer to 1975, 1985 and 1993, respectively Source: World Bat& 1998. WorldDevelopmentIndicators, World Bank, Washington,DC. RURAL DEVELOPMENT IN INDONESIA 167 steadyincrease in life expectancy at birth. The population growth rate began to decline substantially,to 1.9 per cent in the 1980s and 1.7 per cent in the 1990s. The figure in the 1990s was still higher than the 1.2 per cent of Thailand, but significantly lower than the 2.3 per cent of the Philippines and the low-income country average of 2.1 per cent (World Bank 1999). In fact, the total fertility rate in Indonesia declined from 6.4 per cent in the early 1950s to 2.8 per cent in the mid 1990s (Caldwell 1997). The success of Indonesia's family planning program is regarded as one of the critical achievements of the New Order government. As H d and Jones (1994) point out, however, institutional changes in marriage systems had already occurred prior to government intervention, and these were also major factors in the fertility decline. Whatever the impact of government policies on fertility, a remarkable decline in both fertility and the population growth rate contributed to the economic development of the world's fourth most populous country. In the course of the rapid economic growth of the past four decades, the industrial structure of the Indonesian economy has changed significantly. In 1953, non-mining primary sectors, including agriculture, forestry and fisheries (hereafter referred to as the agricultural sector), contributed 57 per cent of total GDP (Table 6.3), while manufacturing industries contributed only 11 per cent. The share of the agricultural sector began to decline in the mid 1960s,although it was still as high as 47 per cent in 1970. Drastic change occurred from the mid 1970s on, with the agricultural sector falling to 25 per cent in 1980 and just 16 per cent in 1997, while manufacturing reached 27 per cent in 1997. These figures should be interpreted with caution, however, given the significance of oil and natural gas to the Indonesian national income, which was highly dependent on world oil prices. In the 1970s, high oil prices coupled with increased oil production significantlyincreased the contribution of the mining and quarrying sectors. In 1980, mining and quarrying accounted for 26 per cent of GDE with the shares of other sectors significantly lower. Although the economic structure consistently shifted hom agriculture to industry throughout this period, the shift actually occurred more slowly than reflected in national income statistics. If the contribution of mining and quarrying is removed from the GDP calculation, the agricultural sector's decline can be seen to have been much milder than statistics suggest. From the late 1980s on, manufacturing's share of GDP increased rapidly, while that of mining and quarrying dropped to 10-15 per cent, a change 168 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH - - Table 6.3 Percentage distribution of Indonesian GDPby industry, 1960-96 Agriculture, forestry, and fisheries Mining and quarrying Manufacturing Electricity, gas, and water supply Construction Trade, hotel, and restaurant Transport and communication Banking Other services Annual value (billion R P ) ~ Vreliminary figure At current market price. Source: Badan Pusat Statistik, various issues. Statistical Pocketbook of Indonesia, Badan Pusat Statiscik,Jakarta; Bevan, D., Collier, I?and Gunning, J., 1999. The PoliticalEconomyof Povery, Equity and Growth:Nigeria and Indonesia,Oxford University Press, New York:304 (for 1953); Badan Pusat Statistik, 1968. Almanak Indonesia, vol. 2, Badan Pusat Statistik, Jakarta (for 1960). due partly to lower oil prices and partly to the expansion of the non-oil manufacturing sector. This sector, which excludes petroleum refining and liquefied natural gas, increased from 10-15 per cent in the late 1980s to about 25 per cent in the late 1990s. Again, we should keep in mind that large regional differences greatly affected the economic benefits brought by these developments, as discussed in later sections of this chapter. Transformation of the trade structure The boom in manufacturing exports that occurred from 1985 on was generated by liberalised policies (Hill 1996). Although exports from Indonesia have expanded more than six-fold in real terms over the past three decades, the change in export commodities was more important. Throughout the 196Os, 50-60 per cent of Indonesian exports were agricultural commodities, including sugar, natural rubber, palm oil, coffee, and tea, all of which had been produced since the colonial era.5In the 1960s, petroleum and its products were the second most important export, accounting for another 30-40 per cent of exports. Export of manufactured products was negligible at that time. RURAL DEVELOPMENT IN INDONESIA 169 The trade structure drastically changed in the early 1970s. Exports of agricultural commodities decreased and were replaced by oil products. From 1971 to 1974, export prices of crude oil rose more than sixfold. The sharp price increase, coupled with expanding crude oil production, caused a dramatic rise in petroleum exports during the oil boom. Petroleumaccounted for 70 per cent of export earnings during this period and earned 22 per cent of GDP and 55 per cent of total government revenue (Woo et al. 1994). Petroleum and its products remain important Indonesian exports. In the 1990s, however, earnings from these products declined to 20-30 per cent of total exports. This decline occurred as world oil markets sofiened and non-oil manufacturing exports accelerated. In the 1990s, the non-oil manufacturing sector earned more than the oil and agricultural sectors. Among the manufactured exports, however, exports based on agriculture- such as plywood, processed rubber, palm oil and processed food-played an important role. A series of structural adjustments in the 1980s enabled Indonesia to shift its economy from its heavy dependence on oil. Importance of the rural sector Although the relative position of the agricultural sector has declined significantly over the past four decades, its importance to the Indonesian economy has not decreased. Over the past three decades, the distribution of labour has changed, with the share of agricultural labour in total labour Ming from 72 per cent in 1961 to 50 per cent in 1990, and the shares of industrial and services labour increasing from 8 and 19 per cent in 1961 to 12 and 33 per cent in 1990, respectively (Hill 1996). The per cent decline in agricultural employment was slightly lower than that of agricultural GDC indicating a widening gap in labour productivity between farm and non-farm sectors. The rural sector, however, is still very important for employment and income, especially for the rural poor. According to the 1990 census, nearly 70 per cent of the population resided in rural areas. Even in Java, the most urbanised area of the country, 64 per cent of the population was rural. In 1961, the rural sector was much more dominant, with 85 per cent of the population residing in rural areas and more than 72 per cent of the labour force engaged in agriculture. Urban manufacturing employment was negligible, with only 2.4 per cent of the total labour force, while 3.5 per cent were employed in rural manufacturing a~tivities.~ 170 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH In 1971, even after a decade of steady economic growth, over 60 per cent of non-firm employment was located in rural areas. Agriculture-related manufacturing and services (such as cottage industries and trading farm commodities and their processed products) were extensive (Manning 1998). REGIONAL DIMENSIONS AND DEVELOPMENT POLICIES A discussion based on nationally aggregated data provides a straightforward overview of the Indonesian economy, but it easily overshadows regional differences. Not all regions benefited from rapid economic growth, and regional income differences are still large. Industries vary from region to region; in some regions agriculture dominates. In a discussion of the Indonesian economy, regional perspectives are critical. With a population of nearly 200 million, Indonesia has the third largest population in Asia after China and India.' The archipelago's 13,677 islands cover an area of 1.9 million square kilo metre^.^ Regional differences in this vast nation are striking. Previous discussions have looked at Indonesian development from a national perspective, but it is essential to examine Indonesian rural development from a regional perspective. The conflict between pribumi (indigenous Indonesians) and ethnic Chinese has received widespread publicity; issues among pribumi themselves, however, are more complex. The indigenous population is divided into nearly 300 ethnic groups, each of which has its own culture and lang~age.~ Although 88 per cent of Indonesians practice Islam, this average figure does not guarantee religious homogeneity. Some ethnic groups in the Outer Islands practice other religions, such as Christianity in North Sumatra, East Nusa Tenggara, North Sulawesi, Maluku and Irian Jaya, and Hinduism in Bali. Thus, re ional diversity in Indonesia is not only geographic. g Cliford Geertz described the Indonesian economy as not simple, but compound. It is neither unitary nor systematic, but multiple and irregular. In desi nin policies, he stressed the importance of regional dimensions. g g ...[Olver-all development policies need to be much more delicately attuned to the particularities of local social and cultural organisation. [However,] [tlhe ideologies of modern nationalism...arising as they do out of an intense concern with massive social reconstruction, show a strong tendency toward a neglect, even an outright denial, of imporcantvariations in domestic culrural pauerns and of internal social discontinuities. (Geenz 1963b:1565). RURAL DEVELOPMENT IN INDONESIA 171 Despite Indonesia's regional diversity, the Indonesian government is highly centralised, and the autonomy of regional governments is strictly limited. Most of the revenue of the regional governments is transferred from the central government; they have few revenue sources of their own. Rural development policies were designed and implemented with strong leadership at the centre under the slogan of national unity, while regional factors were ne lected. The central government'sscepticism about regionalism may have g stemmed fiom its bitter experiencewith secessionist movements in the 1950s. Strong central government contributed to the integration of the national economy. The national slogan of 'Unity in Diversity' (Bhinneka Tungal Ika) became a reality through decades of sustained economic growth driven by a highly centralised government with abundant financial resources (Hill and Weidemann 1989). Huge oil revenues collected by the central government were transferred to the regions in the form of publicinvestments in physical and social infrastructure. This system helped integrate the nation, and the central government took the initiative to implement rural development policies. On the other hand, the highly centralised system neglected regional diversity. Since Indonesia became independent, Java has maintained its dominance, while resource-richregions in the Outer Islands were the major earners of foreign currency through traditionalexport commodities and oillnatural gas. Thus, the economic interests of Java and the Outer lslands conflict with each other, making regionalism a sensitive issue for the Indonesian government. Under the New Order regime, regional development was brought about through massive public investment in rural infrastructure and human resources development Uones and Hull 1997). Demographic dynamism in the form of rural-urban and rural-rural migration was counted as a key to economic growth (Hugo et al. 1987), hence the government-sponsored 'transmigration' programs by which the rural poor of densely populated Java were moved to labour-scarce Outer Islands.lo Despite the rapid economic growth of the past three decades, arid massive public investment in the regions, income disparity across the regions is still large. Per capita GDP in Eastern Indonesia is only about one-third that in Java (Barlow arid Hardjono 1996), arid the regional development process has been uneven. 172 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH RURAL PESSIMISM, OPTIMISM, AND DILEMMAS IN DEVELOPMENT Any economic policy is an amalgam of political compromise and some economic rationale; thus, to understand how and why they were adopted, economic policies must be examined in historical context. The Indonesian nationalism that flared in the colonial era'influenced policymakers, since the newly born government had somehow to obtain legitimacy. Indonesian policymakers were also influenced by other political and cultural factors, as well as by prevailing economic thought, though they may not have been conscious of these influences. In the 1950s and 1960s,the prevailingthought on economic development was rural pessimism. Rural pessimism and optimism In the 1950s and 1960s, pessimism about rural stagnation took hold throughout the developing world. The idea of rural pessimism originated in Java, and the Indonesian economy was regarded as a good example. The core of rural pessimism rests on an approach attributed to the series of classical studies on Java by J.H. Boeke, who studied Java from the colonial era to the early twentieth century. Boeke (1953) developed his pessimism theory on rural sector growth on the basis of observations made in the late colonial period, when massive irrigation investments failed to increase paddy area and food production (Booth 1988).According to Boeke, the traditional sector is a dsphere whose inhabitants engage in primary industries such as agriculture, forestry and fisheries. The rural sector, which Boeke referred to as 'villages', was regarded as prehistoric and typical of pre-capitalism (Boeke 1953). Boeke defined villages as peasant communities whose boundaries were restricted by the extent to which community members could regularly interact for collective actions under traditional agricultural technologies (Boeke 1953).Accordingly, the limit of villages was very narrow, and the scope for expansion seemed limited. Boeke's pessimistic view of agricultural production was shared by Clifford Geertz, who enhanced and extended Boeke's pessimism through his anthropological studies in East Java and Bali. Geertz observed that modern technologies and investments had little effect on Javanese agriculture (Geertz 1963a), however Geertz's pessimism was not limited to agricultural production, but extended to all economic activities in rural areas. Geertz pointed to a lack of entre reneurship within the peasant class, and argued p RURAL DEVELOPMENT JN JNDONESIA 173 that entrepreneurshipwas limited to members of the population with 'extra- villa e status', such as ethnic Chinese traders in East Java, because g '[el ntrepreneurial groups stand outside the immediate purview of village social structure' (Geertz 1963b3148). Geertz argued that Javanese agriculture was static in nature and trapped in a low-levelequilibriumwith its path to modernisation blocked. Observing that 'the prerequisite for success [for entrepreneurs] is rather organisational and administrative skill,' and that '[ilt was not the technically most skilled...who became managers, but the mostskilled handlersof men' (Geertz 1963b:151-2), Geertz argued that innovations were hampered less by technical problems than by organisational problems. Geertz's pessimism also applied to distribution of output among the rural population, which Geertzsaw as involving an 'involution process'. According to Booth (1988:6), '[ilt was in fact the distributive side of the involutionary process which accounted for the passivity and resistance to change and modernisation'. As noted by Booth (1988), while pessimistic about production, Geertz was optimistic about the distributional aspects of Javanese agriculture. Under the pressure of increasing numbers and limited resourcesJavanesevillagesociety did not bifurcate, as d ~ that of so many ocher 'underdeveloped' nations, into a group of d large landlordsanda group of oppressednearserfs. Rather it maintaineda comparatively high degree of social and economic homogeneity by dividing the economic pie into a steadilyincreasingnumberof minute pieces, a process to whichI have referredelsewhere as 'shared poverty' (Geercz 1963a:97). Geertz 's pessimism has been examined and criticised by many scholars, an indication of the influence of his ideas (Van Niel 1992). In the late 1960s and 1970s, however, Geertz's ideas were displaced by new views shared by many critics, who observed vivid but polarised growth processes in Indonesia's rural sector (Collier 1981; White 1976). The critics were more optimistic than Geertz about production but more pessimistic about income distribution. Van Niel (1992), for example, cast doubt on Geertz's model in arguing that Javanese villages adjusted by being increasingly involved in outside markets, while still retaining social distinctions. More recently, however,Geertz's pessimism has been revived. For example, his pessimism about rural entrepreneurship has been incorporated into the theory of 'traders' dilemma', which argues that the social norms of peasant communities that urge well-to-do villagers to redistribute their income to poor neighbours prevent village-born traders from accumulating capital 174 RURAL DEVELOPMENT AND AGNCULTURAL GROWTH and becoming modern entrepreneurs (Evers and Schrader 1994). The modern version of production pessimism is also expressed by recent critics who are reluctant to accept the positive role of agriculture.ll Basic ideas and dilemmas of Indonesian economic development Economic policies are born from political compromise, so it is necessary to know the major players in the Indonesian political arena. During the New Order regime, the Indonesian armed forces (ABRI) and Western-trained technocrats were major actors, but the most powerful man was Suharto, the country's president. Indonesian economic policies swung back and forth between relatively open liberalism and nationalism under the authoritative government. The largest political organisation in the country, ABRI possess a strong ideological and legal mechanism to support the government. The armed forces also stress their political legitimacy in their role as the supporter of 'integralism', as reflected in the state ideology of pancasih.12TG-doctrine of integralism regards the state and society as an organic whole in which the primary emphasis is placed on social obligations rather than individual rights (Ramage1995). Furthermore, the Indonesian armed forces are unique in their double function in both civilian and military affairs, a situation unheard of in Thailand or the Pldippines (Vatikiotis1993).At the beginning of the New Order rule, the leadership of the armed forces was beholden to a small group of 'financial generals' who headed large state-owned enterprises and had access to enormous funds. Use of these funds was obscure and corruption was the norm.13Since the New Order government depended heavilyon theloyalty of army leadership, politicalaction againstthese financial generals remained a delicate issue even for the President (Mackie 1970). As the pardian ofpancasih, ABRI stressed economic growth with justice and national unification, to be achieved only through state intervention (Chalmers and Hadiz 1997). ABRI's brief on pancasila and nationalist sentiment is expressed in Ali Moertopo's statement of 1972:14 This process of modernisationwill naturally involve conflicts,as new norms come into conflict with traditional norms. Consequently,modernisation requires planned social and cultural change.. .As'planned change', modernisation must clearly determine the direction which will be taken. This requires that the model of the pancmila state be clarifiedand the ~rinci~lesofpancmibapplied to everyaspectoflife, to everygovernment institution and state organisation, as well as to all levels of urban and rural society (Chahers and Hadiz 1997:77). RURAL DEVELOPMENT IN INDONESIA 175 As the state ideology, 'pancailrz democracy' was inflexible, because the government, backed by ABRI, so closely controlled its institutions. As the economic ideology, however, 'pancaih economy' was subject to a broader interpretation (Liddle 1982), still closely related to the interpretation of Paragraph 33 of the Indonesian constitution of 1945, which states that 1 economic affairsare to be organised as a joint effort based on family principles 2 branches of production which are important for the state and which control (supply of) the basic needs of the masses are to be controlled by the state 3 the land and water and natural wealth contained within are to be controlled by the state and used as much as possible for the prosperity of the populace.15 Sentence (1) of Paragraph 33 is interpreted as meaning that cooperatives should become the basis of economic enterprises. The New Order government thus promoted the development of cooperatives in rural areas, while restricting competition among private firms, since competition was believed to conflict with the principle of 'a joint effort based on family principles' (Rice 1983:665). Promotion of cooperatives did not mean they were encouraged to organise self-motivated cooperatives as common in Western nations; on the contrary, the government strictly controlled the design and establishment of cooperatives in villages. The most controversial clause is Sentence (2),which provoked debate as to how to interpret the phrase 'controlled by the state'. The Sukarno (Old Order) government interpreted it to mean that the economic sectors related to basic needs should be owned by state enterprises, while the New Order government interpreted it to mean that the government could control private enterprises without owning them (Rice 1983). For Sentence (3), the New Order government relied on the private sector to develop natural resources without permitting private ownership of those resources. Whatever the interpretation given to these sentences, they are all clearly characterised by state intervention. The state-interventionist ideas on which they are based are expressed in the following statement by Soedjono Hoemardani, a co-founding member of CSIS. The growth of modern sector occurs at the cost of and weakens the traditional sector.. .That sector of society organised into cooperativesis comprised of economically weak groups, and they need to be given special attention. On the other hand, the state 176 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH sector, namely the government,has the task of supervising, directing, protecting and, if necessary, subsidising this section of the society so that it becomes a viable economic force (Soedjono Hoemardani 1981, quoted in Chalmers and Hadii 1997934). Western-trained technocrats have often been contrasted with the voices of the Indonesian armed forces, yet according to Glassburner, 'they are nationalistic and are strongly influenced by their cultural herita e' g (1978:28). This is the reason foreign advisors from international donor agencies are often frustrated when their advice is ignored. Emil Salim, one of the key technocrats, expressed mixed sentiments about open-door policies and state intervention.lG What we are pursuing, however, is not only a society with 'an equitabledistribution of income', but alsoone that is prosperous.This meansthat nationalincomemust alsogrow more rapidly...it is dear that apancdsikzeconomy not only rejects free flight liberalism but also statism.. .the branches of production which are important to the state and which affect the livelihoodof the mass must be controlled by the state (30June 1966).17 The Indonesian economist Suhadi Mangkusuwondo (1973) expressed the dilemma his country faced as a 'hard choice between economic growth and social justice. After relatively open economic policies in the late 1960s had enabled the Indonesian economy to grow, various problems such as an increased income gap across regions and economic sectors arose. Apparently presuming rapid economic growth and social justice to be mutually exclusive, Mangkusuwondo argued that [elconomicdevelopment is only one part of the totalityof nationalprogress.Democracy, nation building,and social justice are some of the other ingredients of social progress. By over emphasisingthe economicsideof the developmenteffort, thereis great danger that Indonesia will be heading toward some serious political or social trouble in the near future.. .a lower growth rate is conducive to more government control of the economy, since sufficient investment in the social sector could only be achieved through extensive state intervention (Mangkusuwondo 1973:28, 33). In addition to ABRI and the Western-trained technocrats, two other groups of Indonesians cannot be ignored in a discussion of Indonesia's economic development-Muslims and ethnic Chinese. Although nearly 90 per cent of Indonesians profess Islam, their voices are not uniform. According to Ramage (1995:15),'many nominal Muslims do not channel their political ambitions through Islam. For example, in what is widely regarded as the most "fair" and accurate of Indonesian elections, the 1955 parliamentary polls gave all Islamic-oriented political parties a combined total of only 43.5 per cent of the vote'. RURAL DEVELOPMENT IN INDONESIA 177 This does not mean, of course, that Muslim voices were not influential in the political arena. Islamic groups played key roles in the struggle for independence and the rise to power of the New Order government, yet some Islamic leaders were unhappy that no special role was assigned to Islam in the constitution.Their demand for the inclusion in the constitution of Islamic law and Muslim status for the presidency was promised to them once, then rehed at the last stage of drafting because Mohammed Hatta feared that Christian-dominated Eastern Indonesia would not join the Republicif the state seemed to endorse Islam. The Islamicgroups' frustration led to a series of Islam-inspiredarmed revoltsagainst the central government, called Dam1hhm, between 1948 and 1962. These were suppressed by the armed forces (Ramage 1995),which regarded Islamic fundamentalism as a danger to national integrity. About 6 million Indonesian citizens of Chinese origin form an ethnic minority in Indonesia but are said to dominate the national economy. It is widely believed that 70 per cent of the Indonesian economy is owned by ethnic Chinese Indonesians. However, Kwik Kian Gie, a leading economic commentator and politician, has argued that the economic dominance of the Chinese is just a myth. According to Kwik, 90 per cent of the economy is operated by public sectors, while the other 10 per cent is shared by diverse private sectors, including foreigners, pribumi, and ethnic Chinese. The share owned by ethnic Chinese is very small, he maintains.18Since reliable statistics are not available, any figure is largely conjecture. Whatever the situation, however, the government has tended to regard pribumi as being at a disadvantage in relation to ethnic Chinese in the business arena, thus it has adopted various economic polices that favour the pribumi. Since most indigenous Indonesians (pribumz) practice Islam, while ethnic Chinese do not, ethnic and religious boundaries have !generally been preserved. P~ibumiIndonesians are frustrated by the idea of Chinese dominance, while ethnic Chinese feel their social position is insecure and unfairlydiscriminated against. Pribumi-Chinese relations are a very sensitive political issue. Politically, ethnic Chinese have tended to remain relatively invisible, withdrawing from politics out of a sense of self-preservation (Vatikiotis 1993). The positions of the key players in the Indonesian political arena have generally been nationalisticand interventionist,characteristics that originated from the discourse on state ideologies practised in the years leading up to independence. In addition to this political heritage, three other hctors may 178 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH have constrained or influenced the ideas of policymakers &hen formin g rural development policies-that is, rural pessimism as an academic thought, rural community sentiments based on the communal ties characteristic of villages, and anti-colonial sentiments, as explained below. First, in the 1950s and 1960s, rural pessimism was common in academic thought throughout the world and industrialism the norm in the developing world. It is hardly surprising, therefore, that Indonesian policymakers were sceptical about the potential of traditional sectors as a source of economic growth. This scepticism could have been the foundation for the neglect of traditional sectors in economic development policies. Second, nostalgia for a rural communal way of life was based on the intimate relationships characteristic of peasant villages, where population density is high and community members tend to be tightly bound by kinship and location. The main means to a livelihood in such societies in Indonesia is intensive irrigated rice culture, and hming operations are often collective actions. Similarly, social events such as weddings and funerals tend to involve collective efforts. In fact, mutual help is the norm in Indonesian villages. Born in a small Javanese village near Jogjakarta, Suharto always remained conscious of his rural roots, perhaps even choosing to emphasise his identity as a farmer from a smd village out of political expedience (Vatikiotis 1993). It is possible that Suharto thought his village origin would appeal not only to farmers, but also to urban dwellers with village origins. Third, anti-colonial sentiments were strong in Indonesia after independence, and institutions that existed during the colonial era, such as free markets and foreign capital, tended to be regarded as an evil. We can now outline the basicideas of economicdevelopmentin Indonesia, especially those related to rural development policies, in terms of three ideologies and three cultural factors. The three ideologies are strong state nationalism, which rationalises the broad intervention of the government in the economic sectors of basic needs, including agriculture integration, as proclaimed in pancmikz a strong belief in the positive role of cooperatives formed among economically weak groups, including peasants. The three cultural hctors are the prevalence of rural pessimism in academic thought family farming in the peasant community strong anti-colonial sentiments. RURAL DEVELOPMENT IN INDONESIA 179 These ideologiesand culturalfactors can be seen to form a basicframework for the rural development policies of the New Order government. ORIGINS OF RURAL DEVELOPMENT IN INDONESIA The New Order government either inherited or intentionally ignored the economic and political vestiges of the late colonial, Japanese occupation, Revolution, and Old Order periods. The economic achievements of the last stages of Dutch colonial rule in the early twentieth century and the post-recession recovery of the late 1930s also played their roles. In the previous section, we examined the success of the Indonesian economy since 1966, implicitly setting the 1950s and early 1960s, when the economy was stagnant and exhausted, as a basis for comparison. For the three decades following the early 1930s,the Indonesian economy s&red severely from successive external and internal difficulties. The world depression of the 1930s hit Indonesia's cash crop exports hard. Four years of Japanese occupation during the Second World War and the following struggle for independence did not present much opportunity for economic recovery. Indonesia's economic problems were serious enough to compare with those faced by the developing countries of Africa As a result, the Indonesian economy remained severely depressed until the mid 1960s (Hill 1996). It would not be fair to discuss only recent development experience and ignore the preceding growth records, thus we begin with the Dutch colonial era of the nineteenthcentury. Major events in rural developmentand policies that might have affected the rural population are summarised chronologically in Table 6.4. Late colonial era (before 1941) By about 1910, most of the regions now included in the Republic of Indonesia had been brought under Dutch rule, but a common sense of Indonesian identity had yet to form among the populace. Indonesian leadership groups and Indonesian identity were formed gradually, as anti- colonial movements took place throughout the 1920s and 1930s (Ricklefs 1993). After the economic stagnation of the last two decades of the nineteenth century, accelerating economic growth expanded agricultural production, exports, and imports. Smallholder sectors expanded rapidly, 'as a spontaneous response on part of the indigenous population to the new opportunities which international trade offered' (Booth 1998:35).The rapid Table 6.4 Chronologyof ruraldevelopmentinIndonesia w Paddy 00 0 (million metric Major economic and political events tonnes) Mallaca the centre of a vast trading network. Spices (clove and pepper), coffee, and sugar from Outer Century Islands and rice from Java are traded extensively: Cloves produced only in Maluku in eastern Indonesia. Traders from Europe and Asia compete for cloves centuries until the Dutch monopolise production and trade in the mid 1650sb Cultivating system introduced (1830-70). Production of sugar, coffee, indigo and other crops is pushed to compulsory planting. Tobacco and tea introduced in the 1830s and cinchona in the 1870s. Major export crops-sugar and coffee. Opening of Suez Canal stimulates cash crop trade. Major export crops-sugar, coffee, tobacco and tea. Cultivating system abolished. Liberal period (1870-1900), when private capital from European nations is intensified. Rapid growth of agricultural exports by Dutch plantations.'Major export crops-sugar, coffee, tobacco, tea and cinchona (c.1900). Introduction of ethical policy by Dutch liberal government, which pays attention to welfare and education of the colony. Investment in irrigation and education. Export crops reach their peak (c. 1915). Modern natural rubber variety (HeveaBrasiliensis) introduced in Sumatra. Major export crops-sugar, coffee, tobacco, tea and cinchona (c. 1910). Palm oil introduced to the east coast of Sumatra as a large plantation crop. Major export crops-sugar, coffee, tobacco, tea, cinchona and rubber (c. 1920). Sugar production in Java reaches its peak.dMajor export crops-sugar, coffee, tobacco, tea, cinchona, rubber and palm oil. Japaneseoccupation (194245). Proclamation of independence (1945). Sukarno announces Pancasila, the state philosophy based on five interrelated principles. War for Independence (194549). Linggajati Agreement-Netherlands-Indonesia Union under the Dutch crown with a loose federal arrangement (194647). Netherlands recognises the RUSI (Republic of the United States of Indonesia) as an independent state. All federal states absorbed into the Republic of Indonesia. Parliamentary democracy (1950-57). Secessionist movements in Outer Islands (1950-56). First general election. Sukarno declares Guided Democracy (1957-65). Seizure of Dutch properties. Three-year program of increased rice production using imported fertiliser.' Irian Jaya brought under Indonesian administration. First trial of BIMAS program launched Failed coup d'ktat attempt. Sukarno removed from power. Bloody purge of PKI (Indonesian Communisc Party) and Chinese. Suharto and Indonesian military take political control. Total foreign debt comes to US$2.3 billion. Suharto nominated as acting president of Indonesia. IGGI, an international group of lenders, is formed to coordinate multilateral aid to Indonesia (1967-92). Indonesia adopts open-door kissez-fdire policies favoured by the IMF and the World Bank and investors from western nations. lrian Jaya becomes 26th province. Repelita I (First Five-Year Plan, 1969-73) gives high priority to irrigation rehabilitation and e~tension.~Introductionof new credit policy for agricultural investment. Diffusion of rice HYVs contributes to rapid growth of paddy yield (Green Revolution). Devaluation of rupiah. Rise of nationalism. General election (Golkar's victory). Rice shortage. Windfall income and increase of public development expenditures (irrigation and education). First oil boom. MPR (People's Consultative Assembly) re-elects Suharto to second five-year term. Village cooperatives (BUUDIKUD) assigned to be sole distributors of fertiliser and pesticides for food crop sector.= Malari riots (anti-Japanese) twn government to authoritarianism." Repelita I1 (197478) focuses on agriculture, employment and regionally equitable development. More emphasis placed on non-rice food crops such as maize and sorghum. Democratic Republic of East Timor declared. Indonesian forces occupy East Timor. Pertamina crisis. East Timor made 27th province. Marketing of fertiliser deregulated and subsidised.' Devaluation of rupiah. Repelita I11 (1979-83) emphasises development of agriculture-related and other industry. INSUS (intensified extension system) implemented on food crops, especially rice. Economic crisis. Devaluation of rupiah (28 per cent). Repelita IV (1984-88) concentrates on basic industries. Rice self-sufficiency atcained. Devaluation of rupiah (31 per cent). Major deregulations of trade policy Fertiliser subsidy reduced (July). Repelita V (1989-93) emphasises transport and communications. Subsidy on pesticides is eliminated. Importance of regional development in eastern Indonesia stressed (Presidential speech).k Dili incident. Rice production falls because of drought. General election. CGI formed to replace IGGI and excludes the Netherlands. Repelita VI (1994-98) calls for strengthening of regional autonomy (Presidential speech)! Financial crisis. Suharto steps down from the presidency and is replaced by Habibie. General election. $ aRicklefs, M.C., 1993. A History ofModern Indonesia since c. 1300, 2nd ed., MacMillan, London:19-21. u Bulbeck, D., Reid, A,, Tan, Lay Cheng and Wu, Yiqi, 1998. Southeast Asian Exports Since the 14th Century:cloves,pepper, coffeeandsugar,Instituteof ~3 Southeast Asian Studies, Singapore:l7, 20. Bevan, D., Collier, I? and Gunning 1..1999. The PoliticalEconomy ofPouerg,Equityand Growth:Nigeka and Indonesia, Oxford Univeni~yPress, New R C York:197. s Bulbeck, D., Reid, A,, Tan, Lay Cheng and Wu,Yiqi, 1998. Southeast Asian Exportr Since the 14th Century:cloves, pepper, coffeeand sugar,Institute of Southeast Asian Studies, Singapore:l10. E Booth,A., 1988. AgriculturalDevelopmentin Indonesia,Allen & Unwin, Sydney:114,147. 0 Booth,A., 1988. Agriculh~ral Developmentin Indonesia,Allen & Unwin, Sydney:114,147. gSlayton,T.M. and Exawirya, I.G.N., 1978. 'The fertiliser situation', Bulletin of Indonesian Economic Studies, 14(2):70. Bevan, D., Collier, l? and Gunning, J., 1999. The Political Economy flowerg, Equity and Growth:Nigma and Indonesia, Oxford University Press, New York: 249. Booth,A,,1988. AgriculturalDevelopment in Indonesia,Allen & Unwin, Sydney:158. jPangestu, M. andhis, I.J., 1994. 'Survey of recent developments', Bulletin oflndonesian Economic Studies, 30(2):30;Booth, A., 1988. Agricultural Developmmtin Indonesia,Allen & Unwin, Sydney:149. Booth,A., 1988. AgriculturalDevelopmentin Indonesia,Allen & Unwin,Sydney: 149. 'Pangestu, M. and his, I.J., 1994. 'Survey of recent developments', Bulletin oflndonesianEconomic Studies, 30(2):28. Source: US Library of Congress, 1992. Indonesia:a counrrystudy, US Library of Congress, Washington, DC; Various authors, various years. 'Survey of recent developments' , Bulletin of Indonesian Economic Studies,various issues. RURAL DEVELOPMENT IN INDONESIA 183 gowth of the first two decades of the twentieth century, however, was interrupted by world depression in the 1930s,with demand for Indonesia's major export crops suffering severely. In 1939, despite recovery after the depression, per capita GDP was still lower than in 1913 (Booth 1998). During the early twentieth century, production of export crops increased steadily. In the 19OOs, though still dominated by traditional crops such as sugar, coffee, tobacco, tea and cinchona, Indonesian export markets had begun to shift to new crops such as natural rubber and oil palm. Natural rubber, Hmea brasilimh, was first ~lantedin Java and Sumatra in the early twentieth century. Oil palm was introduced as an estate crop. The indigenous population did not have much interest in palm oil because coconut trees were already used extensively for food and as a source of edible oil. In 1911, the fust large-scale oil palm plantation was established on the east coast of Sumatra, which later became the major centre of oil palm production in Indonesia (Mansvelt and Creutzberg 1975). In 1930, sugar production in Java reached its peak (Bulbeck et al. 1998), and natural rubber and palm oil quickly became important export crops, surpassing traditional crops such as tobacco, tea and cinchona. In the 1910s and 1920s, the production of natural rubber increased rapidly, becoming the dominant crop after the First World War. Oil palm production also increased rapidly in the 1920s and 1930s. The world depression of the 1930s hit exports of some of these cash crops. Production of sugar cane, one of the most important export crops at the time, declined to one-fifih during the period of 1930-35.19 From independence to the end of the Old Order (1942-65) Dutch colonial rule ended with Japanese occupation in 1942. The three and a half years of Japanese occupation dismantled the colonial system and brou ht about the extraordinary changes that enabled Indonesian g independence. It took another five years of fierce struggle from 1945, when Independence was proclaimed, to final recognition as an independent state in 1949. In 1945, Sukarno announced pancasila as the philosophy on which the nation would be based. In the following year, the Linggajati Agreement was established, forming the Netherlands-Indonesia Union under the Dutch crown with a loose federal arrangement, however this was soon broken by the Dutch reoccupation of Java in 1947, called a police action'. In the face of increasing of international criticism and direct intervention by the United Nations, the Netherlands finally recognised 184 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH RUSI (Republic of the United States of Indonesia) as an independent state in 1949 (Ricklefs 1993). In the following year, all federal states under RUSI were absorbed into the Republic of Indonesia; thus, a new Republic of Indonesia with a unitary constitution was established, though with many religious, ethnic, cultural and economic issues yet to be resolved. In 1950, post-war rehabilitation efforts were made under a fragile parliamentary democracy that lasted until 1957. There was a political 'agreement on the desirability of a non-Western type of democracy, but what this meant remained unclear' (Bevan et al. 1999:209). Although the Japanese occupation severely eroded the colonial system and the 194549 revolution dismantled colonial rule, the new Indonesian government and economy inherited much from the former colonial ruler. Policies of the period were characterised by heavy government regulation, a predilection for intervention,and the dominanceof Dutch capitalin the economy (Bevan et al. 1999). Anti-colonial sentiment and nationalism were strong. The Indonesian currency, the rupiah, was set at a high level, and the economic interests of export areas in the Outer Islands suffered. The basic structures of the Indonesian government and economy were set as highly centralised, thus the resource-rich Outer Islands were frustrated not only by the political dominance ofJava, but also byeconomicsuppression. Secessionist sentiments flared up throughout the 1950s. Rebellions in the Outer Islands broke out in South Sulawesi in 1951,Aceh in 1953, Sumatra in 1956 and Eastern Indonesia in 1957; these movements were suppressed by force (Ranis and Stewart 1994). Political conditions were unstable, but the government made some rural development efforts, and agricultural production recovered slightly. A three- year program to increase rice production using imported fertiliser was launched in 1959 (Booth 1988). In 1963 the first agricultural extension service for rice (Bimas, or bimbingan massal) was launched. Rice production (paddy in dry stalk) in the late 1940s came to about 7 million tons in Java and Madura and 4.5 million tons in the Outer Islands. By the early 1960s, these figures had increased to 9.6 million tons in Java and Madura and 6.8 million tons in the Outer Islands, higher than yields achieved during the colonial period in the early1930s (van der Eng1996). In termsof agricultural labour productivity,however, the level of the colonial ~eriodwas not attained until the early 1970s (van der Eng 1996). In 1957, Sukarno announced the implementation of his 'Guided Economy' policy, characterised by pro-socialist sentiment and a strong anti- RURAL DEVELOPMENT IN INDONESIA 185 Dutch and anti-US orientation. It was a fluid and constantly changing system, dominated by Sukarno's ~ersonalit~(Ricklefs1993). Also in 1957, Dutch-owned enterprises were seized. An aggressive foreign policy brought Irian Jaya under Indonesian administration in 1963." The economic results of the Guided Democracywere miserable.Accelerated inflation, deteriorated infrastructure, stagnant output, and widespread poverty and hunger were the norm (Booth 1998). Sukarno was heavily committed to maintaining equilibrium between the army and leftist political patties, thus only a few programs were actually implemented. Land reform was one of the few, beginning in September 1960 (Utrecht 1969). The idea of land reform was announced by Sukarno in 1959, in a political manifesto address in which he terminated property rights to land. Although his idea seemed to be motivated by a socialistic ideology, he also assumed that redistribution of land to farmers would enable them to cultivate more intensively and thus contribute to increased production. Considering the very small size of typical farms in Indonesia, especially in Java, the potential economic benefit of the reform was doubtll. After a fierce political struggle and compromise, the SupremeAdvisoryCouncil (DPA)and then parliament set the framework for the land reform program under the Basic Agrarian Law (Undang-Undang Pokok Apria) of 1960. This law was based on Article 33 of the 1945 Constitution and the state ideologypancaih, which declared that land has a social function and must therefore be controlled by the state. It provided for the implementation of a comprehensive land system by which two old property rights would be replaced-traditional a h t law without formal registration and a Westernised system introduced by the Dutch colonial government (MacAndrews 1983). The new law forbade ownership of land in excess of a certain area and absentee ownership, and authorised the government to seize surplus land for redistribution.The program began in 1961, but progress w a very slow, partly because officials were unskilled but mainly because a land registry could not be established, so the program had to be based on colonial land tax registers (van der Eng 1996). By 1964, only 400,000 hectares had been redistributed to farmers, and Sukarno expressed concern over delays. Sukarno's Guided Democracy came to an end with the coup of 1965, which brought enormous political and social turmoil and a shift of power from Sukarno to Suharto, as well as the end of the land reform program. Although the program still existed in the late 196Os, the new government did notshare haveSukarno's interest in it.Altogetherabout 1million hectares, 186 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH or about 5 per cent of total farmland, were redistributed under the land reform program. Since nearly 70 per cent of the redistributed land was transferred from former public lands, the land redistributed from landlords to farmers was at most 130,000 hectares. Furthermore, in the course of the suppression of PKI members and sympathisers after the coup, some redistributed land was illegally returned to former owners as peasants who had obtained land with support from PKT were either murdered or escaped to cities (Utrecht 1969). Thus, in the end the land reform program had very little impact on the land tenure system in Indonesia. RURAL DEVELOPMENT AND GOVERNMENT POLICIES Success of rice under the New Order (1966 - mid 1980s) It is widely recognised that Indonesian agriculture has achieved remarkable growth since the early 1970s (Booth 1988; van der Eng 1996; World Bank 1993),with food crops, especially rice, experiencing strong growth (Mears 1981). After the failed coup d'ht in 1965, Sukarno was removed from power. A bloody purge of PKI members and an attack on ethnic Chinese were extensive, especially on Java. In 1966, Suharto and the Indonesian military took political control, and in 1967 Suhartowas nominated acting President. The most urgent issue facing the new regime was the enormous foreign debt of US$2.36 billion, inherited from the era of Guided Democracy. In 1967, the Inter-Governmental Group on Indonesia (IGGI), an international group of lenders, was established to coordinate multilateral aid to Indonesia. Initially, the New Order government adopted open-door policies, partly to placate debtors and attract investors from Western nations, including the IMF and World Bank, and ~artlybecause policymakers recopised the limits of state interventionand the role of the private sector.This is reflected in the 1996 speech by Sultan Hamengkubowono IX,a key figure in the early stages of the New Order regime.21 The Government deliberatelyrestricts its program within the limit of its capacityand the prevailing atmosphere in the conviction that it will not reach its goal without public support and aid...The governmentwill abandon its attitude of regardingmedium- and large-scale enterprisesas...enemies (12April 1966, reproducedin Chalmers and Hadiz 1997:47-8). Although US-trained technocrats gradually took control of economic management, there was strong political pressure for government intervention RURAL DEVELOPMENT IN INDONESIA 187 under the name of social justice and economic equity. As discussed earlier, these technocrats basically shared nationalistic sentiments. The open-door policy of BAPPENAS (National Development Planning Agency) was somehow modified during the drawing up of the First Five-Year Plan in 1968 (Chalmers and Hadiz 1997:42). Throughout the late 1960s, relatively open ~olicieswere maintained. The Indonesian government adopted the principle of a balanced budget, which was somewhat tricky because foreign aid and borrowing were used to attain the nominal balance, but the policy did reduce inflation and help maintain macroeconomic stability (Hill 1996; Woo et al. 1994). In 1963, an extension trial was carried out, and later broadened to become one of the key agricultural policies of the New Order government. The program was intended to intensify food crop production, especially rice. The First Five- Year Plan (1969-73) placed a high priority on rural development. Rehabilitation of irrigation facilities and enhanced extension services were emphasised (Booth 1988), and a new credit policy on agricultural investments was introduced in 1969. The period that followed the implementation of the First Five-Year Plan was characterised by the rise of economic nationalism and an economic policy shift toward protectionism. Food crops, especially rice, achieved monumental success, and, despite persistent pessimism, Indonesia attained rice self-sufficiency in the mid 1980s. In the political arena, an authoritarian regime gradually formed. In 1969, parliament passed an election law establishing memberships of a new parliament and People's Consultative Assembly (MPR) and their selection process. Under this law, the government could appoint 22 per cent of parliament seats and 33 per cent of MPR seats. No government employees were allowed to join any political parties other than Golkar (JointSecretariat of Functional Groups), the de facto ruling party and the body that coordinated army-civilian cooperation. The general election of 1971 resulted in a tremendous victoryfor Golkar, which gained 62.8 per cent sf contested seats. This enabled the government to control 73 per cent of parliament seats and 82 per cent of MPR seats (Ricklefs 1993). In 1973, the newly formed MPR re-elected Suharto to a second five-year presidential term. When the Japanese Prime Minister visited Jakarta in 1974, massive anti- Japanese protests broke out. Eventually known as the Malari riots, these protests reflected nationalistic sentiment, which regarded foreign capitalists, in this case the Japanese, as exploiters of the Indonesian economy. The 188 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH incident marked ashift in the government toward authoritarianism,followin g which the removal of critics of the President became routine (Bevan et al. 1999; Ricklefs 1993). Aggressive action was taken in external affairs. When East Timor, the former Portuguese colony, declared itself the Democratic Republicof East T i o r in 1975, Indonesian forces invaded. In the following year, East Timor was made the twenty-seventh province of Indonesia. The shift towards authoritarianismand the rise of nationalism were made possible by a great improvement in the government's financial position. The first oil boom in 1973, coupled with high prices in the preceding years, brought a windfall to the Indonesian government. In 1967, oil revenue accounted for only 12 per cent of total revenue; by the mid 1970s this figure had risen to about 60 per cent. This new revenue enabled the government to increase public development and promote the social welfare policies of the Second Five-Year Plan, launched in 1974.22Further funds for the first year of the plan were allocated to various social policies and subsidies for imported rice and fertiliser (Arndt 1974). As Hill (199643) observed, '[tlhe central government has had hitherto unimaginable resources...[alnd until the early 1980s, most of the government's revenue was raised painlessly, financed abroad by taxpayers and oil consumers'. During the 1960s and 1970s, rice production increased by nearly 250 per cent (Pearson et al. 1991), and the goal of self-sufficiency was finally achieved in 1985. Furthermore, as a result of the activities of BULOG, the State Logistics Agency, rice prices were more stable than the world market (Timmer 1996, 1997).23Behind this success lay government ~oliciessuch as investment in irrigation and adoption of high-yieldingvarieties (HYVs), though it might also have been helped by costly fertiliser subsidies. From the late 1960s on, diffusion of HYV rice contributed to the rapid growth of paddy yields. Developed by the Philippines-based International Rice Research Institute (IRRI) since in the mid 1960s, these new rice varieties were fertiliser-responsive semi-dwarf varieties, which could yield more than twice the amount yielded by traditional rice varieties. Furthermore, new varieties ripened in a shorter period and were insensitive to ~hotoperiod,making possible double and triple cropping. Higher yields could only be attained when sufficient fertiliser was applied under well- managed water control, thus irrigation and chemical fertiliser became crucial for desired productivity increases. New varieties disseminated by the IRRT in the mid 1960s have * I-.- *,-- *- + - -- -- *I-I f". I I - - (productionminus dorrestic consunption) 0 0 1961-65 19W70 1971-75 197HO 198145 198C90 199145 199C97 Source: Food and Agriculture Organization of the United Nations, n.d. FAOSTAT, FA0 online database, http://apps.fao.org/defadt.htm. Table8.3 Per capita monthlyfoodexpendituresof rural and urban dwellers,1962 and 1992 (Urban=100) Rice and cereals Meats and poultry Meats Poultry Fish and sea food Vegetables Fruits Oils and fats Source: National Statistical Office, 1966. Hoi~seholdExpenditure Si~tvey1963, National Statistical Ofice, Bangkok; National Statistical Ofice, 1994. Report of the 1992 Household Socio-Economic Survey, National Statistical Ofice, Bangkok. Agriculture's contribution to farmer welfare The rapid developmentof the non-farm sector made it possiblefor farmers to increase their income. In 1995, the average household income of farmers was nearly four times as large as the average household income of farmers in 1970 in real terms. Meanwhile, however, the income gap between rural and urban residents continued to widen right up until the mid 1990s, even though poverty reduction had been at the top of the political agenda since the 1970s. Environmental conservation As agricultural production in Thailand expanded, the country's forestland shrunk, and upland crops for export crept into idle paddy fields. In 1961, nearly 60 per cent of Thailand was covered in forests; by 1991 only 27 per cent of Thailand was covered in forests. When frozen shrimp exports boomed in the 1980s,Thailand's mangrove swamps were rapidly developed into shrimp farms, leading to the disappearance of 37 per cent of all mangrove swamps between 1979 and 1989 (Royal Forestry Department 1993). Thailand has achieved success in agricultural exports by expanding arable land rather than increasing land productivity, thus at the same time as it has succeeded economically it has failed environmentally. TRADE OPPORTUNITIES IN WORLD MARKETS Thai agriculturaldevelopment has depended heavily on overseas demand. In the 1950s' Thailand was still behind other countries in rice and rubber export, and only much later did it catch up to other exporters. Other major exporting countries after the 1960s had no significant export record in the 1950s, suggesting that the increase in Thai agricultural exports was due not solely to market expansion, but also to its higher agriculturalproductivity relative to other exporters (Honma and Hagino, this volume). Thai agriculture developed as well as it did because it relied on export markets and Thai farmers responded smoothly to overseas demand. The role of government was limited. Most studies indicate that Thailand performed better than competing exporters for three main reasons Thailand had a land surplus relative to its population, making it possible for it to respond productively to overseas demand (Siamwalla 1999; Dixon 1999; Siarnwalla et al. 1992; World Bank 1987) 314 RURAL DEVELOPMENT AND AGRICULTURALGROWTH a good feedback system, consisting of middlemen at several levels, delivered the necessary information on overseas demand to Thai farmers (World Bank 1987) Thai farmers were rational enough to respond to market prices (Siamwalla 1999). During the period 1961-85, increases in farmland per farm labourer and positive developments in education-thou h not irrigation, research g or extension-were significant factors contributin to the enhancement g of Thai agricultural labour productivity. Policymakers' expectations notwithstanding, irrigation did nothing to enhance rice productivity, though it clearly had positive effects on other crops. Research and extension focused on import substitutes, rather than export commodities. (Siamwalla et al. 1993; Siamwalla et al. 1990; TDRI 1995) Because chemical fertiliser did not play a crucial role in agricultural development, a government policy promoting it was not important (World Bank 1987). Thailand had a natural abundance of land, which the rural population reclaimed individually. The government did work to improve education in rural areas, but its efforts in this respect were not aimed at hastening agricultural development. Other government actions, such as the overvaluation of the baht, damaged the trade competitiveness of exports (Siamwalla and Setboonsarng 1991); rice and rubber, the country's major exports, were subject to government levies up until the 1980s. As a result, scholarly views of government agricultural development policies tend to be negative, and most scholars agree that Thai agriculture developed not because of government action, but rather through its inaction-that is, because government did not actively or efficiently intervene in production and trade (World Bank 1987). Siamwalla (1999) in fact states that the government should not intervene in the decision making of producers. Despite these generally negative perceptions, the role of government cannot be completely ignored, since, as a more detailed examination of each commodity shows, certain government practices did promote exports. For example, while irrigation may not have affected labour productivity, it did contribute to increased production for export. A rapid increase in maize exports was made possible by a contract between Japanese importers and the Thai government. The success of Thai sugar exports can be attributed to the government policy of protecting farmers and stabilising consumer prices. Ago-industrial commodities such as frozen chicken and canned pineapple received government assistance in the form of investment promotions. It is worthwhile identifying the policies that were in fact effective in promoting exports. Since market conditions differ from commodity to commodity, we examine each commodity separately, looking at rice and rubber as traditional commodities, maize, tapioca, kenaf and sugar as commodities from the first wave of diversification, and frozen chicken and canned pineapple as agro-industrial commodities. From the 1960s to the present, these eight commodities together accounted for 50-80 per cent of Thai agricultural exports. Rice Thailand exported rice even before the Second World War. For a time, because rice production in Burma-previously the world's largest rice exporter-was seriously damaged by the war, Thailand became the world's lar est rice exporter. After the Second World War, Burmese rice exports g again exceeded those of Thailand, continuing to do so up until the early 1960s (Figure8.5). In the late 196Os,however, price policies unfavourable ,to farmers caused Burma to abandon the world market. From the time of Burma's withdrawal to the late 1970s,the world's largest rice exporters were the United States and China. Even though it could not satisfy its own domestic demand for rice, China exported rice in order to obtain forei n currency. When the world rice market expanded in the late 1970s, g China could not increase its exports, thus from the early 1980s onwards the United States and Thailand were the only major rice exporters. In 1980, the United States and Thailand each exported about 3 million tonnes of rice; by 1985, however, Thailand had become the leader. The United States introduced export subsidies in an attempt to regain the lead, but this attempt failed. In the late 1980s, Vietnam entered the export market and became a new competitor, but Thailand has maintained its dominant market position to the present. In the late 1970s,several Middle Eastern and African countries became rice importers, and the world rice market expanded. Though the Asian market had already matured, Thailand managed to increase its share in the region as well as enter new markets in other regions (Fi ure 8.6). g The United States, on the other hand, could neither expand its exports in Africa nor maintain its market share in Asia. 316 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Figure8.5 Riceexportsfrom majorexportingcountries (yearly average in each period) Source: Until 1989, International Rice Research Institute, various issues. World Rice Statistics, International Rice Research Institute, Los Banos; afier 1990, Food and Agriculture Organization of the United Nations, n.d. FAOSTAT, FA0 online database, http://apps.fao.org/default.htm. A major role in Thailand's penetration of new markets was played by the private sector, with private enterprises exporting most of the country's rice and even shipping the market segment established by inter- governmental agreements. As the structure of the world rice market changed, new players emerged among Thai exporters. In the days when most Thai rice was exported to Asia, the market was dominated by Sino- Thai traders with connections to importers in Hong Kong and Singapore. From the mid 1970s onwards, however, new exporters broke this monopoly. Six of the ten largest rice-exporting companies in 1980 had entered the business in the 1960s and 1970s. Thailand was able to supply new markets because its abundant supplies of rice. Between the early 1970s and the late 1980s, Thai rice exports increased in volume by about 4 million tonnes. Production had continuously increased from the 1960s, but most of this was rice for Figure 8.6 Expansion of Thai rice markets Source: Institute of Developing Economies, n.d. World Trade Database (AIDXT), Institute of Developing Economies,Japan External Trade Organization, Chiba. domestic consumption. When per capita rice consumption fell in the mid 1970s, Thailand had more rice available for export. In the 1980s, partly because of the United States' introduction of rice export subsidies, world rice prices fell. In 1986, however, Thailand responded to this situation by reducing its rice export tax to zero, thus reducing the negative impact of the fall in world rice prices on domestic farm prices. Thus, Thailand captured the world rice market in the 1980s with the aid of US policies of the time. During the commodity boom of the 1970s, the United States relaxed its restrictions on rice planting causing a stockpile to appear by the late 1970s (Dethloff 1988). Under a new farm bill, rice farmers who restricted their planting area were allowed to sell their rice at a government-supported price, then repurchase it at a lower price. Because only farmers who restricted their planting area were allowed to participate in this program, however, US rice production actually decreased at a time when the export market was expanding. 318 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Thus, the United States not only failed to capture new markets, but also had to retreat from traditional markets. In summary, four factors made it possible for Thailand to become the world's leading rice exporter: active marketing by Thai rice traders in emerging markets; increased export capacity (due to a combination of increased production and decreased domestic consumption); bad US rice policy; and the elimination of rice export tax. Rubber Before 1960, rubber was Thailand's second-most important export commodity, and Thailand was the world's third-largest rubber exporter, after Malaysia and Indonesia (Figure 8.7). After the Second World War, the world rubber market expanded continuously, and Thailand showed Figure8.7 Rubber exports from major producing countries Malaysia \ Source: 1946-59 data: International Rubber Study Group, 1974. World Rubber Statistics Handbook 194659, vol. 1, International Rubber Study Group, London:16; 1960-83 data: International Rubber Study Group, 1986. World Rubber Statistics Handbook 1960- 85,vol. 3, International Rubber Study Group, London:l7; 198485 data: International Rubber Study Group, 1990. Rubber StatisticalBulletin, 45(3):8; 1986-91 data: International Rubber Study Group, 1992. Rubber StatisticalBulletin,47(3):11; 1992-97 data: International Rubber Study Group, 1998. Rubber StatisticalBulletin, 52(12):11. remarkable growth, especially during 1977-81 and 1984-97. Production growth after 1984 made Thailand the largest exporter of natural rubber in the world. Thailand's success in rubber production and trade was made possible by a complex interplay of production methods, traditional import partners, tire manufacturing methods and economic conditions in both rubber producing and rubber consuming nations. Most natural rubber is used to manufacture automobile tires. Thus, from the 1960s onwards, the three largest rubber consumers were the United States, Europe, and Japan, also the world's largest users of automobiles. Among rubber producers, Malaysia and Indonesia concentrated on producing Technically Specified Rubber or block-type rubber, which was sold mainly to the United States and Europe, while Thailand produced mostly RSS (ribbed smoked sheet) rubber for the Japanese market (Yamazaki 1988). Compared to their European and US counterparts, Japanese tire makers were latecomers. In the 1960s,Japanese tire makers bought cheap, lower-quality rubber from Thailand and used a labour-intensive manufacturing process to remove foreign matter from the rubber sheet. In the 1970s, rubber demand in Europe and the United States stagnated. The Japanese tire industry, however, recovered quickly after the oil crisis and increased its imports, mostly from Thailand. From the mid 1980s onwards, this situation changed, with Thailand continuing to increase its rubber exports while at the same time diversifying its markets. Diversification was possible because the quality of Thai rubber had improved enough-through the efforts of a Japanese tire maker, Japanese traders and Thai smoke houses-to attract Western tire makers (Shigetomi 1995a). In about 1980, Japanese tire makers automated their previously labour- intensive manufacturing process and could no longer use low-quality rubber. Bridgestone (the largest Japanese tire maker) assigned an engineer to Hatyai, the market centre for Thai rubber, and through the engineer's work production methods were standardised, rubber quality improved, and Thailand became the leading rubber exporter of the 1980s. All of this was achieved without government intervention. At the same time, however, Thai rubber yields were increasing thanks to a government program of replanting that had begun in the early 320 RURAL DEVELOPMENTAND AGRICULTURAL GROWTH 1960s and accelerated in the late 1970s. From 1988 onwards, Malaysian rubber exports decreased by about 1 million tonnes, while world demand increased by 500,000 tonnes. The Thai rubber industry was able to exploit this market opportunity, partly because planted area expanded but also because yields increased as a result of the government program. Kenaf During the 1960s and 1970s, Thailand was a significant exporter of kenaf (Table 8.4). Cheaper but of lower quality than jute, kenaf was used as a substitute for jute in order to save money on gunnysacks; thus, demand for kenaf depended on the price gap between jute and kenaf. At the time, Bangladesh held a near monopoly on jute export, with direct implications for Thailand's kenaf market. When natural disasters or political turmoil led to shortages of jute from Bangladesh, demand for kenaf surged (Samakhom 1975; Charoenloet 1971). When Bangladesh devalued its currency against the dollar in 1974175,however, kenaf exports from Thailand were af5ected negatively (Sarnakhom 1975). World exports of jute-like fibres have long been dominated by Bangladesh and Thailand. When demand for kenaf arose, Thailand was the only country in a position to meet it, and it could supply the market because it already had to the capacity to make gunnysacks. Before the Second World War,Thailand was a major rice exporter but had to import Table8.4 Exportsof juteand jute-likefibresfromBangladeshand Thailand in comparison to world totals, 1961-97 Jute (tonnes) Jute-like fibres (tonnes) World Bangladesh Thailand World Bangladesh Thailand 581,173 535,741 1,037 814,686 707,576 227 734,250 633,160 3,187 429,995 312,943 1,379 481,804 339,479 3,297 380,759 265,287 624 440,387 325,557 6,423 288,083 248,944 3,831 288,464 263,000 2,062 Source: Food and Agriculture Organization of the United Nations, n.d. FAOSTAT, FA0 online database, http://apps.fao.org/default.htrn. gunnysacks from Pakistan in order to transport the rice (Department of Factory 1959:43). After the war, in order to avoid having to import gunnysacks, the Thai government supported the production of kenaf, a fibrecrop traditionally used withinThailand as asubstitutefor jute imports. The first Thai pnnysack factory began operations in 1952, and two more factories opened the following year. Since both productivity and quality were low, the government protected the industry by forcing jute importers to purchase a certain amount of domestic product (Samakhom 1975). This protection stimulated cultivation of kenaf and thus prepared the industry for sudden demands from abroad. When demand from abroad rose, the price of kenaf rose too, and the early 1960s saw kenaf cultivation expand drastically. In the late 1960s, however, plastic bags began to replace gunnysacks and bulk transportation of agricultural commodities became more common, thus demand for jute and jute-like fibres began to fall (Chintayarangsan 1983). From the mid 1970s onwards, kenaf exports decreased rapidly, and Thai farmers turned to cassava, whose price was more attractive than kenaf's. Maize Thailand's maize exports have never exceeded 7 per cent of world trade volume in a market dominated by the United States. Despite its poor standing in the world market, Thai maize used to be the country's third- largest export commodity, in certain years accounting for nearly 13 per cent of total export value. Its importance as an export commodity stemmed only from its capacity to capture nearby markets; it had no comparative advantage outside Asia. Commercial production of Thai maize began in 1953, when the United States introduced the Guatemala variety as part of a technical assistance program (Muscat 1990). In 1958, when Japan increased its maize imports from Thailand by more than 300 per cent, planted areas began to expand rapidly (Figure 8.8) and 80 per cent of Thai maize went to Japan. From the late 1950s onwards, the number of swine and chickens in Japan increased rapidly, and Japanese farmers relied increasingly on imported feed grain (Hayakawa 1986). Most of the imported feed grain came from the United States, but Thai maize interested Japanese importers for several reasons- 322 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Figure8.8 Destinations of Thai maize exports Source: 1955-87 for all areas and 1955-95 for Japan and Taiwan: Ministry of Finance, various dates. Foreign Trade Statisticsof Thailand,Ministry of Finance, Bangkok; 1988-95 excepr Japan and Taiwan: Institute of Developing Economies, n.d. World Trade Database (AIDXT),Institute of Developing Economies,Japan ExternalTrade Organization, Chiba. diversification of exporting countries would stabilise feed quantities. Thai maize was exported during the off season for US maize. chicken farmers peferred the more yellow colour of Thai maize. in the 1950s, Japan had drastically scaled down rice imports, thus the Japanese government needed more imports from Thailand to decrease the trade surplus, a serious diplomatic issue between the two countries at the time. Japanese traders were eager to reach import thresholds in order to be allowed to join a government-supported feed program (JETRO 1979; Asami 1976). Between 1957 and 1959, Japan greatly increased its total maize imports, including those from Thailand. Because most Thai maize exporters were petty traders who had no access to modern storage and transport facilities, there arose quality and delivery problems associated with moisture, foreign matter and weight loss (Gaimusho 1960; Mizoguchi 1962). In order to control competition among importers and pressure Thai exporters to solve quality and delivery problems, Japanese importers formed the Thai Maize Importers' Council (TMIC), with strong assistance from the Japanese government. Many Thai traders had already formed the Thai Produce Exporters Association (TPEA) in 1956 (Nititham 1978), and the two associations made their first agreement for the 1959160 growing season. This agreement focused on quality control. It defined maize quality standards for each shipping month, and designated inspection companies to check cargoes. In order to support this private contract, the Thai Ministry of Commerce declared maize a 'standardised commodity', meaning its quality would be legally defined. These measures helped to eliminate existing quality problems and persuaded Japanese importers to buy more maize from Thailand. Over the next few years, a series of quality, delivery and price problems plagued maize trade between the two countries, and a contract between them ended in 1981. By this time, maize exports to Japan had already declined considerably because of Thai exporters' dissatisfaction with the contract price and inability to compete with US exporters, who used large bulk loading facilities. After export contracts to large markets disappeared, Thai maize exporters turned to Asian markets that did not have bulk loading facilities andlor were very close to Thailand. When Thai maize production decreased because of a crop failure in 1987, exports dropped to about half the previous year's level. m e n production increased in the following year, export volume failed to return to its previous level because importers had shifted to other exporters, and demand for maize in Thailand increased. Because the domestic feed market was protected by high import duties, Thai maize could be sold domestically at a higher price than in export markets. Now, Thailand is a net maize importer. Tapioca Up until the mid 1980s, Thailand shipped most of its tapioca products to Europe (Table 8.5). Even before the Second World War, European countries had used tapioca as animal feed; in the 1960s,however,demand increased tenfold, for the following reasons new intensive feeding systems needed more non-fibre feeds the European supply of such feeds was limited t;i Table8.5 Globaltapiocaexports, exportsfromThailand and selected areas,Thaimarketshare, and per cent of A volumeimportedbyWesternEurope World Thailand Exports from other areas (tonnes) %2 Per cent Per cent Latin 2 Exports imported by Exports of world America, (tonnes) W. Europea (tonnes) exports Indonesia Africa Caribbean Other aImports to Western Europe as a percentage of total world imports F Source: Food and AgricultureOrganization of the United Nations, n.d. FAOSTAT, FA0 online database, http://apps.fao.org/default.htrn. 0 after 1962, while the European Community imposed high duties on imported feed grains, it did not on certain tapioca products (UNCTADJGATT 1977; Siriprachai 1987). In 1962, the volume of tapioca exported by Thailand was only one- ninth the volume exported by the continent of Africa, the world's leading tapioca exporter. In 1963, European livestock producers increased their imports. At the same time, however, African exports decreased, because tapioca was an important staple in the diet and a large part of production had to be reserved for domestic consumption. Thailand, on the other hand, could swiftly respond to the increased demand becauseThai people do not eat tapioca as a staple. Thus, in 1964, Thailand became the leading tapioca exporter, shipping nearly 75 per cent of the world total; it remains the leading exporter today. Many scholars regard 1968's Kennedy Round of trade talks as responsible for Thailand's leading position; in fact, however, because tapioca is not a staple food in Thailand, the country was able to take the lead even before 1968 In 1965, the European Common Agricultural Policy finally imposed import duties on tapioca; in 1968, however, the Kennedy Round of trade talks capped these import duties at 6 per cent of import value, making tapioca more competitive than feed grains. Thailand could exploit this advantage because its private exporters were able to respond quickly to changes in market demand, and in 1969 Thai tapioca exports increased !greatly. Tapioca pellets became the preferred form of animal feed in Europe, and Thailand responded quickly by building pellet production plants, something African countries were unable to do (Ithijarukul n.d.; UNCTADJGATT 1977). Shipping facilities are another area where Thailand gained market share in tapioca. As a low-priced commodity, tapioca is shipped in bulk. In the 196Os, tapioca shipping ports were limited to ships smaller than 15,000 tonnes; in the 1970s, however, a private company set up a loading facility at Si Chang Island in Chon Buri Province to load pellets directly from silos to cargo ships (TDRI 1987). Indonesia had no such facility, so shipping &om it was more e~pensive.~ Compared to private companies, the government did little to promote tapioca exports except apply the factory code to tapioca facilities and set quality standards (Siriprachai 1987). Quality inspections were performed by private companies. The government's passivity in 326 RURAL DEVELOPMEKTAND AGRICULTURAL GROWTH supporting tapioca production can be attributed to the belief then current that cassava, the raw material of tapioca products, negatively affected soil fertility. The boom in tapioca exports ended when France set import controls in 1977 (Ruam P~achachat,8 April 1987). Tapioca prices fell considerably, and the Thai government took control of the construction and expansion of tapioca product factories. In 1979, the Ministry of Commerce issued a decree to control the volume of tapioca exports. In 1980, it made an agreement with the European Community (EC) to limit Thai tapioca exports to the EC to 5 million tonnes per year during the 1981 and 1982 season^,^ and 4.5 million tonnes per year during the following two seasons. Trade volumes for the 1985 and 1986 seasons were reduced still further. By this agreement, Thailand could no longer expand its tapioca exports. When trade controls were needed, the government appeared as an important player. Since the EC still taxed imported feed grains, the tapioca price in the EC was more favourable than in any other market. This situation led the government in 1981 to adopt a quota system for exporters, includinga 'bonus' quota as a reward for exports to countries outside the EC. In 1987, the bonus quota increased, and exporters became ever more eager to ship to non-EC countries (Tasaka 1991). As a result, during the 1980s, Thai tapioca exports increased: even after the EC had imposed import limits (Table 8.6). The bonus quota system thus contributed to expansion of both the volume of Thai tapioca exports and their market. The reason Thailand became the major tapioca exporter was that no other country could supply the market when demand in the EC increased rapidly in the 1960s. The large and stable market of the EC stimulated Thai firms to invest in production and transportation facilities, and Thai exporters to form a carrier company to compete with the big European carriers (TDRI 1987). The government's role in the expansion was limited to the establishment of quality standards. When export volumes needed to be controlled, the government took on a major role, introducin a quota system to cover exports to the EC and at the same g time promoting exports to countries outside the EC. Though many scholars condemned the system because it enabled government officials to engage in rent seeking and corruption, the bonus quota positively affected Thai tapioca exports. Table 8.6 Tapioca exports, 1981-97 ('000 tonnes) Year Stock quota Bonus quota Exported to EC Total exporB 1981 5,500 1982 7,331 1983 4,964 1984 5,506 1985 4,920 1986 4,833 1987 5,500 1988 5,500 1989 5,492 1990 4,583 1991 5,252 1992 5,250 1993 5,747 1994 4,397 1995 3,107 Source: Total exports: Office of Agricultural Economics, various dates. Agn'mltural Statisticsof Thailand, Office of Agricultural Economics, Bangkok; all others: Siriprachai, Somboon, 1998. Controland Rent-Seeking:the role of the sute in the Thai cassavaindustry, Lund University Press, Lund:164. Sugar In the past, Thailand was a net importer of sugar, though sometimes it exported sugar when the government offered export subsidies. Today, Thailand is a major sugar exporter (after the European Union and Brazil), in competition with Australia and Cuba (Figure 8.9). In 1976, exports increased dramatically to more than 1 million tonnes; in 1982, they passed 2 million tonne^.^ Between 1976 and 1981, the world market for raw sugar expanded by nearly 2 million tonnes; from 1982 to 1994, however, it changed very little. Thailand was able both to capture a large share of the world market during the period of expansion and enlarge its share during the period of stagnation that followed. Up until the 1960s, the Thai sugar industry was an import substitution industry, the government's policy being to control production so as to satisfy domestic demand, not create a surplus. In reality, however, sugar prices fluctuated widely, depending on domestic harvests and world market movements, and a rising price in one season led to production increases in the following season, thus creating a s~rplus.~ 328 R U W DEVELOPMENT AND AGRICULTURAL GROWTH , Figure8.9 Sugar exportsfrom majorexportingcountries -- Mllions of mtric hropean~ m n i t (15 nations)-Australia y - tonnes -Brazil Qba -Philippines --Thailand Source: Food and Agriculture Organization of the United Nations, n.d. FAOSTAT, FA0 online database, http://apps.fao.org/default.htm. During the 1960s, the Thai government implemented three strategies to stabilise the domestic sugar price. In 1967, the government established the Sugar Price Stabilization Bureau (SPSB) to manipulate buffer stock. Because the fund the Bureau had to draw on was small, however, when world prices were extremely high or low the buffer stock system did not work well. In 1961, the government's Sugar Code set an upper production limit. The 1968 Sugar Code allocated quotas for domestic supply (white sugar) and export (raw sugar) to each factory. Under this policy, the government determined which type of sugar each factory should produce first each season (Wannitikul 1981). In 1969, it prohibited the construction and expansion of sugar processing plants (Manarungsan and Kaeothep 1989); once prices rose, however, such controls stopped working, moving the government to relax them in order to lower prices. The limited success of these first two strategies led to a third: in the 1960s, the government recognised exports as a way to stabilise the local sugar price. With the government's allocation of export quotas in 1969 and quotas for both domesticand export markets in 1970, sugar millers could recoup their losses from the export market with returns from the domestic market, where the price was set higher than in the world market (Yamamoto 1998). Recognising the difficulty in balancing domestic sugar supply and demand, the government began to consider conversion of the sugar industry into an export industry. Between 1972, when a world sugar boom began, and 1977, the volume of world sugar trade expanded by nearly 7 million tonnes. In June 1972, the Thai government declared an official retail price for sugar. When this proved to have no effect, the government allowed the production capacity of sugar mills to increase, and the number of sugar mills rose from 30 in 1972173 to 41 in 1974175 (Jessadachatr 1977). Meanwhile, total production capacity rose to nearly 2.5 billion tonnes (Yamamoto 1998), and in June 1974 the government began to levy an export premium on sugar (Manarungsan and Kaeothep 1987). During this boom, the structure of Thai sugar exports changed significantly. Before the boom, the government had allowed only the Thailand Sugar Corporation (TSC) to export sugar; during the boom, however, the TSC made an unfavourable long contract with an international broker. The contract was subjected to severe public criticism, and in December 1974 the government was forced to admit new exporters, thus introducing competition into the sugar export business (Ando 1979; Jessadachtatr 1977). In 1976, the sugar boom passed its peak, and world sugar prices fell. Thailand, however, already had in place a system to cover export losses with profits from the domestic market. Beginning in 1974, government off~cialsmonitored sugar mills to ensure the fair and accurate allocation of quotas (Yamamoto 1998). Since 1982, the subsidisation of sugar exports using profits from domestic sales has been formally institutionalised as the 'sharing system'. Under this system, total output from sugar mills is channelled into three quotas: Quota A, for domestic supply, its price set by the government; Quota B, for export by a government-appointed exporter; and Quota C, for export by private exporters. The sugarcane price is automatically set at 70 per cent of the sugar price, as calculated from the prices in these three market channels. 330 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Thailand's involuntary consumer subsidy for sugar exports does not completely explain the country's success in sugar exports, however, since other countries also subsidise sugar exports. When Thailand entered the export market, it faced high costs; by the late 1980s, the cost of Thai raw sugar was the second lowest in the world after that of South African raw sugar (Lichit 1994). Thailand's domestic sugar price dropped below the sugar prices of most other major exporting countries, and its subsidy was also relatively low. In the world sugar market, the price had been set at an extremely low level that did not cover production costs for most exporters; in such a situation, only reduced production costs would help alleviate a defi~it.~ Production costs could be reduced through increasing mill size. The sugar boom of the 1970s stimulated such expansion, and between 19651 66 and 1985186 the production capacity of Thai mills increased ninefold (Yamamoto 1998). Because of the high salvage cost of reduced production capacity and the significant political influence of cane growers, which made reductions in production unacceptable, the government chose to dump the country's surplus sugar on the export market with a subsidy. Thus, Thai sugar exports were able to expand because of government intervention in the market, even though the government's intention was not to promote exports but to balance domestic supply and demand and stabilise the domestic price. Thai sugar's export success was an unexpected result of this government policy of attempting to satisfy both farmers and consumers. Frozen chicken Thailand's share of the world marketfor frozen chicken has never exceeded the 7 per cent share it achieved 1991. In that year, the United States led frozen chicken exports with three times the volume of Thai exports, and Brazil was second with about twice the volume of Thai exports. Although Thailand has never been a major exporter of frozen chicken, Thai policymakers have focused their attention on it in the hope that, together with other high value-added commodities, it will lead the country into a new agricultural era. Frozen chicken accounts for 7 per cent of Thailand's total agricultural exports. Though its export value remains small, the rapid growth it has made since the 1970s is noteworthy. Thailand began commercial exports of frozen chicken in 1976. Within two years, the volume of Thai frozen chicken exports had increased fivefold; by 1986 the volume was 33 times larger than in 1976. The rapid increase continued until 1992, a crucial drivin factor being the growth of Thai frozen chicken's share of g the Japanese market (Table 8.7), from just 2 per cent in 1976 to 36 per cent in 1992. Thailand's first exporter of frozen chicken was a feed producer, Charoen Pokphand Co., Ltd. Charoen Pokphand began producing chicks for farmers in a joint venture with Arbor Acres in 1970. The emphasis at the time was on the domestic market, made up mostly of US armed forces based in Thailand during the Vietnam War. Realising that it could not rely solely on the domestic market, in 1973 Charoen Pokphand signed its first export contract with Asahi Broiler. In the beginning, Charoen Pokphand's processing technology was very primitive. When its first shipment of chicken did not meet Japanese standards, it suspended exports and sought technical advice from Asahi Broiler. Another Japanese company, Kanshoku Co., Ltd., began to import chicken from Thailand to satisfy the Japanese people's changing dietary needs at a time when the country's major companies refused to change their product (Yamada 1986). Though Thai chicken was more expensive than US and Chinese chicken, Thai chicken exports to Japan increased (Chikusan shinko jigyodan 1980). Crucial to this increase were changes in the forms in which chicken was consumed in Japan. Increasingly, Table 8.7 Volumeof chicken exported fromThailand (tomes) Year Total volume Exports to Japan Percent of total 1973 135 135 100.0 1974 337 337 100.0 1975 373 367 98.4 1980 18,503 17,430 94.2 1985 37,840 33,147 87.6 1990 138,859 108,121 77.9 1995 149,799 115,233 76.9 1996 137,167 101,678 74.1 1997 150,596 95,626 63.5 Source: Office of Agricultural Economics, various dates [c]. Thailand Foreign Agricultural Trade Statistics, Office of Agricultural Economics, Bangkok; Ministry of Finance, various dates. Foreign Trade Statisticsof Thailand, Ministry of Finance, Bangkok. 332 RURAL DEVELOPMENTAND AGRICULTURAL GROWTH Japanese were eating processed foods and buying food in take-out markets. By the end of the 1980s, around 60 per cent of chicken was consumed in such forms (Nihon shokucho kyokai 1989). Because of Thailand's low labour costs, its processed food production was very competitive, and because of their location Thai processors were able to respond quickly to changing Japanese demand. Thailand only realised its competitiveness in production costs, however, when it learned to control quality to meet Japan's strict standards for imported foods. To this end, Kanshoku offered technical assistance to its then counterpart, Saha Farms, while Centaco, a joint Thai-Japanese venture, also introduced Japanese technology into processing lines. Such efforts on the part of both Japanese and Thai companies helped Thai frozen chicken earn a good reputation for qualityin terms of freshness, shape and freezing temperature (Chikusan shinko jigyodan 1980). This swift response to processing changes was aided by competition among processing companies, whose total capacity grew fourfold from 1975 to 1978 (Suehiro 1987). Before the broiler industry boom, in order to control volumes, seven companies had built facilities that integrated the entire chicken production process, from chick hatching to product sales. Though three of these companies were joint Thai-Japanese ventures, they made these daring investments in broiler production even without any long-term forecast of Japanese demand (Yamada 1986). In all of this development, the government played a secondary role. Though chicken processing was included among the projects eligible for the government's investment promotion scheme in 1967, it was only ten years later that it took any action in this respect. In an amendment to the 1977 Investment Promotion Act, the Board of Investment (B0I)'s authority to make promotion decisions was strengthened, and import duties on facilities and raw materials were waived at 100 per cent and 90 per cent respectively. The Act also gave the same preferential treatment it gave to factories located around Bangkok to factories in the provinces (Chantharanimi 1988:42), thus allowing all broiler processors to receive investment promotions (Christensen 1993). Another role played by the government lay in negotiating with Japan to reduce import duties. Previously, Japan had levied high duties on imported chicken meat. In 1980, however, following the GATT Tokyo Round of trade talks, Japan divided chicken into two tax categories- bone-in and boneless. The former category, made up mostly of chicken from the United States, enjoyed a tax reduction from 20 per cent to 17.5 per cent, then to 13.8 per cent in December 1981 (Saito 1985). At Thailand's urging, Japan reduced the duty on boneless chicken from 20 per cent to 18 per cent in April 1984, to 14 per cent in January 1986, and to 12 per cent in 1989 (Miyazaki 1986; Nikkei Sangyo Shinbun, 13 June 1986; Prachachat Thurakit, 1 February 1989). As a result, boneless chicken imports from Thailand to Japan increased. It should be noted that the government did not control the slaughter of chicken as it did the slaughter of cattle and swine (World Bank 1987). Even before the emergence of the modern broiler industry in Thailand, many merchants had already been engaged in chicken slaughtering, and the government continued to allow private enterprises to play the major role in the business (Puaphonsakon 1980; JETRO 1984). Canned pineapple World trade in canned pineapple has increased constantly for many years (Figure 8.10). Thailand entered this export market in the late 1960s and became a major exporter in the late 1970s. Even though it was a latecomer compared to other exporters, in less than a decade from the time of its entrance it became the second largest exporter after the Philippines, and it overtook the Philippines in 1985. Thailand increased its exports and for a time controlled half the world market. When the United States, the largest importer of canned pineapple, imposed anti-dumping duties on the Thai product in July 1995, the volume of Thai canned pineapple exports decreased drastically (Chulalongkorn University 1997). In 1998, however, Thailand was still the second largest canned pineapple exporter, after the Philippines. The major markets for Thai pineapple are the United States, Europe and Japan. In the United States, up until the 1950s, Hawaii was the major pineapple-producing area. In the early 1960s, rising land and labour costs in Hawaii caused the major producers to move overseas (Business in Thailand 1976). As US pineapple production decreased, US imports of canned pineapple increased. The Philippines and Thailand both seized this opportunity, though each in its own way. The major US producers and distributors, such as Dole and Del Monte, moved out of Hawaii and established their plantations and processing factories in the Philippines, which became the top exporter in about 1970. Canned pineapple production in Thailand, on the other hand, began in 334 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Figure 8.10 Exports of canned pineapplefrom major producing countries 600 Thousands of ,200 I mtric Thousands of metric tonnes) tonnes Source: 1963-72: Taiwan Canners Association,1972. Taiwan Exports of Canned Food, 1972,Taiwan CannersAssociation,Taipei; 1973-78: Department of FarmersService, Taiwan, 1982. Agrcultural Trade Statisticsof Eiwan, R 0.C., Department of Farmers i Service, Taiwan, Taipei; 1979-85: Taiwan Statistics Office, 1986. AgriculturalTrade Statistics of Republic of China 1985, Council of Agriculture, Executive Yuan,Taipei; 1986- 93: Taiwan Statistics Office, 1994. AgriculturalTrade Statistics of Republic of China 1993, Council of Agriculture, Executive Yuan, Taipei; others: Food and Agriculture Organization of the United Nations, n.d. FAOSTAT, FA0 online database, http:ll apps.fao.org/default.htm. a government factory and was aimed only at the domestic market (Kiatpanapikul 1978). The first factory to produce for the export market-the Thai Pineapple Canning Industry-was established in 1967 as a joint Thai-Japanese venture (ESCAP 1979), and was supplied by already existing domestically oriented pineapple producers. Another Thai firm built a processing plant in 1970, and a year later Dole entered Thailand. Up until the mid 1970s, Thailand had eight large factories producing for export. These factories purchased 70-80 per cent of their pineapples from small farmers and the balance from their own plantations (ESCAP 1979; Bangkok Post, 12 January 1987). Though foreign investment gave the Thai canned pineapple industry its start, Thai firms were quick to respond, and domestic capital has always dominated production in the country. In 1978, Dole's share of production was only 18 per cent. Pineapple canneries continued to be built until 1996, when their number passed 30, and this expansion made Dole's share of the market even smaller (Chulalongkorn University 1997). In the US market, Thailand's largest competitor was the Philippines. Del Monte had had a plantation in the Philippines since the 1920s; Dole entered the country in 1963 (Hayami et al. 1988). This foreign investment enabled the Philippines to increase exports rapidly in the 1970s, with Dole and Del Monte dominating the industry (IADE CRC c. 1996). In the 1980s, however, as investment from new countries enabled Thailand to increase production, plantation expansion in the Philippines became nearly impossible and wages soared, leading to the stagnation of the country's exports (Ministry of Industry c. 1985; Business Review 1984). Philippines pineapple exports also suffered from political instability and natural disasters (Rakthammakit 1990; Kiatikamchai 1992; Chakatpanyawai 1992). In Thailand, on the other hand, because of the large domestic and fresh market, a large area had been planted in pineapple from the beginning, and enough pineapple was produced to increase the country's exports. Reorienting production from domestic to export markets, however, was only possible with the cultivation of a new variety of pineapple. From the late 1970s onwards, though harvested area remained stable, exports increased rapidly (Figure 8.11). Another important market for Thai pineapple exports was Europe, where CBte d'Ivoire was the major competitor. This African country had a transportation advantage over Thailand; also, as a former French colony, it was exempted from import taxes (TheInvestor, 9-1 1July 1976). CBte d'Ivoire's pineapple-planted area, however, was only about 10 per cent of Thailand's. This disparity enabled Thailand to increase vastly its pineapple exports to Europe during the period 1975-90. In addition to the Philippines and CBte d'Ivoire, Taiwan and Malaysia were also competitors in canned pineapple exports. Taiwan, however, faced the same problem of high land prices and wages as Hawaii, and it withdrew from the export market in the early 1980s. Meanwhile, Malaysia soon turned to oil palm, which was more profitable than pineapple (Kiatpanapikul 1978). 336 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Even before it entered the export market, Thailand's capacity to produce pineapple was large. The only resource lacking was capital. Thus, while foreign firms invested first in Thai pineapple canneries, Thai firms were quick to follow. Though unable to sell their products in the United States under their own brand names, they could sell raw material to well-known distributors. Government investment promotion policies supported this industry. All factories established in the 1970s benefited from the Board of Investment (TheInvestor, 13-18 July 1976; Kiatpanapikul 1978); later, factories located in provincial areas received special privileges, such as tax exemptions on imported equipment. In addition, foreign investors were granted the right to own land, received guarantees that their land would not be nationalised, and were allowed to transfer money to other countries. Figure8.11 Pineapple-harvestedarea in the Philippines,Thailand and CBted'Ivoire ,Thousands of Thailand - Cote d'lvoire 0 - 1961 1966 1971 1976 1981 1986 1991 1996 Source:Food and Agriculture Organization of the United Nations, n.d. FAOSTAT, FA0 online database, http://apps.fao.org/default.htm. PRODUCTION RESPONSES TO MARKET OPPORTUNITIES As noted above, Thailand has successfully responded to demand from overseas for certain agricultural commodities by quickly adapting production to market changes. In this section, we analyse the resources and institutions that enable robust agricultural production. Land resources and the land system During the export boom period, all of the export commodities discussed in the previous section except chickens rapidly expanded their planted area. Thailand could expand its farmland because large areas of land had not yet been occupied or cultivated in the 1950s. When overseas demand reached Thai farmers, this unused land was converted to farming. Farmland expansion was conspicuous in the 1960s but slowed in the mid 1970s (Fi ure 8.12). Production increases in maize, cassava and g kenaf occurred at the same time as this rapid farmland expansion; other commodities enjoyed production increases at different times. The considerable increase in rice exports occurred in the late 1970s, rubber exports increased rapidly from the early 1980s onwards, and sugar became a major export commodity after the late 1970s. Meanwhile, canned pineapple exports soared without adding new production areas. Farmland could only expand if forest was sacrificed (Figure 8.12). As late as 1961, Thailand's ample flats and gentle slopes were covered with forests, especially in the Northeastern region. By 1988, forest area in the Northeastern region had declined to one-third the 1968 area. In the mountainous, upper Northern region, it declined by only 20 per cent. Moreover, the land system allowed and even promoted occupation of forest areas, which the government defined as public land but also allowed farmers to clear for productive use, since it could not provide any revenue to the government as forest. If people occupied a certain plot of land (chap chong), the government issued them a certificate called bai chong. If they turned this plot into farmland within a certain period, they received a land title or a certificate equivalent to land ownership. Land titles, however, could only be drawn up on the basis of cadastral maps, and the government's progress in cadastral surveying, begun at the beginning of the twentieth century, was very slow, which meant that most land titles were issued in the 1990s. Certificates equivalent to land ownership, on the other hand, were given out increasingly rapidly from the late 1970s onwards (Department of Land 1976, 1992). 338 RURAL DEVELOPMENT AND AGRICULTURAZ.GROWTH Figure 8.12 Trend of farmland expansion and proportion of forest land Roportion of forest land (per cent) Roportionof forest 25 Farmland(nillionsof Farniand(nillionsof hectares) 20 15 10 5 0 Source: Office of Agricultural Economics, 1979. Land Utilization of Thailand 1950/51- 1977/78,Office of Agricultural Economics, Bangkok;Office of Agricultural Economics, 1991~.Kan cbai tbidinpbuea Kan kaet khongpratbet tbaipi2534 [Land utilisation for agriculture in Thailand, 19911,Ofice of Agricultural Economics, Bangkok; Royal Forestry Department, 1988. Forestry Statistics of Thailand 1988,Royal Forestry Department, Bangkok; Royal Forestry Department, 1996. Forestry Statisticsof Thailand 1996 Royal Forestry Department, Bangkok. Delays in government certification did not prevent people from occupying forests, however, since local people could take possession of unoccupied land through common recognition among neighbours. Certain wealthy people also laid claim to large areas of forest, seeking capital returns. In 1988, uncontrolled deforestation caused a landslide in Southern Thailand, moving the government to consider forest protection and to ban occupation of forestland. Infrastructure affecting farmland expansion Road construction also affected farmland expansion (Mundlak 1997). One of its positive effects was to give regional farmers access to world markets, as for example when the major cassava and kenaf production area in the Northeastern region was connected to Bangkok by a national highway in the 1960s (Phongpaichit and Baker 1995; Siriprachai 1987). Other effects of road construction on farmland expansion, however, are the subject of considerable disagreement. A study on the dynamics of Thai agriculture during the period 1961-85 carried out by the Thailand Development Research Institute (TDRI) found that the coefficient between farmland expansion and road length was low (TDRI 1988) and that road construction in fact tended to intensifjr land use. When Tongpan (1970) summarised the factors related to maize area expansion, he attributed greater importance to malaria eradication than road construction. Cropper, Griffiths and Mani (1997) also found the coefficient between road construction and farmland expansion to be low, but concluded that more than 30 per cent of deforestation was caused by the increased road density. There is clearly no agreement about the specific effects of road construction on farmland expansion, though all observers agree that road construction in one way or another positively affects agricultural development. Irrigation is also important for arable land expansion, but quantitative data on farmland increases caused by irrigation construction have yet to be collected. For dry-season planting, only data for paddy are currently available. The area of the land planted in dry-season paddy has been about 7 per cent of total paddy land for many years, indicating that land planted in the dry-season crop has expanded less rapidly than land planted in the rainy-season crop. Agricultural research and extension Improving productivity is another way to increase overall production. The crops considered to have most increased productivity as a result of agricultural research and/or extension are rice, rubber and maize, with the increase for rubber the most dramatic (Table 8.8). Sugarcane also experienced a notable yield increase in the 1960s, because it was planted on newly cleared-and very fertile-land. After it began to be exported in the 1970s, however, yields stagnated until very recently. Rice yields also increased. The yield hike that occurred during the 1960s was brought on by the expansion of irrigation systems in Central Thailand, which stabilised water procurement in rain-fed paddy fields, and by new cultivation of rich forest land in the Northern region. From the late 1970s onwards, dissemination of new varieties and irrigation 340 RURAL DEVELOPMENT AND AGRICULTURALGROWTH expansion were major factors contributing to increased yields (Dixon 1999).This was the period when Thailand responded to increased export demand, and dissemination of new varieties and dry-season paddy cultivation enabled it to compete with its rivals, especially the United States. Even before the Second World War, the Thai government had established a research institute to run a seed improvement program. After the war, the FA0 and the World Bank strongly recommended seed improvement, with the World Bank providing considerable funding in support of Thailand's program (World Bank 1959). Following the establishment of the International Rice Research Institute ( I N ) in the Philippines, the Thai government embarked upon a joint research program with IRRI in 1966 (Ministry of Agriculture and Cooperatives 1992). This research program produced RD1 (in Thai, Ko.Kho.l), which it released in 1969 (NESDB 1978). RD1 inherited the characteristics of IR8: non-photoperiod sensitivity, short gowth period, short stalk, and yields three times greater than those of indigenous varieties (NESDB 1978; Welsch and Tongpan 1973). Research and extension activities were not limited to introducing new varieties; indigenous varieties were also improved. The improved indigenous varieties were broadly disseminated for a rainy-season crop, while RD varieties were dominant for the dry-season crop. The improved varieties soon became popular. In 1971, Sriswasdilek (1973) surveyed three villages in the Don Chedi District of Suphan Buri province, one of the most progressive rice-growingareas in Thailand, and found that 41 per cent of the rainy-season crop area was planted in high-yielding varieties (HYV). At the time, modern varieties (including improved indigenous varieties) prevailed only in the irrigated area; later, however, they became dominant in other areas as well (Setboonsarng et al. 1991). In 1986, 95 per cent of rain-fed paddy in Northeastern Thailand and 66 per cent of rain-fed paddy in Northern Thailand were covered by the modern varieties,while in the irrigated areas of the Central, Northern and Northeastern regions, the dissemination rate reached 89 per cent, 77 per cent and 95 per cent respectively. During the years when Thailand was becoming the top rice exporter, seed improvement and dissemination were already well underway. Research was almost always carried out by the government, which distributed successfully improved varieties to selected farmers for Table 8.8 Yield trends for seven crops, 5-year averages with base year of 100 Rice Rubber Maize Sugarcane Cassava 100 100 100 69 100 102 98 105 87 94 108 94 110 100 84 102 99 106 98 91 111 106 116 90 98 117 199 122 104 88 123 304 137 103 84 132 363 172 123 88 a'Jute-like' fibres in the original database. Y~elddata from before and after 1975 are inconsistent. Source: Food and AgricultureOrganization of the United Nations, n.d. FAOSTAT, FA0 online database, http://apps.fao.org/default.hm. multiplication. These farmers then disseminated the improved varieties. Of the total production increase from 1974175 to 1984185, 26 per cent came from the dry-season crop. While Siamwalla et al. (1990) argue that irrigation did not contribute to improved labour productivity in Thai a riculture, it clearly played a significant role in increasing total g production during the period when world rice demand surged. Improved varieties of maize have also been successful in Thailand. When the area planted to maize began to expand rapidly, a cooperative program between the Thai and US governments introduced the Guatemala variety (Muscat 1990; Setboonsarng 1990), which produced much higher yields than traditional varieties. In 1956, Thai breeders started research to develop Phra Bhutthabat (PB) varieties from Guatemala (CPCDT c. 1962; NESDB 1979). In the mid 1960s, a group of Thai breeders returned from studying in the United States and joined a maize improvement program supported by the Rockefeller Foundation (Setboonsarng 1990). The first Phra Bhuttabat variety was released in 1965, and by 1974 the share of Phra Bhuttabat varieties had climbed to 80-90 per cent of the total area panted to maize (Setboonsarng1990). A year after release, Phra Bhuttabat varieties were affected by downy mildew,' despite having performed better than Guatemala during research trials. The government then developed a resistant variety, Suwan I, and released it in 1974. By 342 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH 1981, this variety had been adopted by about 75 per cent of farmers (Setboonsarng 1990), and in later years was used as the basis for developing hybrid maize. The government improvement program that developed these maize varieties is considered a success (TDRI 1995). Rubber also contributed to the successful growth of agricultural exports, with new varieties being replanted from the late 1970s onwards. A rubber tree will start to yield latex at an age of seven years, thus areas replanted in the late 1970s provided the mid 1980s yield increases that allowed Thailand to expand its exports. Up until 1976, money from the government budget and a levy on rubber exporters supplied the funds for replanting, but this was not enough to expand the planted area. A 1977 loan from the World Bank made possible more, and more rapid, land expansion. Most of the replanted trees were of the PRIM600 variety developed in Malaysia (Picciotto and Kopp 1994). While the loan and the variety on which the increased production of the 1980s was based came from outside the country, nonetheless the government's replanting program was also crucial to the success of rubber exports. Supply and demand for capital and capital goods Fertiliser and machinery have been important factors in the development of Thai agriculture since the Second World War. Four-wheel tractors, useful for clearin forests (TDRI 1988), tended to be owned by rich g farmers and merchants, and hired by small farmers to clear and cultivate their land (Moerman 1968). The two-wheel tractor (power tiller) has been the most popular machine among small farmers, but there is disagreement about whether this machine helped farmers to increase output or decrease production costs. Fertiliser was an important factor in the increase of production volume. Siamwalla et al. (1987) argue that fertiliser's contribution to improved productivity has not been statistically proven. While it is true that up until the 1980s, the volume of fertiliser input per acre in Thailand was much lower than in neighbouring countries, this does not necessarily mean that higher rates of fertiliser did not contribute to increased production. The largest portion of fertiliser (30-50 per cent) has been used for paddy cultivation. Fertiliser input per unit of area increased considerably from the mid 1970s, because dry-season paddy, which requires 7-8 times more fertiliser than a rainy-season crop, was expanding at the time. Beginning in the 1980s, inputs for rainy-season paddy also increased, to the point where the volume of fertiliser used in 1985 was twice that used in 1975. These increases were also made necessary by the introduction of new varieties, since most upland crops need much more fertiliser than does rice. The government began promoting fertiliser very early in the process of development. In 1963, though fertiliser had already been produced domestically, the government established the Thai Fertilizer Co. and allowed it to monopolise fertiliser imports so as to keep the domestic price high (Manarungsan 1976; Prichametta 1996). Government control ended in 1973, however, and ever since, except for a short period in 1978, fertiliser imports have been carried out without government intervention (Siamwalla and Setboonsarng 1989). During this period of duty-free fertiliser imports, Thailand was affected by two world oil crises, one of the effects of which was to cause world fertiliser prices to soar. In 1973, a natural gas field was found in the Gulf of Thailand, and the government began a new project of domestic fertiliser production. Because it failed to persuade private companies to invest in this project, the government was forced to become the main investor (Prichametta 1996). To distribute cheap fertiliser to farmers, the Marketing Organization for Farmers, a government agency, sold fertiliser below the market price and allowed farmers to pay for it after the harvest (Tongpan 1993). Bungling by the government and its agent, however, often caused fertiliser to be distributed too late. Government-promoted farmer groups distributed fertiliser to individual farmers, but problems occurred in money collection (Phasuk 1985). Since then, most farmers have purchased fertiliser from merchants in markets. The situations of sugarcane and rubber are different. Sugar factories extended loans, partly subsidised by the government, to their contract growers. Meanwhile, at the time of the replanting of rubber, fertiliser for rubber was distributed for free. The government has also implemented a program for farmers to acquire production capital. In the 1970s, informal sources of funding came to be partly replaced by formal loans (Table 8.9). The most important institution in this respect is the BAAC, which provides loans to farmers and farmer organisationsat a lower interest rate than commercial lenders. 344 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH The expansion of the BAAC's service, however, remained limited up until the late 1980s, when the majority of Thai farmers became clients. Siamwalla et al. (1993) argue that in the 1980s, the BAAC's clients were mainly better-off farmers, and that most farmers could not access its or the co-operatives' loans because more than 40 per cent of these loans were collateralised by land. Because the collateral was indispensable for a long-term loan but land titles took so long to be issued, most farmers found it impossible to acquire low-cost capital. Table8.9 Originand share (percent)of loanamountsduringfarm year per farmhousehold" 1953b Non-institutional 83.4 Relatives 43.8 Neighbows n.a. Private businesses 39.8 Institutional 15.2 BAAC Commercial banks 0.0 Cooperatives 15.2 Farmers Groups - Others n.a. Tota1 100.0 aFigures represent average farm household loans calculated on the basis of total households, including those that did not borrow. Figures for 1953 are adjusted from original data that show average farm household loans calculated on the basis of households that actually borrowed. Source: 1953: Office of the Under-Secretaryof State, Ministry of Agriculture, 1955. Thailand Economic Farm Survey, 1953,Ofice of the Under-Secretaryof State, Ministry of Agriculture, Bangkok; 1962: Thisyamondol, Arromdee and Long, 1965. Agrcultural Credit in Thailand: theory, aha, policy, Kasetsart University, Bangkok; 1978179: Ofice of AgriculturalEconomics, 1980. Phawa nishin khong kasetrakon pi kan phopLk 2521/2522 [Situation of farmers' debt, 19781791, Office of Agricultural Economics, Bangkok; 19801 81: Office of Agricultural Economics, 1983. Phawa nishin khong kasetrakonpi kanpho pLk 2523/2524[Situation of farmers' debt, 19801811, Ofice of Agricultural Economics, Bangkok;1986187: Ofice of Agricultural Economics, 1988. Nisin kasetrakonpiphopbk 2529/30[Farmers' debt, 19861871, Office of Agricultural Economics, Bangkok; 1988189: Office of Agricultural Economics, 1991a. Nisin kasetrakonpiphopbk 2531/32[Farmers' debt, 19881891, Office of Agricultural Economics, Bangkok. Agribusiness and investment promotion policy An important role in the change of production systems to enable farmers to meet overseas demand for Thai exports was played by agribusiness. Agibusinesses were especially helpful in the maize, rubber, sugar, canned pineapple, and frozen chicken sectors. In the years preceding the release of the maize variety Suwan I, the government played a significant role in seed improvement; in 1980, private companies entered the business and developed hybrids from Suwan I, the first of which was Charoen Pokphand's DK 888, released in 1982. From 1985 onwards, subsidiaries of major international agibusinesses dominated the sector (Setboonsarng 1990). Even for Suwan I, an open-pollinated variety, the private sector plays a significant role in multiplying and distributing seed, since the production capacity of the government is limited. In 1987, 20 per cent of Suwan I was produced by the private sector (Setboonsarng 1990). It has become common for the government to cooperate with the private sector in research and development, with government research institutes providing manpower and the use of facilities to the private sector, and amicable relations among the researchers facilitating data sharing (Setboonsarng 1990). For Thai exporters to compete with other countries in rubber and sugar, factory-level quality and productivity had to be improved. Quality improvement of rubber sheets was realised through the efforts of Thai smokehouses and their Japanese counterparts. Sugar refineries have expanded their capacity to the highest level in the world and reached high levels of productivity. Canned pineapple and frozen chicken were introduced by agribusinesses, which expanded exports by competing with each other and received investment promotions from the government. The Thai government's investment promotion policy began with the establishment of the Industrial Promotion Committee in 1954. In 1960, the first Investment Promotion Act (IPA) was enacted and the Board of Investment (BOI) set up to make decisions on promotion. The 1954 law had already defined tax exemptions and reductions on imported equipment, imported raw materials, business income and export duties. Until the third version of the IPA was enforced in 1962, however, promotions tended to be given to investors who had connections with powerful politicians (Chantharanimi 1988). The situation began to change in 1966, with the establishment of the secretariat of the BOI, whose chief was elected from among technocrats. The BOI became increasingly independent from political influence. In 1977, when a new regulation declared that the chair of the BOI would automatically be occupied by the Prime Minister, the institute acquired much greater authority than before (Chantharanimi 1988; BOI 1978). While it has never given special treatment to the agro-industry sector, the I P S policy of promoting industries located in provincial areas has effectively meant agribusinesses receive more privileges than other industrial sectors. This was the institutional background that stimulated investment in canned pineapple and frozen chickens in the mid 1970s. During the second half of the 1980s, when the government emphasised NAIC-type development, the share of net direct foreign investment in agriculture and food processing increased. ALLEVIATING POVERTY: SUCCESSES AND FAILURES Reduction of absolute poverty Thailand is often cited as an example of a country that has successfully reduced rural poverty in absolute terms (World Bank 1996; UNDP 1999). While researchers have used different methods to calculate the incidence of poverty in Thailand (Krongkaew c. 1981), the results of such research generally show similar trends for rural poverty (Figure 8.13). At the beginning of the 1960s, 60 per cent of the rural population lived below the poverty line; within two decades, however, the figure had dropped to less than 30 per cent. During the first half of the 1980s, because of falls in the world prices of agricultural commodities, especially rice, the incidence of poverty increased, returning to the level of the mid 1970s. Soon, however, it went down again; in 1992, the rural poor made up less than 20 per cent of the population. Several social indicators of rural non-income poverty have also improved. In the mid 1970s, most of the rural population relied on public wells and rivers for water; in 1998, nearly one-half had water piped into their houses. In the mid 1970s, nearly 70 per cent of rural people had no toilet facilities; in 1998, the figure was negligible. In 1975, only 7 per cent of rural households had electricity; the census of 1990 showed that households without electricity were very scarce. Since the early 1970s, life expectancy at birth has increased by ten years; , Figure 8.13 Incidence of rural poverty in Thailand Share of population under thepoverty Source: Laovakul, Duangmanee,1993. Decomposition of Income Inequaliy and Povery in Thailand, Thammasat University, Bangkok;Ashakul, Charuma, 1985. Income Distributionin Thailand:an analysiswith policyimplications, UniversityMicrofilms International,Ann Arbor; Krongkaew, M., 1996. Thailandpoverymsessmentupdate, Thammasat University, Bangkok. meanwhile, the infant mortality rate has declined to about one-half the rate of the 1970s. From the mid 1970s to 1994, the number of physicians and nurses per head of population roughly doubled. The lower incidence of rural poverty can be partly attributed to increased farm household income. Between 1970 and 1995196, the average cash income of a farm household increased from about 5,400 baht to about 21,000 baht (deflated by CPI). Up until about 1980, net on-farm income and other income increased; from the 1980s onwards, however, farm cash income deflated by CPI declined, while other income increased very rapidly (Figure 8.14). This suggests that Thai farmers enjoyed a considerable rise in income during the during the commodity boom of the 1970s, but had to rely more on off-farm and non-farm income after it. 348 RURAL DEVE1,OPMENT AND AGRICULTURAL GROWTH Despite significant exports, farm income did not increase in real terms. From the late 1970s onwards, world commodity prices stagnated and even dropped. Because Thai agricultural commodities are delivered from farms to ports for shipment through free market channels, farm prices reflect world market prices. In the mid 1970s, because farmland in Thailand was reaching the limits of its expansion, average farm size began to decrease. Farmers responded to this situation by seeking income sources off the farm rather than intensifying productivity. Among the sources of off-farm income, 'non-farm custom work' and 'salaries' have become increasinglyimportant, while income from 'hired farm work' has decreased in importance. Few opportunities remain for farmers to increase their incomes in rural areas, thus urban areas have become more important to the farm household economy. Figure 8.14 On-farm and off-farmincome trends,1976 prices 1 16,000 lncom (baht;1976 prices) 1 12,000 Net on-farmagriculturalcash incom fromcrops 1 and livestock 1 Cash incom fromall other sources Source: Bank of Thailand, various dates [a]. Monthly Report, Bank of Thailand, Bangkok; Ofice of Agricultural Economics,various dates [b]. Raidai raichai khong kmetrakon [Farmers' Income and Expenses], Office of Agricultural Economics, Bangkok. Rural development policies and non-income poverty alleviation Government rural development policies have played an important role in the alleviation of non-income poverty. In this context, 'rural development' refers to deliberate activities to deliver resources to rural people in order to enhance their welfare. In the late 1950s, the Thai government formed a s~ecialisedagency for rural development and began to create a system for resource distribution (Yingvorapunt 1966). In 1962, the agency became the Community Development Department (CDD) of the Ministry of Interior. Up until the 1970s,Thailand's rural development policy was directed against anti-government and communist movements in rural areas. The first program for rural development involved the creation of a Mobile Development Unit (MDU), to be operated by the army with assistance from the United States, which feared another communist country in Asia. The MDU was deployed to remote areas, where it constructed roads and bridges and provided medical services (Huff 1967; Judd n.d.; Mingmaninakhin 1988). In the 1970s, the threat of communism became more realistic. In 1975, the Kukrit government embarked upon the Tambon Project, injecting 500,000 baht per tambon (sub-district) to enable people to carry out infrastructure projects (Phiphatseritham 1975). About 55 per cent of the total budget was used for road construction, while 26 per cent was spent for water resource development (Nonaka 1983). Just one year before the implementation of the Tambon Project, the CDD began to promote savings groups (Woradilok 1984), or micro- finance organisations formed by local people to pool money for borrovving and lending among themselves. This was the first time rural Thais had taken the initiative to organise themselves and manage their own resources for economic development. After 1976, when the student and peasant movement was brutally crushed by the government, many activists joined the Communist Party of Thailand (CPT). The anti-government movement became stronger and more determined than ever, despite military operations led by the Thanin government. In late 1977, the hardline Thanin government was ousted by the army, led by General Kriangsak, who took a more lenient attitude towards anti-government activists. Kriangsak declared 1979 the Year of the Farmer and planned a rural development program 350 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH for less developed areas. Because Kriangsak resigned soon after the start of the program in 1980, however, this program bore no fruit. The office of head of state was then occupied by General Prem Thinsulanonda, who had formerly been the army commander in the Northeastern region, where he had tried to weaken communist influence through rural development projects. Under the Prem government, the fifth five-year development plan (1982-86) set rural development as its main target, and a new policy framework for rural development called the National Rural Development Program (NRDP) was formed. Coordinated by a national committee headed by the Prime Minister, the NRDP included various government agencies' projects aimed at making activities more effective, and also persuaded rural people to participate in project conception and creation, and the formation of development-oriented organisations. During the period 1982-86, this framework was typically applied to the Poverty Area Program (PAP), in which the ministries of the Interior, Agriculture, Public Health and Education joined to create 34 projects for villages identified as 'poor' in Northeastern, Northern and Southern Thailand. These projects could be placed in three categories: projects that promoted communal activities at a village level; projects to enhance production under poor agricultural conditions; and projects for basic human needs (BHN), mostly for public health care and non-formal education (Shigetomi 2000). Among the positive effects of these projects, according to the Village Database Survey implemented every two years, was to cause the indicators of BHN and infrastructure related to livelihood to rise from 1982 to 1986. Projects designed to improve agricultural production and farm income, however, were not satisfactory (NESDB n.d., 1985, 1991). Infrastructure development has formed the core of all government rural development programs, which include construction of rural feeder roads, creation of rural off-farm employment opportunities, meeting basic human needs, and provision of medical services and primary healthcare. Since the 1980s, participation and organisation by rural people have also become key elements. In the late 1970s and early 1980s, following the success of the savings groups, many other projects to organise local ~ e o p l ewere formed (Shigetomi 1998). A fund to support local organisational activities was established under the NRDP and upgraded to an independent institution for community activities in 2000. Most of these projects could be implemented by transferring resources, including information, through organisational channels from resource providers (such as government and non-government agencies) to local people. Thus, infrastructure construction and the BHN approach have helped raise living standards for the rural population for more than three decades. Disparity between rural and urban areas Despite the reduction of absolute poverty in rural areas, poverty in relative terms persisted and in some cases even increased. Gini coefficients calculated by various scholars show a similar trend in income disparity. Most fi ures indicate that the income gap in Thailand increased up g until the early 1990s, then began to narrow. After the economic crisis, however, the gap seemed to widen again, to the point where the income gap in Thailand is serious compared to that in neighbouring countries. Among developing countries in Asia, Thailand scores second highest in Gini coefficients (Figure 8.15). Moreover, among Asian countries, Thailand alone increased its income gap from 1972 to 1992 (World Bank 1996:15; lkemoto 1991:24-5). The income disparity between urban and rural dwellers is reflected in the Gini coefficient. From 1975 to 1986, the gap between average monthly expenditures in rural and urban areas in Thailand widened (Table 8.10). M e r 1994, however, the width of the gap clearly trended downward. The widenin rural-urban gap can be partly explained by g the income structure of farm households. After the 1980s, non-farm income was becoming a major part of the farm household's income, with farmers increasingly relying on non-farm work for income. Job opportunities in urban areas, however, have not necessarily been stable for immigrants from rural areas. In the 1960s, most jobs were found in the service sector, which absorbed many female migrants as domestic service workers. In manufacturing, most women labourers were hired in the textile mills, which at the time were talung the lead in the sector. Male immigrants, on the other hand, could not easily find jobs in the formal manufacturing sector, and many were hired as construction workers and taxi drivers. From 1960 onwards, the manufacturing sector continued to grow, hiring more people than other sectors. Meanwhile, the metal and machinery sector also showed conspicuous growth. 352 RURAL DEVELOPMENT AND AGRICULTURALGROWTH Figure8.15 Distribution of per capitaGDPand Ginicoefficients amongselected Asian countries 25 L- la, Note: Per capita GDP is from 1998. The Gini coefficients are based on data from various years, indicated in parenthesis for each country. Source: World Bank, various dates. World Development Indicators,World Bank, Washington DC. Available online at http:llwww.worldbank.org/data~onlinedatabasesl onlinedatabases.htm1. Table 8.10 Averagemonthlyexpenditureper householdbyarea (bahtper month) Rural 1,619 3,011 4,129 4,751 5,817 6,919 7,889 Urban 3,151 6,020 7,726 9,447 10,702 12,739 13,446 Urbanlrural 1.95 2.00 1.87 1.99 1.84 1.84 1.70 Source: National Statistical Ofice, various dates. Report of the Howehold and Socio- Economic Survgi Whole Kingdom, National Statistical Office, Bangkok. THAILAND'S RURAL SECTOR AND THE GOVERNMENT 353 Though increasing numbers of rural workers were absorbed into the urban sector, nonetheless job opportunities remained unstable, and about one-half of rural-to-urban migrants returned to rural areas (Table 8.11). This pattern of urban-to-rural return migration was apparent for people above 30 and below ten years old, suggesting that migrants to urban areas tended to return home with children when they became old enough to have them. Their movements can be explained by the characteristics of employment opportunities in large cities, especially Bangkok and the surrounding area. Comparatively stable and high-paying jobs in large factories (producing garments and electrical appliances, for example) were offered to young and mostly female labourers. This employment structure for rural people, which formed in the 1960s and continued to operate until very recently (Shigetomi 1995b), may have prevented a drastlc decrease in the rural population of Thailand.' The uncertainty of urban jobs for migrants from rural areas is related to their level of education. According to the World Bank (1996) in 1988 and 1992, education level determined 34 and 44 per cent of income inequality respectively. There is a clear gap, especially for secondary and higher educational levels, between the urban and rural employed population (Table 8.12), but the gap is declining rapidly because most rural parents provided a secondary education to their children after the 1990s. Since rural people are employed in unstable jobs compared to those who grew up in urban areas, the income gap cannot be reduced without enhancing farm income. The government started an agricultural price intervention program in the 1970s and even tried to control agricultural production after the late 1980s. Today, most agricultural policies, implicitly or explicitly, aim to increase farmer income rather than productivity. Agricultural price policies In its first attempt at price intervention for rice in 1975, the government forced rice mills to buy paddy at a higher-than-market price. Severe criticism of the scheme from consumers, however, forced the government to abolish it (Christensen 1993). Since this failed attempt, the government has taken repeated measures to support the paddy price. These measures fit into four categories (Tongpan 1993) w wl Table 8.11 Migrationpatternsbetween rural and urbanareasby agegroup ('000people) JA Age group Rural to Urban to Rural to Urban to Rural Urban to E urban rural urban rural to urban rural F (a) (b) (a)-(b) (a) (b) (4-6) (a) (b) (4-6) u 5-9 27.7 30.2 -2.5 23.7 34.7 -11.0 28.5 38.3 -9.8 10-14 39.7 23.6 16.1 35.2 25.5 9.7 47.2 32.2 15.0 2 r 15-19 81.2 19.0 62.2 108.0 29.9 78.1 179.8 48.2 131.6 20-24 85.7 24.3 61.4 110.3 54.5 55.8 207.4 141.3 66.1 25-29 40.5 21.5 19.0 60.4 46.0 14.4 112.5 86.6 25.9 9 30-34 25.3 16.6 8.7 32.7 24.4 8.3 64.4 56.6 7.8 35-39 15.8 14.2 1.6 17.7 21.8 4 . 1 33.6 31.9 1.7 5 0 40-44 10.2 11.0 -0.8 11.2 15.4 4 . 2 21.1 24.4 -3.3 iin 45-49 6.3 5.8 0.5 6.8 8.0 -1.2 14.7 16.9 -2.2 50-54 4.4 5.0 -0.6 4.9 6.2 -1.3 10.9 10.7 0.2 s 55-5 9 3.3 3.2 0.1 3.0 5.5 -2.5 6.8 8.2 -1.4 60-64 2.7 2.4 0.3 1.O 2.6 -1.6 4.7 5.6 -0.9 ci 65- 4.3 3.3 1.O 3.7 3.8 -0.1 6.8 8.0 -1.2 E Unknown 0.9 0.3 0.6 - - - - - - Tota1 348.0 180.4 167.6 420.6 278.3 142.3 738.4 508.9 229.5 Note: The total figures for rural to urban migration in the original dataset do not match the sum of the rows. Source: National Statistical Office, various years. Population Census Subject Report No. 2: migration, National Statistical Office, Bangkok. Z the government or its agents purchase rice from farmers, traders and millers at a supported price the government forces traders and millers to stock a certain amount of rice the government or its agents provide loans with preferential conditions to traders and millers with the condition that the latter purchase paddy at a supported price the government or its agents accept a pledge of paddy from farmers at a supported price. These schemes have also been applied to many other commodities. These schemes were integrated and managed not by just one agent, but by several, including the Marketing Organization for Farmers, the Public Warehouse Organization (PWO), the BAAC, and the Bank of Thailand (BO'T). The Marketing Organization for Farmers lies under the Ministry of Agriculture and Cooperatives, the PWO is an agent of the Ministry of Commerce and the BAAC is supervised by the Ministry of Finance. Besides these major agents, the Ministry of the Interior and the National Army also had small schemes to support prices. The various agents did not coordinate with one another; instead, each agent and the Table 8.12 Shareof employed personsbyeducationallevel Agriculture Elementary and kindergarten Secondary or equivalent University Others None Elementary and kindergarten Secondary or equivalent University Others None Source: National Statistical Office, various years. The Labor Force Survey, Bangkok, Round 2 (July-September) of 1975, 1980, National Statistical Office, Bangkok; National Statistical Office, various years. The Labor Force Survey, Bangkok, Round 3 (August) of 1985, 1990, 1995, 1998, National Statistical Office, Bangkok. 356 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH politicians who supervised it made the decisions on what market intervention it should make. Because there were no set criteria for intervention, decisions were generally based on political considerations. During the 1970s and 1980s, urban income increased, and resistance to price supports for farmers among urban dwellers disappeared. At the same time, partly because of a farmers' movement that took place during the 1970s, rural poverty came to be recognised as a problem urgently needing to be tackled. As a result, since the 1980s, the government has paid much more attention to supporting farm-gate prices than to controlling consumer prices, and price intervention to support farmer income continues today. One major policy development that occurred was the establishment of the Integrated Farmers' Aid Fund (IFAF) in 1991. Compared to former policies, the IFAF has certain features that make it unique. It integrates the funding sources that were previously drawn on by various governmental agencies without any coordination. It sets criteria for intervention. Thus, intervention can only be considered when the relevant commodity's market price falls lower than the average of the past three years (excluding exceptional years), while the final decision on whether intervention is to be undertaken must be made by the National Farmers' Aid Committee (NFAC), chaired by a Deputy Prime Minister. It covers all commodities. Since its establishment, the IFAF has come to fund almost all price supports (DOBE 1996). Because the actual decision for intervention is made by the NFAC, political pressure is still an important element (Tongpan 1993). The trigger price for intervention is the average of the past three years, meaning that the rationale behind intervention is to keep the price above that average. This is not a measure to ensure production costs for producers, but a measure to show the government's response to a fall in price. Although the implementation agencies are not yet integrated (Manarungsan 1997), they can now obtain money for intervention more easily than before because they do not have to ask permission from the Bureau of the Budget, the government agency that scrutinises spending, thus they can respond more quickly to price falls and political pressure for price intervention. Production control policies The government began its attempts to diversify agricultural production in the early 1990s. In 1991, it implemented the Agricultural Rehabilitation Plan (ARP), by which farmers chose from a list of recommended commodities and proposed their own production plan to a governmental committee at the local level. When the committee had approved the ~ l a nthe government would supply technical support , and a loan (at 5 per cent interest) through the BAAC (Kharnpha 1993). The ARP remained in force for five years in 44 provinces demarcated as poor areas, benefitin over 220,000 farm households, or about 22 per g cent of all farmers in the target area (DOAE c. 1997). In 1992 and 1993, following the implementation of the ARP, agricultural commodity prices fell considerably, and there was a severe drought in the area along the Chao Phraya River. In 1994, recognising that Thailand could no longer rely on foreign markets and natural resources as it had in past decades (Chulalongkorn University n.d.), the government implemented the Project for Restructuring Agricultural Production Systems (RAPS). This project's implementation was similar to the ARP's, with the following differences. The RAPS covered the entire country. A large budget was established (33.8 billion baht for loans compared with 646 million baht for the ARP) (DOAE 1994, c. 1997). Loan conditions were 5 per cent interest and a 15-year repayment period. In addition to crops and trees, livestock, inland fisheries and mixed farming were also recommended. Certain crops and acreage had to be reduced. This was the first time the government had clearly indicated which crops to discourage. (TDRI 1995) The diversification that occurred in the 1990s was different to earlier waves of diversification in several respects. First, it was the government that prepared the list of recommended commodities; and even though farmers had the right to choose a commodity and make their own production plan, they received very limited information about the listed commodities. Often all the information they received was an oral description of the commodities recommended in their village, the requirements for farmers who chose a certain commodity, and the costs 358 RURAL DEVELOPMENTAND AGRICULTURAL GROWTH and benefits that could be expected for each recommended commodity. The farmers were given only a very short time in which to consider, and no intensive consultations between field officers and farmers were carried out. As a result, the diversification programs were seen to be government led. While the government scheme to provide loans was effective, extension support was not as well organised; because of the attractive loan conditions, however, the number of participants exceeded the target. The government's goal in beginning these projects was to raise farmer income rather than promote export of agricultural commodities, and the RAPS was aimed at appeasing the farmers' unions demonstrating at the Prime Minister's office to demand price supports in 1993. Meanwhile, reduced crop production presented a dilemma: when the acreage of a crop was successfully reduced, its price naturally tended to rise, which meant that the government had no reason to enforce an area reduction. Moreover, since most Thai commodity markets are linked to world markets, their prices can change irrespective of the government program. These factors brought the feasibility of RAPS into question, and in 1997, the BAAC, reluctant to continue providing longterm and low-interest loans, changed the loan conditions. This led to a drastic decrease in the number of participants, an indication that the main incentive for farmers to participate was to secure a loan. Another method used to control production was zoning. Zoning appeared in a 1991 Agricultural Assembly bill that created a special committee for the planning of production, marketing and zone demarcation (Chulalongkorn University 1994). Farmer organisations opposed the bill and proposed that it be replaced by a Farmers' Assembly bill, referrin to an assembly made up only of farmer representatives, as g opposed to the Agricultural Assembly, which would include representatives from private businesses. Because the assembly was to have the authority to formulate agricultural development plans, opposition to it led to serious political conflict. Even the bill proposed by farmer groups, however, tried to regulate production and marketing in order to prevent falling farm prices (Chulalongkorn University 1994). Changing the policy target from production to income When rural poverty became a political issue, many programs initially aimed at raising a ricultural production were transformed into income g enhancement policies. For example, beginning in the 1980s, irrigation construction schemes shifted their focus from large-scale to small and mid-sized systems, with the small projects designed to raise farmer income rather than intensify agricultural productivity (Tongpan 1993; Siamwalla et al. 1993). Later, the primary goal of the agricultural extension policy of 1987-91 was to raise farmer income, with no mention made of intensifying agricultural ~roduction(DOAE 1986). In a review of 100 years of the Ministry of Agricultureand Cooperatives (MOAC), research and extension are recopised as contributors to the economic welfare of farmers (MOAC 1992). Na Chiangmai (1989) describes how agricultural extension policy changed at the end of the 1980s, moving away from its late 1970s emphasis on income distribution. Siamwalla et al. (1990) also argue that the government only increased the budget for research on agricultural commodities came only after farmers had already widely accepted the relevant commodities. According to the history of the MOA, the government placed more importance on raising farmer income than on the market competitiveness of certain commodities, and even regarded projects promoting the use of fertiliser as income subsidies for farmers (Tongpan 1993). According to Siamwalla et al. (1993), a policy more supportive of agriculture was enacted in about 1982. It is doubtful, however, whether this policy or pricing and production restructuring contributed to reductions in rural poverty, especially in relative terms. Because major Thai agricultural commodities depended on world markets, the government's attempts to control domestic prices through intervention in markets and production were largely in vain. New crop diversification and rural differentiation Though the effectiveness of the government's diversification policy is doubtful, diversification is nonetheless occurring, although no new commodity has become a major export. Diversification is important for the farm economy; for example, for the last ten years, cash income from livestock has steadily increased, with large-scale increases in the incidence of livestock production, such as integrated farming of chicken and swine, responsible for much of the additional income. In the crop sector, vegetables, fruits and flowers have increased their production over the past ten years and are gaining a larger share of income, though they are yet to match the income from rice (Table 8.13). 360 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Table 8.13 Shareof crop income by commodity, 1988-99 (totalcrop income=100) Crop Rice Rubber Maize Cassava Sugar cane Pineapple Oil crops Fibre crops Tobacco Vegetables Fruitsltree crops (except rubber) Flowers - - Source: Office of AgriculturalEconomics, 1991. Raiddi-raichai Rhong kasetrakonpipho pluk 2531/32[Farmers' income and expenditure in crop year 19881891, Ofice of Agricultural Economics, Bangkok; Ofice of Agricultural Economics, various dates. Agricultural Statistics of Thailand, Ofice of Agricultural Economics, Bangkok. These new crops are different to more traditional crops in several ways. They require more capital, especially livestock, which requires facilities and inputs. They are labour and technology-intensive. The modern system of raising livestock requires farmers to use time and technology to maintain and improve the wellbeing of their animals. Crops such as vegetables, fruits and flowers, if intended for sale in a market, require much more careful cultivation. Their producers must be able to market them, especially those producers who do not have a purchase contract. Not every farmer can meet these expectations. In 1989 and 2000, the author conducted intensive farm household surveys in Ban Thon, a small village in Northeastern Thailand. The changes undergone by this villa e's farm economy during these 11 years show who has been able to g adapt to the new agricultural economy. Thon is located in Khon Kaen province, about 20 kilometres from Khon Kaen city (Shigetomi 1998). When the author surveyed Thon in 1989, its only marketable commodities were rice, cassava and soya. All farmers had livestock, but they raised it in a traditional way. Non-farm income was also important, with most farmers finding farm and non- farm employment, and many young people leaving the village to work in Bangkok and even overseas. Severalfarmers had begun to raise chickens under contract to big agribusiness, and non-traditional livestock such as dairy cattle. When the author visited Thon again in 2000, many changes had occurred, the most conspicuous relating to agricultural diversification. Among the village's 347 households, 42 had large-scale livestock enterprises, including 34 chicken farms, four pig farms and four dairy cattle farms. The household that had started a dairy farm about ten years ago had expanded its size considerably. One villager now cultivates flowers, others grow mushrooms. Among the 145 sample households surveyed in 2000, four were engaged in mixed farming (rai na suan phasom), including production such as rice, fruits, fish and fowl. In terms of land ownership, because a mortgage is needed to secure a loan from a financial institution, most livestock producers belong to the village's upper class. In some cases, landowners have mortgaged their own land for their children's farms. While the government claims that mixed farmin is suitable for small-scale farmers, in fact it requires a g rather large irrigated area. One of the smallest farms is 1.44 ha, while the average area of irrigated land in Thon is 0.88 ha. Most of these farm households have at least one male family member who is a full-time farmer; many, however, no longer have enough labour even for rice cultivation. In 1989, the author often found that if one of a group of related households lacked family labour, the related households jointly cultivated land. When the author visited Thon again in 2000, such cooperation seemed to be diminishing, since most households could not find relatives with enough labour for farming, and many households had taken to lending out their paddy fields to non-related households that had enough family labour. This suggests a polarisation among the village's farm households, with the majority group reducing farm productivity and the minority still struggling to survive by farming. CONCLUSION The success of Thai agricultural exports over the past four decades may be attributed to the swift and adequate responses to world market opportunities of not only agibusinesses but also small-scale farmers. 362 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH The Thai government did not act as a trader or producer in isolation, and though it did sometimes intervene in markets to support farm prices, its impact was very limited. This does not mean, however, that it did nothing to foster agricultural development in the country. Table 8.14 summarises the Thai government's role in relation to the major export commodities, as discussed above. The government's positive contributions include provision of services that. cannot be procured through market transactions (seed improvement, irrigation) * supportive actions lying within the jurisdiction of government (investment promotion, tax exemptionlreduction, negotiation with foreign governments to reduce import duties) provision of subsidies to improve productivity (program to replant rubber) authorisation of quality standards and inspection (maize, tapioca) protection of infant industries (promoting industrial use of kenaf as a raw material) subsidisation of exports through compulsory redistribution of economic surplus (sugar, tapioca). From certain perspectives, the last two government actions in this list may be assessed negatively. The positive result produced by the second- last was unexpected, since the government did not intend to promote exports of kenaf products, while the last may have had a negative impact on other sectors. We should not forget, however, the fortunate conditions that have allowed Thailand to achieve such successes. These include the availability of significant amounts of land for the expansion of maize, tapioca, and kenaf cultivationin response to market expansion Thailand's relative proximity to Japan, the world's largest importer of agricultural commodities Thai food habits, which have made it possible for tapioca products to dominate the European market. On the other hand, Thailand has failed to reduce the rural-urban income gap. Even though their economic welfare did improve with the expansion of agricultural production and increases in employment opportunities in the non-farm sector, Thai farmers still regard themselves as the country's poor. Urban dwellers have tended to share this perception Table 8.14 Factorsthatpromotedagriculturalexportsof major commodities Government policies Other factors Rice Seed improvement and dissemination Entrepreneurship of exporters Irrigation to increase total production Policy mistake by United States Reduction of rice premium in early 1980s Rubber Rubber replanting program (subsidy) Private sectors' efforts to improve quality Japanese tire makers' preference for Thai rubber Kenaf Establishing gunnysack factories and No other producers of jute-like protecting the domestic industry products before the market using kenaf opportunity appeared Land availability for expanding plantation Maize Negotiating with Japanese traders and Land availability for expanding monitoring performance of Thai plantation traders during the contract Close to Japanese market Seed improvement and dissemination Defining quality standards Tapioca Setting up and managing bonus quota No other country that can steadily system provide the export market (Thai Defining quality standard people do not eat tapioca as their staple food) Preferable tax conditions by GATT agreement Sugar Setting up and managing cross-subsidy Active investment to expand system facilities by private sector Close to Japanese market Frozen Investment promotion Active investment by private sector chicken Negotiating with Japan to reduce Close to Japanese market import duty Cannecl Investment promotion Active investment by private sector pineapple Sufficient land available to provide enough raw material to processors Source: Prepared by the author. 364 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH of farmers, making rural poverty a key political issue in Thailand since the mid 1970s. Several factors have contributed to the continuing existence of the urban-rural income gap. Because the major Thai agricultural commodities are linked to world markets, slumps in the latter have directly affectedThai farm income. Thai agriculture's reliance on market mechanisms has made it diEcult for the government to control farm prices. Budget constraints have made it impossible for the government to guarantee farmer income, and the government's inability to control markets has made production control ineffective in the long term. Low education levels have made it diff~cultfor rural migrants to find stable, well-paying jobs. Under these circumstances, what kinds of policies would narrow the urban-rural income gap? To raise their income, farmers could introduce new commodities. While it is unreasonable to expect that new boom crops such as maize and cassava will appear again, farmers may still find products for which demand is not huge but which yield a good profit. In this context, farmers should not just follow their neighbows; they should take an entrepreneurial approach, introducing commodities new to their areas, and develop the general intelligence needed to find and exploit market opportunities. They also need be shown how to master new technologies and management practices, and should be given access to capital for new investments. In these respects, the government can help by providing education, research on new technologies and loans, but not in picking new products for dissemination,since this might disturb markets for existing producers and lead potential new growers to regard the products as unsuitable. Thailand's agribusinesses and farmers are capable of finding and adapting to new market opportunities themselves, thus the government should limit its hnction to the provision of such services as irrigation, roads, research and education. Only a minority of farmers, however, may be in a position to use such strategies. Most farm households rely on non-farm income for their economic well-being, thus it is important to increase educational opportunities in rural areas so as to reduce rural residents' disadvantage in non-farm labour markets, even though this may move some rural residents to leave the farming sector permanently. Such a polarisation seems inevitable; in fact in some areas it is already occurring. ACKNOWLEDGMENTS The author is gateful to Dr. Sompop Manarungsan for advice on survey methodology and textual matters, Mr. Bhumisuk Khananurak for data collection in Thailand, and Mr. Yosuke Noda for data processing at the Institute of Developing Economies. NOTES NationalEconomic and Socii Development Board (NESDB)databaseon nationalincome. Interview with a tapioca exporter in Bangkok, September 1988. This agreement allowed coverage of 10 per cent, so Thailand could sell at most 5.5 million tonnes to the EC. * Since raw sugar is the most common Thai sugar export, this analysisconcentrates on its market. For example, in 1963 the sugar price soared becauseof decreasedproduction in Thailand, Cuba and Europe,and theThaigovernment allowedsugar factotiesto increase production. In 1965, however, the price fell sharply, and the government was forced to pay export subsidies in order to dispose of the surplus in the external market (Jessadachatr 1977). For example, while the average NewYork spot market price in 1986187 was about 6.0- 6.7 cents per pound, no country could actually produce sugar at a cost below this price (Lichit 1994). ' Interview at the Department of Agriculture, Ministry of Agriculture and Cooperatives, September 1999. Siamwalla et al. (1990) argue that Thailand is unique because of its high share of agricultural labour relative to countries with similar per capita GDP, but also that this may be largely a result of statistical bias in the population census. REFERENCES Ando, K., 1979. Kumpawapi 13 nen no ayumi (Kumpawapi sugar: its 13-year progress), unpublished paper. Asami, K., 1976. Shiyo sangyo hatten goju nensbi: Shiryo hozei kojo seido sosetsu goju shuunen wo kinen shite (50 Years' Development of Feed Industry: On the Occasion of 50 Years' Anniversary of Bonded Factory System), Tokyo. 366 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Ashakul, Charuma, 1985. Income Distribution in Thailand: an analysis with policy implications, .University Microfilms International, Ann Arbor. Bangkok Post, 12 January 1987. 'Industry report: pineapple canning has more growth potential'. Bank of Thailand, 1980. Quarterly Bulletin, 25(4), Bank of Thailand, Bangkok. -, 1985. Quarterly Bulletin, 28(4), Bank of Thailand, Bangkok. -, 1990. Quarterly Bulletin, 35(3), Bank of Thailand, Bangkok. -, 2000. Monthly Report, September, Bank of Thailand, Bangkok. -, various dates [a]. Monthly Report, Bank of Thailand, Bangkok. -, various dates [b]. Quarterly Bulletin, Bank of Thailand, Bangkok. Board of Investment, 1978. Kan songsoem kan long thun [Investment Promotion], Board of Investment, Bangkok. ----, 1989. 35pi bi o ai: &o yang mankhongpaikap kan songsoem [35 Years of BOI: stable progress with promotion], supplement in Than Sethakit, June. Business,in Thailand, 1976. 'Pineapple canning: a golden future', June:53-8. Business Review, 1984. 'Canned pineapple: starting a strong comeback', August:69-72. Chakatpanyawai, Wilaiphon, 1992. 'Sapparot krapong: naeo nom kan song ok khayai tua' [Canned pineapple: increasing export trends], Sarup khao thurakit, 23(17):28-36. Chantharanimi, Phairot, 1988. 'BOI pisat rai thi sing yu nai san phra phum' [BOI: ghost in the sacred shrine], Phu chat kan, 5(57):3048. Charoenloet, Chira, 1971. The Evolution of Thailand; Economy, Thai Watana Panich, Bangkok. Chikusan shinko jigyodan (Livestock Industries Corporation), 1980. Yunyu Shokkei no ryutsujittai chosa [Study on the Market of Imported Chicken Meat], Livestock Industries Corporation, Tokyo. Chintayarangsan, Rachain, 1983. Rubber and Kenaf in Thailand, National Institute of Development Administration, Bangkok. Christensen, S.R., 1993. Coalition and collective choice: the politics of institutional change in Thai agriculture, PhD dissertation, University of Wisconsin, Madison. Chuchat, Chaiyong, 1960. Nayobai kan kaset wa duai kan wikhro watthuprasong withi kan he kan ddmnoen nayobai kaset khongprathet thai he tangprathet [Agricultural Policy: analysis of targets, methods and policy implementation in Thailand and foreign countries], Prae Bhittaya, Bangkok. Chulalongkorn University, Faculty of Economics, Sun wichai sethakit, 1991. Final Report on NAIC Development: a new dimension in Thai- Japanese economic partnership, Institute of Asian Studies and Faculty of Economics, Ch~lalon~kornUniversity, Bangkok. -, 1994. 'Khrong kan phatthana santiwithi nai sangkhom thai. Rai ngan kan sammana satharana ruang sapha kan kaset' [Report of Seminar on Agricultural Assembly], Chulalongkorn University, Bangkok. -1997.Brongkanwichairuangkanphoemkhitkhwamsamatnai , kan khaeng khan khong utsahakam thai nai sethakit lok: Rai ngan kan suksa khrong kan yoi thi 4: Kan phoem khit khwam samat nai kan khaeng khan khong utsahakam prae mp ahan [Research Project on Competitiveness of Thai Industry in the World Economy: Report No. 4: the food processing industry], Ch~lalon~kornUniversity, Bangkok. --,Samnakborikanwichakan.n.d.,Raingankanpramoenphon khrong kan prap khrong sang lae rabop kan phalit kan kaset [Assesment Report of the Project of Restructuring Agricultural Production System], Bureau of Budget, Bangkok. Coordinated Program for Corn Development in Thailand (CPCDT), c. 1962. 1962 Progress Report, Coordinated Program for Corn Development in Thailand, Bangkok. Cropper, M., Griffiths, C. and Mani, M., 1997. Roads, Popuhtion Pressures, and Defovestation in Thailand, 197689, Policy Research Working Paper IVo. 1726, World Bank, Washington, DC. Department of Farmers Service,Taiwan, 1982. Agricultural TrddeStatistics of Taiwan, R.0.C , Department of Farmers Service, Taiwan, Taipei. Dethloff, H.C., 1988. A History of the American Rice Indusovy, 1685- 1985, Texas A&M University Press, College Station. Dixon, C., 1999. The Thai Economy: uneven development and internationalisation, Routledge, London. 368 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Department of Agricultural Extension (DOAE), 1986. Phaen phatthana kan songsoem kan kaset pho.so.2530-2534 [Development Plan for Agricultural Extension 1987-19911, Department of Agricultural Extension, Bangkok. -, 1994. Khrong kan prap khrong sang Iae rabop kan phalit kan kaset pi 2537-2539 [Project of Restructuring the Agricultural Production System, 1994961, Department of Agricultural Extension, Bangkok. -, c. 1997. Rai ngan phon kan damnoen ngan khrong kan sanap sanun phaen phalit khong kasetrakon phai tai phaen @n@ kan kaset [Report on the Result of Implementing the Project for Promoting the Farmers' Production Plan under the Agricultural Rehabilitation Plan], Department of Agricultural Extension, Bangkok. Department of Business Economics, 1996. Report to World Trade Organization, Committee of Agriculture, Bangkok (unpublished). Department of Factory, 1959. Rai ngan kitchakam pracham pi [Annual Report], Krom rong ngan utsahakam, Bangkok. Department of Land, 1992. Khrong kan phatthana krom thidin he reng rat kan ok chanot thidin thua prathet 2535 [Project for Developing Department of Land and Accelerating Issuing Land Title in Entire Nation in 19921, Department of Land, Bangkok. --,1976.Rainganprachampi2518[I975AnnualReport], Department of Land, Bangkok. Dowling, J.M. Jr. and Soo, D., 1983. Income Distribution and Economic Growth in DeuehpingAsianCountries,Asian Development Bank, Manila. Food and Agriculture Organization (FAO) of the United Nations, n.d. FAOSTAT, FA0 online database, http://apps.fao.org/default.htm. Gaimusho (Ministryof Foreign Affairs,Japan), 1960. Taikoku tomorokoshi ... j40: Sono seisan to yzwhutu ni tsuite [The Situation of Thai Maize: its production and export], Ministry of Foreign Affairs, Tokyo. Hayakawa, O., 1986. 'Nihon chikusan to shiryo shijo no tenkai katei' [The development of the Japanese livestock and feed industry], in K. Yoshida (ed.), Chikusan butsu no shohi to yutsu kiko, Nozan Gyoson Bunka Kyokai, Tokyo:204-37. Hayarni, Y., Adriano, L.S., Angnes, M.A. and Quisumbing, R., 1988. Agribusinessand Agrarian Refm: a viewfiom the banana andpineapple phntations, University of Philippines, Los Banos. THAILAND'S RURAL SECTOR AND THE GOVERNMENT 369 Huff, L.H., 1967. 'The Thai mobile development unit program', in l? Kunstadler (ed.), Southeast Asian Tribes, Minorities, and Nations, Princeton University Press, Princeton:425-86. Institute of Agribusiness Development and Policy, c. 1996. Green Pages: Philippine agribusiness and food marketfict book, 1995-96, Institute of Agribusiness Development and Policy, Center for Research and Communication, Manila. Ikemoto, Y., 1991. Income Distribution in Thailand: its changes, causes and structure, Institute of Developing Economies, Tokyo. Institute of Developing Economies, n.d. World Trade Database (AIDXT), Institute of Developing Economies, Japan External Trade Organization, Chiba. International Rubber Study Group, 1974. World Rubber Statistics Handbook 1946-59, vol. 1, International Rubber Study Group, London. --,1986.WorldRubberStatisticsHandbook 1960-85, vol. 3, International Rubber Study Group, London. -, various issues. Rubber Statistical Bulletin, International Rubber Study Group, London. International Rice Research Institute, various issues. World Rice Statistics, International Rice Research Institute, Los Banos. Ithijarukul, Suchada (ed.), n.d. Manufacturing in Thailand, Business Information and Research Ltd, Bangkok. Japan External Trade Organization (JETRO), 1979. Tai no meizu [Thai Maize], Japan External Trade Organization, Tokyo. --,1984. Tdinoburoira[ThaiBroiler],JapanExternalTrade Organization, Tokyo. Jessadachatr, Pitsanes, 1977. A history of sugar policies in Thailand: 1937-75, MA thesis, Faculty of Economics, Thammasat University, Bangkok. Judd, L.C., n.d. In Perspective: trends in rural development policy and programs in Thailand, 1947-87, Research Report Series No. 41, Phyap University Centre for Research and Development, Chiang Mai. Jumbala, Prudhisan, 1987. 'Interest and pressure groups', in Somsakdi Xuto (ed.), Government and Politics of Thailand, Oxford University Press, London:110-67 370 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Khampha, Ladawan, 1993. 'Phaen fun fu kan kaset: thang ok khong kaset rai yoi' [Agricultural rehabilitation plan: the way out for small farmers], Warasan sethakit lae sangkhom, 30(4):22-6. Kiatikamchai, Somchai, 1992. 'Sapparot: phut kaset utsahakam thi samkhan khong thai' [Pineapple: the important crop of Thailand], in Thanakhan Krungthai (ed.), Rai ngan sethakit, January:52-7. Kiatpanapikul, Chirapan, 1978. Panha kan phalit lae kan chamnai sapparot krapong khong kitchakan thi mi kan Pluk saparot eng [Problems in the production and distribution of canned pineapple enterprises with own plantations], MA thesis, Facul* of Accountancy, Chulalongkorn University, Bangkok. Kirakul, Krisada, 1975. Sugar cane procurement in the eastern and western regions of Thailand, MA thesis, Faculty of Economics, Thammasat University, Bangkok. Krongkaew, Mehdi, c. 1981. Current interest in and research on poverty and income distribution in Thailand, unpublished paper. -, 1989. Poverty and Income Distribution in Thailand, revised version of paper presented at the International Conference on Thai Studies, the Australian National University, Canberra, 3-6 July 1987. -, 1996. Thaihnd: poverty assessment uppdate, Thammasat University, Bangkok. Laovakul, Duangmanee, 1993. Decomposition of Income Inequality and Poverty in Thailand, Thammasat University, Bangkok. Licht, F.O., 1994. 'Philippines Sugar 2000: an industry finding its feet', in EO. Lichtj International Sugar and Sweetener Report, 126:749-54. Manarungsan, Sompop, 1976. The history of fertilizer policies in Thailand: an economic study, 1960-75, MA thesis, Faculty of Economics, Thammasat University, Bangkok. -, 1997. Rai ngan wichai phatthana kan khong phak kaset he krathop to chao rai chao na thai nai chuang hng samai songkhram lok khrong thi 2 - pho.so.2536 [Study Report on the Development of the Agricultural Sector and its Impact on Thai Farmers from the end of World War I1 to 19931, Ch~lalon~kornUniversity, Bangkok. --andKaeothep,Kanoksak,1987.Rainganphonkanwichai utsahakam namtan nai prathet thai [Research Report on the Thai Sugar Industry], Institute of Asian Studies, Chulalongkorn University, Bangkok. Mingmaninakhin, Wanrak, 1988. fin phatthana chonnabot thai [Rural Thai Development], Samnakphim mahawithayalai tharnmasat, Bangkok. Ministry of Agriculture, 1957. Agriculture in Thailand, Ministry of Agriculture, Bangkok. Ministry of Agriculture and Cooperatives (MOAC), 1992. Khrop rop 100 pi Krasuang kaset lae sahakon [I00 years of the Ministry of Agriculture and Co-operatives], Ministry of Agriculture and Co- operatives, Bangkok. Ministry of Commerce, various dates. k d e Statistics and Economic Indicators of Thailand, Ministry of Commerce, Bangkok. Moerman, M., 1968. Agricultural Change and Peasant Choice in a Thai fillage, University of California Press, Berkeley and Los Angeles. Ministry of Finance, various dates. Foreign Trdde Statistics of Thailand, Ministry of Finance, Bangkok. Ministry of Industry, c. 1985. h i nganphawa sethakit utsahakam choepho praphetpi 2522.utsahakam sapparot krapong [Industry Situation Report, 1984: pineapple canning industry], Ministry of Industry, Bangkok. Miyazaki, H., 1986. 'Chikusan integureshon to shijo saihensei' [Livestock integration and restructuring of market], in K. Yoshida et al. (eds), Chikusan butsu no shohi to yutsu kiko, Nozan gyoson bunka kyokai, Tokyo:238-67. Mizoguchi, F., 1962. Ajia no shiryo sangyo [The Feed Industry in Asia], Institute of Developing Economies, Tokyo. Mundlak, Y., 1997. Land Expansion, Land Augmentation, and Land Saving, University of Wisconsin-Madison, Madison. Murashima, E., 1980. '70 nendai ni okeru tai nomin undo no tenkai: tai nomin no seiji kanyo to seiji kozo' [The Thai farmers' movement in the 1970s: their political participation and political structure], Ajia Keimi, 21(2):2-31. Muscat, R.J., 1990. Thaikznd and the United States: development, security, and foreign aid, Columbia University Press, New York. Na Chiangmai, Chartchai, 1989. Krom songsoem kan kaset [The Department of Agricultural Extension], Thailand Development Research Institute, Bangkok. National Economic and Social Development Board (NESDB), 1978. 'Kasetrakam phaen mai khao ko kho: khwam samret khong nak kaset thai' [New agricultural system, RD varieties: a success of Thai agronomists], Warasan Sethakit lae Sangkhom, 15(3):6-14. 372 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH -, 1979. 'Kasetrakarn phaen mai: Khao phot suwan 1' [New agricultural system: maize Suwan I], Warasan Sethakit h e Sangkhom, 16(3):5-14. -, 1985. Rai ngan kan suksa khwam to nuang khong khrong kan tam phaen phatthana chonnabot phun thi yakchon [A Study Report on the Sustainability of Projects under Poverty Area Program], Kasetsart University, Bangkok. -, 1991. Rai ngan kan suksa sathana phap chonnabot thai pho.so.2534 [Report on the Situation of Rural Thailand in 19911, National Economic and Social Development Board, Bangkok. -, n.d. Phon kan phatthana chonnabot phun thi yakchon nai chuang 3 pi tam nayobai khong rathaban [The Results of Three Years Implementation of the Poverty Area Program], National Economic and Social Development Board, Bangkok. National Statistical Office, 1966. Household Expenditure Survey 1963, National Statistical Office, Bangkok. -, 1968. Statistical Yearbook Thailand 1966, National Statistical Office, Bangkok. -, various years. Population Census Subject Report No. 2: miption, National Statistical Office, Bangkok. -, various years. The Labor Force Survey, Bangkok, Round 2 (July- September) of 1975, 1980, National Statistical Office, Bangkok. -, various years. The Labor Force Survey, Bangkok, Round 3 (August) of 1985, 1990, 1995, 1998, National Statistical Office, Bangkok. -, 1994. Reportof the 1992 Household Socio-EconomicSurvgl, National Statistical Ofice, Bangkok. ----, various dates. Report of the Household and Socio-Economic Survey, Whole Kingdom, National Statistical Ofice, Bangkok. Nihon shokucho kyokai (Japan Chicken Association), 1989. Showa 63 nendo yunyu keiniku ryutsu doko chosajigyo: Yunyu keiniku juyo doko chosa hokokwho [Report on the Market Situation of Chicken Meat in Financial Year 1988: the trend of imported chicken meat], Japan Chicken Association, Tokyo. Nikkei Sangyo Shinbun, 1986. 'Tai san honenashi keiniku nihon muke ~ushutukocho, kanzei sagari endaka mo mikata' [Upside of Thai boneless chicken exports to Japan: import duty reduction and high yen rate help a lot], 13 June. Nititham, Thanon, 1978. Prawat 20 pi khong samakhom pho kha khao phot lae phut phan thai [2O years of Thai Maize and Produce Traders Association], Warasan samakhom pho kha khao phot lae phut phan thai, 2(2):7-25. Nonaka, K., 1983. 'Kaisetsu' [Explanatory], in Nimit Phumthawon, Noson kaihatsu tenmatsuki, Keiso shobo, Tokyo. Office of Agricultural Economics, 1979. Land Utilization of Thailund, 1950/51-1977/78, Office of Agricultural Economics, Bangkok. -, 1980. Phawa nishin khong kaetrakon pi kan pho pluk 2521/2522 [Situation of Farmers' Debt, 19781791, Office of Agricultural Economics, Bangkok. -, 1983. Phawa nkhin khong kaetrakon pi kan pho pluk 25232524 [Situation of Farmers' Debt, 198018 11, Office of Agricultural Economics, Bangkok. -, 1988. Nisin kasemkon pi pho pluk 2529/30 [Farmers' Debt, 19861 871, Office of Agricultural Economics, Bangkok. -, 1991a. Nisin kasetrakon pi pho pluk 2531/32 [Farmers' Debt, 19881891, Office of Agricultural Economics, Bangkok. --,1991b.Raidai-raichaikhongkasetrakonpiphopluk2531/32 [Farmers' Income and Expenditure in Crop Year 19881891, Office of Agricultural Economics, Bangkok. -, 1991c. fin chai thidin phuea kan kaset khong prathet thai pi 2534 [Land Utilization for Agriculture in Thailand, 19911, Office of Agricultural Economics, Bangkok. --variousdates.AgriculturalStatisticsofThailand,Officeof Agricultural Economics, Bangkok. -, various dates. Raidai raichai khong kasetrakon (Farmers' Income and Expenses), Office of Agricultural Economics, Bangkok. -, various dates. Thailund Foreign Apicultural W e Statistics, Office of Agricultural Economics, Bangkok. Office of the Under-Secretary of State (OUSS), Ministry of Agriculture, 1955. Thailand Economic Farm Survey, 1953, Office of the Under- secretary of State, Ministry of Agriculture, Bangkok. Panpiemras, Kosit, 1978. Development Planning in Thailand: a practitioner?viewpoint,United Nations Asian and Pacific Development Institute, Bangkok. 374 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Phanthurat, Wina, 1991. 'Nayobai phoem phon phalit kan kaset: Khwam champen thi tong thop thuan' [Policy for increasing agricultural production: a review is necessary], Warasan sethakit h e sangkhom, 28(4):15-23. Phasuk, Siri, 1985. Klum kasetrakon kap phatthana chonnabot: Korani klum kasetrakon changwat surin [Farmer Groups and Rural Development: the case of Surin Province's farmer groups], Chulalongkorn University Social Research Institute, Bangkok. Phetprasert, Narong, n.d. Rai ngan phon kan suksa nayobai rakha lae kan talat sinkha kaset lem thi I ton thi I: Botbat khong 0.%.KO. lae 0.Kho.So. nai talat khao [Report on Price and Market Policy on Agricultural Commodities, Volume 1, Section 1: the role of the MOF and PWO in the rice market], National Economic and Social Development Board, Bangkok. Phiphatseritham, Krirkkiat, 1975. Wikhro khrong kan phan ngoen [Analysis on Fund Return Project for Villages], Thammasat University, Bangkok. Phongpaichit, Pasuk and Baker, C., 1995. Thailand.. economy andpolitics, Oxford University Press, New York. Picciotto, R. and Kopp, H.E., 1994. Impact Evaluation Report on Thailand Second Tree Crops Project (Ln. 2078-TH) and Third Rubber Replanting Project (Ln.2691-TH), World Bank, Washington, DC. Prachachat'Thurakit, 1 February 1989. 'Kai sot chae khaeng thuk huai khrang yai: Yipun lot phasi nam khao lue 12%' (Frozen chicken wins the big prize: Japan reduces import duty to 12%). Prichametta, Araya, 1996. Rai ngan chabap sombun ruang lu thang lae okat kan song ok h e phon krathop chak kan mi khet kan kha seri asian (Samrap utsahakam pui khemi) [Final report on the way out, opportunities and impact of AFTA: the chemical fertilizer industry], Thailand Development Research Institute, Bangkok. Puaphonsakon, Niphon, 1980. Kan talat lae rakha kai krathong kap anakhot khongphu liang isara rai yoi [Market and Price of Broiler and the Future of Independent Raisers], National Economic and Social Development Board, Bangkok. Rakthammakit, Somchai, 1990. 'Phaen kan phalit lae kan talat sapparot krapong b o n g thai' [Plan for the production and marketing of Thai canned pineapple], Warasan sethkit kan phanit, 21(220):8-12. Ramsay, A., 1987. 'The political economy of sugar in Thailand', Pacifc Affairs, 60(2):248-70. Royal Forestry Department, 1996. Forestry Statistics of Thailand 1996, Royal Forestry Department, Bangkok. --,1993.ForestryStatisticsofThailand1993,RoyalForestry Department, Bangkok. --,1988.ForestryStatisticsofThailand1988,RoyalForestry Department, Bangkok. Ruam Prachachat, 8 April 1987. '10 pi bon sen thang an kharu khara khong mansampalang thai' [Ten years on the rough road of Thai tapioca]. Saito,T., 1985. 'Nihon-ajia shokoku no nosanbutsu boeki mondai' [Trade issues between Asia and Japan], in Mitsugu Kamiya and Tohiko Yuize (eds), Nogyo hogo to nosanbutsu boeki mondai, Norin tokei kyokai, Tokyo:237-93. Samakhom (Samakhom utsahakarn tho krasop thai), 1975. Utsahakam po hephliqhan po naiprathet thai [Kenaf and Kenaf Products Industry in Thailand] (Phim pen anuson nai ngan phra racha phloen sop Nai Phongsawat Suriyothai na meru Wat Thathu Thong wan thi 15 phrutsaka yon 25 18), Bangkok. Samohantharak, Wani, 1988. 'Botbat khong rat kap kan phatthana kan kaset' [The role of government and the development of agriculture], Warasan sethakit lae sangkhom 25(2):40-6. Setboonsarng, Suthad, 1990. Biotechnology and Developing Country Agriculture: maize in Thaihnd, OECD Development Centre, Paris. -, Wattanautchariya, Saran and Phutigorn, Banlu, 1991. The Structure and Performance of the Seed Industry in Thailand, Thailand Development Research Institute, Bangkok. Shigetomi, S., 1995a. 'The transmission of information in the transacting of primary products: the case of quality improvement in Thailand's natural rubber production', The Developing Economies, 33(2):203- 21. -, 199513. 'Tohoku tai ni okeru shugyo kozo no tenkai: ichi noson kara mita tai sengo keizai shi' [Change of labour market structure of a village in Northeastern Thailand after World War 111, in K. Mizuno (ed.), Tonan ajia noson no shugyo kozo, Institute of Developing Economies, Tokyo:163-210. 376 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH --,1998.CooperationandCommunityinRuralThailand:an organizational analysis of participatory rural development, Institute of Developing Economies, Tokyo. -, 2000. 'Noson kaihatsu seisaku: Henkaku ni okeru seido to kojin' [Rural development policy: institutions and individuals in the process of reform], in A. Suehiro and S. Higashi (eds), Tdi no keizai seisaku: Seido, soshiki, akuta, Institute of Developing Economies, Chiba:215- 58 Siarnwalla, Ammar, 1975. 'A history of rice price policies in Thailand', in Prateep Sondysuvan (ed.), Finance, M e and Economic Development in Thailand, Sompong Press, Bangkok:141-65. -, 1999. 'Myths, demons and the future of Thai agriculture', in Khrop rop 60 pi Achan Ammar, Thailand Development Research Institute, Bangkok:237-57. Siamwalla, Ammar; Patamasiriwat, Direk and Setboonsarng, Suthad, 1987. Productivity and competitiveness in Thai agriculture, Paper presented at the Year-End Conference on Efficiency and Competitiveness of the Thai Economy, Thailand Development Research Institute, Bangkok, 28-29 November. Siamwalla, Ammar; Pinthong, Chirmsak; I'oapongsakorn, Nipon; Satsanguan, Ploenpit; Nettayarak, Prayong; Mimgmaneenakin, Wanrak and Tubpun, Yuavares, 1993. 'The Thai rural credit system and elements of a theory: public subsidies, private information and segmented markets', in K. Hoff, A. Braverman and J. Stiglitz (eds), The Economicsof RuralOrganization: theory,practice, andpoliy, Oxford University Press, New York:154-85. Siamwalla, Ammar and Setboonsarng, Suthad, 1989. W e , Exchange Rate, and Agricultural Pricing Policies in Thailand, World Bank, Washington, DC. -, 1991. 'Thailand', in A.O. Krueger, M. Schiff and A. Valdes (eds), The Political Economy ofAgriculturalPricing Poliy, vol. 2, Asia, Johns Hopkins University Press, Baltimore and London:236-80. Siamwalla, Ammar; Setboonsarng, Suthad and Patamasiriwat, Direk, 1990. Thai Agriculture: resources, institutions and policies, Thailand Development Research Institute, Bangkok. -, 1992. 'The response of Thai agriculture to the world economy', in W.T. Coyle, D. Hayes and H. Yamauchi (eds), Agriculture and Trdde in the Pacific: toward the twenty-jrst century, Westview Press, Boulder:149-74. -, 1993. 'Agriculture', in P.G. Warr (ed.), The Thai Economy in Transition, Cambridge University Press, Cambridge:81-117. Siriprachai, Somboon, 1987. 'Phatthana kan kan song ok mansampalang khong thai tang tae lang songkhram khrng thi 2 (2489-2525)' [Development of Thailand's tapioca exports after the Second World War (1946-89)], Warasan sethasat thammasat, (5)3:5-80. -, 1998. Control and Rent-Seeking: the role of the state in the Thai cassava industy, Lund University Press, Lund. Sriswasdilek, Jerachone, 1973. The yield performance and economic benefits of the high-yield rice varieties in Don Chedi, Suphan Buri, Thailand, MA thesis, College of Agriculture, University of the Philippines, Los Banos. Suehiro, A., 1987. 'Tai ni okeru aguri bijinsu no tenkai: shiryo buroira sangyo no 6 dai gurupu' [The development of agribusiness in Thailand: the six groups in feed and broiler industry], in T. Takigawa (ed.) Tonan ajia no nogyo gijutsu henkaku to noson shakai, Institute of Developing Economies, Tokyo:275-321. -, 1989. Capital Accumulation in Thailand, 1855-1985, Centre for East Asian Cultural Studies, Tokyo. -andYasuda,0.(eds),1986.Tainokogyoka:NAIChenochosen [The Industrialisation of Thailand: a challenge for NAIC], Ajia Keizai Kenkyu Sho, Tokyo. Taiwan Statistics Office, 1994. Agricultural Trade Statistics of Republic of China 1993, Council of Agriculture, Executive Yuan, Taipei. -, 1986. Agricultural Trade Statistics of Republic of China 1985, Council of Agriculture, Executive Yuan, Taipei. Taiwan Canners Association, 1972. Taiwan Exports of Canned Food 1972, Taiwan Canners Association, Taipei. Tasaka, T., 1991. Nettairin hakai to hinkonka no keizdigaku: tai shihonshugi no chiiki mondzi [Economics of Deforestation and Poverty: regional problems of capital penetration in Thailand], Ochanomizu shobo, Tokyo. 378 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Thailand Development Research Institute (TDRI), 1987. Distribution Infrastructure of Four Food Product Exports of Thailand tapioca,fiozen chicken, seafood and longan, Thailand Development Research Institute, Bangkok. --,1988. DynamicsofThaiAgriculture, 1961-85, Thailand Development Research Institute, Bangkok. -, 1995. Agricultural Diversijcation/ Restructuring of Agricultural Production Systems in Thailand, Thailand Development Research Institute, Bangkok. Thisyamondol, Pantum, Arromdee, Virach and Long, M.E, 1965. AgriculturalCredit in ThaiZund theory,dzta, policy, Kasetsart University, Bangkok. Tongpan, Sopin, 1970. Thai Corn Export Problems, Kasetsart University, Bangkok -, 1993. Nayobai kaset thai [Phim khrang thi 2, prap prung mail (Thai Agricultural Policy (2nd and revised edition)), Rong phim loetchai kan phim, Bangkok. UNCTADIGATT, 1977. Cassava:export potentialand market requirement, Trade Center UNCTADIGATT, Geneva. United Nations Development Programme, 1999. Human Development Report of Thailand 1999, United Nation Development Programme, Bangkok. United Nations Economic and Social Commission for Asia and the Pacific, 1979. Transnational Corporations and the International Commercialization of Pineapple Canned in Thailand, United Nations Economic and Social Commission for Asia and the Pacific, Bangkok. Wannitikul, Wilaiwan, 1981. The impact of government intervention on the organizational structure of the Thai sugar industry, PhD dissertation, University of Pennsylvania. Warr, PG., 1993. The Thai Economy in Transition, Cambridge University Press, Cambridge. -andNidhiprabha,Bhanupong,1996.Thailand?Macroeconomic Miracle: stable adjustment and sustained growth, World Bank, Washington, DC. Wattanavitukul, Somluckrat, 1978. 'Income distribution of Thailand', in H.T. Oshima and T. Mizoguchi (eds), Income Distribution by Sectors and Over Time in East and Southeast Asian Countries, Council for Asian Manpower Studies, Quezon City and Hitotubashi University, Tokyo:259-89. Welsch, D.E. and Tongpan, Sopin, 1973. 'Background to the introduction of high-yielding varieties of rice in Thailand', in R.T. Shand (ed.), Technical Change in Asian Agriculture, Australian National University Press, Canberra:124-43. Woradilok, Phonphimon, 1984. 'Kong thun phatthana chonnabot phua prachachon chonnabot' [Rural development fund for rural people], in Chonnabotthai 2527, National Economic and Social Development Board, Bangkok:105-15. World Bank, 1959. A Public Development Program for Thailand, Johns Hopkins Press, Baltimore. -, 1987. Thailand: agro-industrial diverstjkation:issues and prospects, World Bank, Washington, DC. ----, 1996. Thailand-Growth,Povertyand Income Distribution: an economic report, World Bank, Bangkok. Available online at http:// www-wds.worldbank.org/servlet/WDSContentServer/WDSP/IB/ 1996112113/000009265~3970311115145/Rendered/PDF/ multi0page.pdf. --, various dates. World Development Indicators, World Bank, Washington DC. Available online at http://www.worldbank.org/data/ onlinedatabases/onlinedatabases.html. Yamada, K., 1986. 'Tai koku san buroira no tainichi senryaku' [Thai broiler exporters' strategy towards Japan], Yokei no tomo, January:4G 54. Yamatnoto, H., 1998. Tdi Togyo Shi: yushutsu taikoku heno kiseki [The History of Thai Sugar Industry: a way to a giant exporter], Ochanomizu Shobo, Tokyo. Yamazaki, J., 1988. 'Tai koku no tennen gomu sangyo ni tsuite' [On the natural rubber industry of Thailand], Tai koku joho 23(4):5-23. Yingvorapunt, Suvitya, 1966. Phatthana chonnabot nai prathet thai (phim khrang thi 2) [Rural Development in Thailand, 2nd printing], Sahapracha phanit, Bangkok. EFFECTS OF DIRECT PROTECTION FOR AGRICULTURE THREE SOUTHEAST ASIAN COUNTRIES TakarnasaAkiYarnaand Kei Kajisa Until the recent financial crisis, the economic performance of East Asian developing countries over the past few decades had been outstanding world Bank 1993).' This performance is usually attributed to successful industrialisation, but there has been little recognitionof thesimilarlyexcellent performance of the region's agricultural sector. Agriculturein these countries has contributed to stable food supplies and prices, foreign exchangeearnings, and poverty reduction, and hence to general economic growth. Several key factors affect the performance of the agriculturalsector, includingtechnology, rural infrastructure and protection rates. In this chapter we focus on direct protection rates. Schultz (1964) made one of the first comprehensive analyses to link price distortions and high taxation in agricultural markets to stagnation of agriculture in developing countries. Until recently, governments in many developingcountries taxed the agriculturalsector heavily for various reasons, including the need for tax revenue, promotion of industrialiation, and the relatively weak political power of farmers. Mundlak et al. (1989) conducted a comprehensive, long-term analysis of the agricultural sector in Argentina, emphasising the important impact of taxation on its performance. Schiff and Valdes (1992) argue that the agricultural sector in many developing countries has suffered from negative protection (taxation). Bates (1981) argues that the main cause of the poor agricultural performance of the Sub- EFFECTS OF DIRECT PROTECTION FOR AGRICULTURE 381 Saharan African countries up until the mid 1980s was primarily large distortions in agricultural markets often controlled by marketing boards, a control deeply rooted in political and social structures. The agricultural sector in many Southeast Asian countries has performed considerably better than those of countries in other developing regions. The World Bank's (1993) study of the so-called East Asian miracle suggests that while there have been a number of market interventions in these countries, their effects have not necessarily been harmful. Hayami claims that, although East Asian countries 'exploited agriculture, they did not neglect to make necessary investments in irrigation and agricultural research to increase land productivity' (Hayami 1998:28). Timmer (1989) argues that policies related to direct protection in the region were pragmatic and designed to stabilise domestic prices in the short run while reflecting international price trends in the long run. In spite of the importance of the effects of protection policies on the performance of the agricultural sector, there has been relatively limited detailed quantitative study of their nature and evolution over time. Most past work has used a cross-country, multi-country approach with a large number of countries as samples (Krueger et al. 1992, Honma and Hayami 1986). An exception is the work by David and Huang (1993) on several Asian countries, but their study is restricted to rice. This chapter m i n e s the characteristicsof agricultural protection policies for key agricultural commoditiesin Indonesia, the Philippines and Thailand, and their evolution over the past few decades. It attempts to answer the following six questions. Are there differences in protection rates among the three countries? Have these policies changed over time? Have the policies for food been different from those for other crops? Have the policies had the effect of stabilising domestic prices? What other variables have af5ected the policies? What has been the importance of direct protection in total protection rates (sum of direct and indirect protection rates)? First, we review agricultural protection policies in the three counties over the past few decades and discuss the analytical framework we use here. Then we discuss the econometric method used in the analysis, and present the results of the quantitative analysis of the protection rates in view of policies actually implemented. 382 RURAL DEVELOPMENT AND AGRICULRJRAL GROWTH AGRICULTURAL PROTECTION POLICIES To examine the levels and the trends of direct protection rates and their importance in total protection rates, we calculated period averages of the direct rates for several key commoditiesover the past four decades, following the method suggested by Schiff and Valdes (1992). Commodities examined for the three countries were selected on the basis of their importance in the respective agricultural sector (Table 9.1) Indonesia-rice, maize, rubber, sugar, and palm oil The Philippine+rice, maize, sugar, and coconuts Thailand-rice, maize, rubber, and sugar. For all three countries, the weighted averages of direct protection rates (the geometric averages of four-five commodities weighted by production value evaluated by border price) have increased substantially fiom negative to positive (that is, fiom taxation to protection). However, because indirect taxation (negativeprotection)rates (taxationon the agriculturalsector through exchange rate and other policies favourable to non-agriculturalsectors) have been high and without discernible trends, the total protection rates for the commodities examined were negative until the early 1990s in Indonesia and Thailand and for the entire period in the Philippines. There are notable differences in the weighted averages of the protection rates among the three countries. Indonesia had the highest direct protection rates over the period, reflecting its policies toward agriculture as discussed below. In addition, petroleum has been an important source of government revenue, hence there has been less need for the government to tax exported agricultural products. Protection rates in the Philippines increased significantly over time but the weighted average for the second half of the 1990s is still the lowest among the three countries. This and the high indirect protection rates reflect the pro-industrialsector policies that have been implemented in the Philippines in the past (Intal and Power 1989). Thailand's protection rates changed drastically, from heavy taxation to si nificant protection, essentially following the pattern of protection in g relation to income observed by Timmer (1988). Indonesia Protection rates in Indonesia have increased substantially over time, with large reductions in taxation for exported commodities and increased protection for imports. Although palm oil and rubber (exported commodities) were taxed in the 1970s when world prices were high, the Table 9.1 Periodaverageof direct,indirectand totalratesof nominalprotectionfor selected commoditiesin Indonesia, the Philippines and Thailand Direct protection rate Oil Weighted Average of total Indirect Rice Sugar Maize Rubber palm Copra average protection rate protection rate Indonesia E Trade position EXJIM EX EX 1967-70 8.3 -7.3 na -8.0 -33.3 -27.5 2 CA 1971-80 22.60 -8.2 -24.7 -8.4 -22.8 -15.7 1981-90 29.80 0.0 -1.7 14.0 -7.2 -18.6 s? 8 1991-95 33.12 0.0 -3.5 32.1 3.5 -21.6 1967-95 24.70 -3.7 -8.5 5.3 -15.2 -19.5 n Philippines 4 Trade position g 1961-70 -28.7 -20.9 1971-80 -41.0 -23.8 2 1981-90 42.8 -24.5 2 1991-95 -19.8 -25.6 9 1961-95 -35.5 -23.4 8 Thailand ;d Trade position EXJIM EX 1962-70 -3.7 -13.2 -38.6 1971-80 -3.2 -19.1 -39.9 1981-90 -0.9 -12.0 4 . 9 1991-95 0.0 -0.1 3.9 1962-95 -1.9 -12.9 -25.5 Note: EX.exported commodity; IM: imported commodity. The average of'the total protection rate is the sum of the weighted average of direct w m protection rates and the indirect protection rates.The indirect protection rate is the average of the number presented in SchifTand Valdes (1992:15, w Table 2-1) and the number we calculated. It is computed using the free trade equilibrium exchdnge rate, defined as the exchange rate expected to prevail if the exchange rate is used to balance the current account under a completely free trade regime (Schiff and Valdes 1992). Source: See Appendix for details. 384 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH level of taxation was not high. Since the 1980s, protection rates have turned positive for all commoditie~,~making the rates very high for imported commodities. This reflects Indonesia's policies for rice and other crops. Indonesids rice policy has had two goals: (i) achieving food self-sufficiency by providing incentives for farmers to undertake production using new varieties and adequate fertiliser (Gonzales et al 1993, GATT 1995), and (ii) providing rice to urban consumers at 'reasonable' and relatively stable prices. To these ends, the parastatal BULOG implemented various policies, including floor prices at the farm gate, ceiling prices for consumers, and control of the international rice trade. Similarly, policies for sugar, maize and palm oil promoted price stabilisation and alleviation of the impact of high prices on consumers. The Indonesian government's policy objective of stabilising palm oil prices was pursued through direct government intervention in the market. Because palm oil was produced mainly by the state parastatal PTP in the 1960s and 1970s, marketing was conducted by a government agency until 1992. This agency controlled the market and maintained stable domestic prices. Even after 1992, the government intervened in the market, mainly through impositionof an export tax, but in a less aggressive way than before (Larson 1996). When world prices increased sharply, the government imposed heavy taxes on palm oil exports, thereby diverting the flow from export to domestic supply and reducing domestic prices so that they were considerably lower than world prices. The Indonesian government had encouraged sugar production on the Outer Islands because it was a strategic commodity for rural development on theseislands.One of theobjectivesof sugarprotection policiesin Indonesia has been to close the gap between incomes on the Outer Islands and incomes on Java (Gonzales et al. 1993). Philippines In the Philippines, the agricultural protection policy also focused on achieving self-sufficiency in rice and protecting agricultural products from imports. The National Food Authority (NFA) intervened in the rice market with the declared objective of stabilising prices. Rice was taxed when it was exported during the late 1970s and early 1980s, and protected during other periods (Intal and Power 1989). Because of decliningworld rice prices, the protectionrate has increasedsubstantiallyin recentyears.The government recognised maize to be an important commodity for the pork and poultry industries and pursued price stabilisation through NFA intervention. EFFECTS OF DIRECT PROTECTION FOR AGRICULTURE 385 Sugar in the Philippines was a heavily taxed commodity subject to considerable government intervention. The stated objective of the intervention was to regulate US sugar quotas before 1974 and to stabilise domestic prices after 1974 (Intal and Power 1989). Price stabilisation was achieved by imposing heavy taxes on those occasions when world sugar prices increased sharply and restricting imports when world prices were low, such as in the first half of the 1990s. The marketing of copra has been under the control of a monopolistic trade association with heavy intervention (Intal and Power 1989). Taxation on copra became high when the Coconut Industry Stabilization Fund put levies on this product from 1981 to 1986. Thailand Thailand is a major exporter of agricultural products. Up until the early 1990s, it heavily taxed most exported agricultural commodities, but especially rice. The effective tax on rice was high until the early 1980s. The objectives of high rice taxation were domestic price stabilisation and rent acquisition from importers because of Thailand's high market share in the world rice trade and the low price elasticity of import demand for rice.3In addition to the premium, quantitative restrictions on exports were used to keep domestic prices low and stable. The Thai government's stance on rice changed in the mid 1970s,when pressure was put on governments 'to pay at least lip service to the idea of "supporting" rice prices to the farmers' (Siarnwalla 1975:45). With the softening of world rice prices after 1982, the government gradually reduced and then abolished all export taxes, the premium, and reserve requirements. Thailand's main policy goal for sugar has been to provide an adequate income to cane producers while keeping domestic prices stable (Yiimamoto 1998; Siarnwalla and Setboonsarng 1989). One reason is political pressure put on the government by sugar producers, considered the best organised among Thai farmers. The government's objectives have essentially been achieved by keeping domestic prices higher than export prices through import restrictions. In 1974 and 1975, the government used an export ban to keep domestic prices lower than world prices, and in 1980, when domesticprices increased sharply becausedrought caused a significant decline in production, a governmental agency imported sugar. Thai maize was exported and taxed until the 1990s. There has been little intervention in this market, probably because of the relatively small size of the sub-sectorand the hct that maizeproducerswere not verystrong politically. 386 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Traditionally, rubber, Thailand's second most important agricultural export commodity after rice, was taxed heavily. However, taxation of rubber has been reduced substantially and effective subsidies have increased as world rubber prices declined from the early 1990s on (Thainugul et al. 1996).*Political pressure from rubber producers probably influenced this policy change. Price instability This discussion suggests that price stabilisation for the main products, especially imported food, was a key policy objective, a suggestion also made by past studies (Timmer 1989; Lindert 1991; Anderson and Hayami 1986). To examine the effect of policies on price stabilisation, we calculated short and long-term instability indicators. The short-term instability indicator is the average over the period of the absolute percentage year-to-year changes. It measures the extent of price instability on the basis of price changes from oneyear to another.Thestandard instabilityindex, the coefficient of variation, is used to evaluate long-term instability. It would indicate whether real domestic prices were more stable than real border prices around the mean for the entire period of about three decades. Table 9.2 shows the short and long-term price variability indices, respectively, for the real border and real wholesale prices, and the ratios of these two variables. If the ratio is greater than 1, it indicates that instability in real wholesale prices is smaller than real border prices. The ratios in Table 9.2 (columns 3 and 6) show considerably smaller variability for the real wholesale prices than for the real border prices, with the exception of Philippine copra, suggesting that the objective of domestic price stabilisation was achieved in both the short and the long run for most commodities. The extent of financial gains for producersfrom price stabilisation can be evaluated using the method developed by Newberry and Stiglitz (1981).5 Although this measure does not take into account the long-term effects of price stabilisation (see, for example, Kanbur 1984), it is accepted, thus we have calculated it for the commodities examined. The results are given in the last column of Table 9.2. For most of the commodities in Thailand and the Philippines (especially rice and sugar), producers gained from long-run price stabilisation. Rather surprisingly, except for maize, the price stabilisation effect on producer income was negative for Indonesia. This was probably due to the fact that Indonesia is a large producer of agricultural commodities, hence prices and EFFECTS OF DIRECT PROTECTION FOR AGRICULTURE 387 Table 9.2 Short-run and long-runprice instability of realborder price and real wholesale price and estimates of the benefits of price stabilisation (7) - - Producer (1) (2) (4) (5) benefits Real Real Real Real from price border wholesale (3)= border wholesale (6)= stabilisation price price (1)/(2) price price (4)/(5) (%I Indonesia Rice 1967-96 21.2 11.4 1.9 Maize 1967-96 17.7 11.5 1.5 Rubber 1967-80 19.3 18.6 1.0 Oil ~al,m1976-95 19.4 8.9 2.2 Sugar 1979-97 25.0 8.8 2.8 Philippines Rice 1960-93 26.4 8.6 3.0 Maize 1960-95 20.1 10.9 1.8 Copra 1960-95 28.5 33.0 0.9 Sugar 1960-95 20.3 13.2 1.5 Thailand Rice 1960-97 14.8 10.8 1.3 Maize 1967-81 13.4 12.2 1.1 Rubber 1960-91 17.5 14.8 1.2 Sugar 1962-95 26.7 9.2 2.9 aproducer bendts from price stabilisation are expressed in terms of percentageof production value. Source:See Appendixfor details. produ'ction of several commodities are correlated, implying the non- applicability of 'small-country assumptions'. This is evident for palm oil, of which Indonesia is a very, large producer, and, to a lesser extent, rubber. (Philippine copra is another case.) It is clear that rice and sugar producers in the Philippines and Thailand benefited significantly from price stabilisation policies. ANALYTICAL FRAMEWORK In this section we describe the approach we took to test empirically whether protection policies stabilised domestic prices and whether any variables affected such policies. Then we discuss the econometric technique we used. To begin with, we specified the relationship between domestic and border prices of agricultural commodities as where DPitis the domestic price of the agricultural commodity i in the year t,and BP, is the border price of the commodity i in the year t converted into local currency at the official exchange rate and adjusted for transport, storage and other costs. The magnitude of the domestic price elasticity with regard to the border price, b,would indicate the extent of interventionin price. Special cases are when p is nil or unity. When P isnil, the domestic price DP, is a and the border price, Bpi? has no effect on the domestic price. Conversely, when P is unity, the Ml percentage change of the border price is transmitted to the domestic price.GWhen the government implements policies to stabilise the domestic price and if the policies have the intended objective, then P will be between zero and unity. Equation 9.1 is slightly moditied so it can be estimated using econometric methods as Both coefficients, a and p, may change over time with some variables. Hence, assuming linearity of the functions, each coefficient is replaced with a, and p,, respectively,.that is where a,reflects an intervention rate constant over time, and 9reflects an intervention that changes over time with the variable (X). Similarly, Pcould change over time with some variables. Hence, in Equation 9.4 3and<. are variables that reflect changes in the economy or political reality that would cause the government to change protection policies. Combining Equations 9.2, 9.3 and 9.4 and adding an error term for econometric analysis, this equation is used to estimate EFFECTS OF DIRECT PROTECTION FOR AGRICULTURE 389 where eifis an error term assumed to be independent and identically distributed.This equation allows us to evaluate how a;and Pithave changed over time as the structure of the economy and/or commodity characteristics have changed. If only a,, is statistically significant, it suggests that the protection rate was constant over time. If both 7 s and @ are statistically si nificant, it suggests that the protection rate and the extent of stabilisation g changed with the variable BP,, 5and/or with <. over time. An empirical problem with this regression form is that the coefficients as and Ps may capture several other effects stemming from policies other than protection policies influencing the domestic price. Marketing, transport and storage costs within a country may systematicallyinfluence the domestic price, and thereby affect either or both coefficients. The coefficients should therefore be interpreted cautiously, with careful reference to a review of the policies actually implemented. To explain these rates, we reviewed past studies of protection rates and thevariables used in them.7Thesevariables can be classifiedinto two g r o u p s commodity-specific variables and variables related to economic, political and sectoral structural changes. The first group refers to characteristics of individual commodities, for example whether or not they are food crops, their trade position, productivity, and price volatility. The second group indudes such variables as income, importance of agriculture to the economy and competitiveness of a country's agriculture in the world market. We examined six variables in our analysis and their correspondence to the variables 5and lnx+ u (11.1) where y is value added, x and s are vectors of inputs and state variables respectively, r(s) and P(s, x) are the intercept and the slope of the function, respectively, and u is a stochastic term. At each sample point, the data consist of aggregated techniques, the compositionof which is likely to change over the sample points. To identify the aggregate production function, it is necessary to loosen the tie between decisions on the implemented technology and decisions on the level of inputs. This is achieved when deviations from the first-order conditions are more pronounced in the input decisions than in the choice of techniques. Variations in the state variables affect the production function coefficients directly as well as indirectly through their effect on inputs. For this reason, estimates obtained under the assumption of constant coefficients provide a distorted view. Often empirical estimates are not robust because they are sensitive to the choice of sample. We can illustrate this by evaluating the elasticity of average labour productivity with respect to a given state variable (say si) The state variablesmay not be independent.A change in one state variable may be associated with a change in the others, but this possible relation is ignored here for the sake of simplification. The first two terms show the response of the implemented technology to a change in the state variables, whereas the last term shows the output response to a change in inputs under constant technology. The elasticities in Equation 11.2 are functions of the state variables and thus vary over the sample points. The innovation DETERMINANTS OF AGRICULTURAL GROWTH 477 in this formulation lies in the response of the implemented technology to the state variables. To isolate this effect, we rewrite Equation 11.2, holding x constant to yield the elasticities When a production function is estimated under the assumption of constant coefficients, the effect captured by Equation 11.3 becomes part of the unexplained production function residual. Equation 11.3 captures the fact that a change in the state variablesmay cause a change in the composition of techniques in addition to a change in the inputs used on a given technique. As such, it is correlated with inputs, thus the estimates are biased. Estimation The estimation of Equation 11.1 requires a specification of the functions T(s) and P(s, x) in terms of the arguments, s and x. The product of P(s, x) with In x will give quadratic terms. The time-series data to be used here are highly intercorrelated (strong multicolinearity), and it is impossible to identi@the coefficients of the quadratic terms properly. The approach to identificationis to use the factor shares, but this information is not available. We therefore impose constant slopes, but allow the intercept to depend on the state variables. This reduces the impact of the term in Equation 11.3 on the residual, and thereby removes the bias due to the correlation of the residual and the inputs. To be precise, this eliminates only the linear component of the residual and the inputs, but for linear estimators this is all that matters. Strong multicolinearitydecreases the precisionof the ordinaryleast squares (OLS)estimates.In that case severalcoefficients are not significantlydifferent from zero, whereas others take on unreasonable values, such as elasticities larger than one. Elimination of variables with non-significant coefficients is inconsistent with our prior knowledge that the variables belong to the equation. For instance, we do not want to eliminate an important input &omthe production function. From a formal point of view, the elimination of a variable is equivalent to an imposition of a linear homogeneousconstraint on the coefficients of the function. There is a less costly possibility, namely to impose a constraint in such a way as to eliminate a linear combination of the variables in the equation instead of a particular variable. In general, when a variable or a linear combination of variables is eliminated from a 478 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH regression, the coeficients of the remaining variables are dected unless the variables are uncorrelated. This suggests that it is desirable to work with orthogonal (uncorrelated) regressors, which can be achieved by constructing orthogonal linear combinations of variables, referred to as principal components (PCs). The analysis begins with the computation of regression in terms of the principal components. The non-significant components are eliminated, thus the question is which and how many principal components to eliminate &omthe regression. For thiswe need a criterion. Herewe followthe algorithm in Mundlak (1981), which seeks to obtain the tightest confidence region for a given level of significance. We thus eliminate as many principal components as possible, subject to the restriction that the null hypothesis- that the coeficients arejointly equal to zer-is not rejected at the 5 per cent level of significance. This means that it is impossible to find a linear combination of the eliminatedprincipal components to add to the regression that would subsequently have a regression coefficient significantly different from zero. The next step is to convert the coefficients of the principal components to those of the original variables. When the regressors are written as a matrix, the number of regressors constitutesthe rank of this matrix.The rank minus the number of eliminated principal components is referred to as the statistical rank. Thus, the statistical rank states the number of linear combinations of the original regressors that exhaust the information embedded in the whole set of regressors. The empirical results show that in most cases the statistical rank is between two and four, reflecting the high degree of multicolinearity. The analysis begins with an estimation of the Cobb-Douglas production function with inputs alone. The inputs are irrigated land, rain-fed land, fertiliser, capital and labour. In general, the sum elasticities of a function with inputs alone are larger than one, some elasticities are larger than one, and others are negative or not significantly different from zero. This is then followed with a gadual introduction of state variables, the carriers of the implemented technology, starting with public goods consisting of measures of human capital and physical capital in infrastructure.The next step is to introduce incentives. In the search, we inspect the sum input elasticities, the Durbin-Watson statistics, and of course the sign of the coefficients. In the case of Indonesia and Thailand, serial correlation is not a problem. The situation is different for the Philippines, where the data show cyclical variations. To overcome this, we transform the variables as explained in the DETERMINANTS OF AGRICULTURAL GROWTH 479 section on the Philippines.The tables include the PC estimates obtained at the 5 per cent significance level and in some cases the OLS estimates. The latter are presented just as background information to illustrate how the choice of technique influences the estimates. In concluding this section, it is important to point out explicitly that we use primal estimates of the production function to derive marginal productivities. This is in contrast to the dual approach where prices are used to identlfy the production function. The dual approach is inferior for several reasons, as discussed in detail in Mundlak (2001). Beyond this, the basic assumption of maintenance of the first-order competitive condition disguises crucial facts needed to understand the development process in the countries under consideration. This becomes clear in our discussion of the empirical results. State variables In our application,state variablesscale production up or down, while leaving marginal rates of substitution unchanged. The state variables are referred to here as carriers of the implemented technology because they are correlated with that component of the residual which reflects changes in the implemented technology. The state variables included in the final results are roads (representing physical infrastructure),measures of education and health (representing human capital), and measures of incentives. Education is represented by the percentage of agricultural workers in Thailand and Indonesia who have no schooling (referred to as 'no schooling') and by the mean accumulated school years of the total labour force ('education') for the Philippines. Infant mortality rate represents the level of health. Both no schooling and infant mortality declined continuously during the period, whereas road length increased. These variables signify overall development during the period. We have also tried other measures such as electricity consumption, but strong multicolinearityprevented their inclusion. These physical and human capital variables are policy variables, since they are largely publicly financed. Their regression coefficients were significant, and this was not seriously affected by the choice of other regressors. As anticipated in the foregoing discussion, the inclusion of the state variables in the regression affected estimated elasticities in the expected direction-namely the sum elasticities became close to one and individual elasticities were mostly positive. As we show below in the discussion of factor growth, state variables account for an 480 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH important part of changes in total factor productivity (TFP). This is consistent with the assumption that introduction of the more productive techniques was supported by improvement in these variables. Unlike the role of policy variables, the role of prices was less consistent, although in general price coefficients had the right sign. The price effect is pronounced in the Philippines, exists but is not robust in Indonesia, and is not important in Thailand. Price variability was also important in the Philippines. Prices impact on growth in several ways. The regression coefficients of prices represent a direct impact of price variations on output, conditional on inputs. The indirect effect of prices on output occurs through their impact on the level of inputs and choice of technology. There is an additional effect which generally goes unrecognised. When there is a gap between the shadow price of an input and its market price, employment of the input will eventually rise. This is a generalisation of the formulation of the migration equation discussed elsewhere,12where the income gap between agriculture and non-agriculturegenerates a flow of labour to non-agriculture. Similarly,for instance, the gap between the marginal productivity of fertiliser and its market price increased fertiliser supply and consequently fertiliser use. This has been the case in all three countries. There can be various reasons for such a gap, which we will not discuss here. What is important for our current discussion is that, as long as the gap prevails, resources will flow, even when product price declines. This situation blurs the impact of prices on output in empirical analysis. GROWTH ACCOUNTING Agricultural technology improved dramatically during the study period. This change in available technology affected factor prices and their supply, which in turn resulted in productivity growth. The changes that took place over time are summarised in the growth accounting in Table 11.5.13The results are based on the tables in the country sections. In all countries, the growth rate of output during the first period (up until 1980 or 1981) was &rly similar, about 3.8 per cent for Thailand and the Philippines, and 3.4 per cent for Indonesia. The rates dedined in the second ~eriod(fiom 1980 on), with most of the decline occurriig in the TFI), not in the total factor. This is true in all countries, but the magnitude of the decline varied, with the steepest decline in the Philippines, from 0.98 per cent in 1961-80 to 0.13 per cent in 1980-98. The mildest change was in Indonesia, from 1.58 per cent in 1971-81 to 1.49 per cent in 1980-98. DETERMINANTS OF AGRICULTURAL GROWTH 481 Table11.5 Sourcesof growth for Thailand,Indonesiaand the Philippines Elasticity Change (%/year) Share of growth (%) Philippines 1961 1961 1980 1961 1961 1980 -98 -80 -98 -98 -80 -98 Output 2.55 3.81 1.38 Inputs Irrigated land 0.155 2.64 3.20 1.15 Rain-fed land 0.425 1.01 1.42 0.18 Fertiliser 0.077 5.36 7.35 4.90 Capital, ag. origin 0.101 3.75 3.47 3.35 Labour 0.181 2.17 2.30 1.50 Capital, machines 0.062 4.55 6.64 0.28 State variables Price 0.320 0.00 1.75 -0.70 Pricespread -0.696 0.03 0.14 -0.01 Inflation -0.104 0.11 0.67 -0.64 Education 0.213 1.65 1.46 1.60 Factor accumulation 2.30 2.84 1.26 Total factor productivity 0.25 0.98 0.13 State variables Portion of TFP due to state variables Indonesia 1971 1971 1981 -98 -81 -98 Output 3.39 3.69 3.04 Inputs Irrigated land 0.457 0.80 0.97 0.69 Rain-fed land 0.230 0.52 -0.25 0.68 Fertiliser 0.084 8.18 14.45 3.69 Capital 0.031 11.59 8.00 12.68 Labour 0.198 1.88 1.32 1.95 State variables Wholesale price ratio 0.127 1.36 0.84 2.37 Pricespread 0.161 0.10 0.75 0.11 No schooling -0.003 -1.30 -1.43 -1.22 Roads 0.073 5.71 5.46 5.09 Infant mortality -0.002 -2.79 -2.97 -2.40 Factor accumulation 1.90 2.10 1.56 Total factor productivity 1.49 1.58 1.49 State variables Portion of TFP due to state variables 482 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Table11.5 Sources of growthfor Thailand,Indonesia and the Philippines(continued) Elasticity Change (%/year) Share of growth (%) Thailand 1971 1971 1981 1971 1971 1981 -95 -81 -95 -95 -81 -95 Output Inputs Irrigated land Rain-fed land Fertiliser Capital Labour State variables Real farm price Inflation No schooling Roads Infant mortality Factor accumulation Total factor productivity State variables Portion of TFP due to state variables Alternative Thailand Output Inputs Irrigated land 0.129 Rain-fed land 0.286 Fertiliser 0.059 Capital 0.377 Labour 0.149 State variables Real farm price 0.023 Inflation -0.295 No schooling -0.008 Roads 0.081 Electricity 0.045 Infant mortality -0.003 Factor accumulation Total factor productivity State variables Portion of TFP due to state variables DETERMINANTS OF AGRICULTURAL GROWTH 483 This seems like a paradox where the technical change is recorded more as a chan e in total factor than a change in TFP; however, it is consistent with g chan es that take place during the transition to more advanced techniques g that are intensive in scarce resources (Mundlak 2000). The magnitude of TFP is path-dependent in that it depends on factor supply. When new techniques are intensive in scarce inputs, producers cannot shift immediately to make full use of the new technique because of insufficient supply of critical inputs. As a result, the shadow prices of these inputs rise, which in turn raises the level of the total factor, thereby reducing the measured TFP change. This does not reduce the impact of a change in the available technology but rather states that part of the technical change is absorbed in the factor prices. As a consequence, the quantity of the scarce inputs increases, and eventually the gap between the shadow price and their long- run supply price tends to disappear. The situation is different if all the inputs needed to implement the new technique are sold in a competitive market with a perfectly elastic supply. In that case, the share of TFP in total growth is expected to be more substantive. Turning to individual inputs, irrigated land accounts for 1&16 per cent of output growth. The contribution of rain-fed land is substantial in the Philippines and Thailand during the first period, and far less important during the second period. This pattern also follows from the choice of techniques framework. Unlike traditional crops, new varieties and crops are intensivein irrigated land, thus scarce resourcesare mobilised to the irrigated land, causing the productivity of the non-irrigated land to suffer. For the period as a whole, fertiliser accounted for 14-20 per cent of growth. The relative importance of fertiliser was stable in Thailand, declined drastically in Indonesia, and increased in the Philippines in the latter period. The increase in the relative importance of fertiliser in the Philippines is partly a result of a decline in output growth (so the same impact of fertiliser carries a higher relative weight) and partly an alleviation of the supply condition. The relative importance of capital grew over time in d three countries, most significantly in Thailand. This is substantive evidence that the new techniques are capital-intensive. There is less uniformity in the contribution of labour to growth. In the first period it was 14 per cent in Thailand, 7 per cent in Indonesia, and 11 per cent in the Philippines. In the latter period a gap opened up, with the contribution of labour to growth almost doubling in the Philippines and Indonesia, and declining to only 2 per cent in Thailand. The decline in 484 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Thailand is consistent with the hypothesis that the initial endowment of rural labour exceeded needs, and that output growth was not in labour- intensive techniques. Altogether, the state variables accounted for a large proportion of TFP growth. Indeed, in Indonesia they practically exhausted it. There is some variability in each country in the performance between the two periods. The elasticities used in the calculations are the same for the whole period, thus there will be short-term departures from the estimated values. Nevertheless, the overall record indicates that state variables serve well as carriers of the implemented technology shocks. Roads as a representative of physical infrastructure accounted for 11-15 per cent of output growth in Thailand and Indonesia This variable was not included in the regression for the Philippines. Schooling made a similar contribution, with some variability over time, as did infant mortality as a measure of health. The price variable made a substantial contribution-in Indonesia it accounted for 10 per cent of output growth in the second period and 5 per cent for the period as a whole. In the Philippines, where prices varied considerably more than in the other countries, it contributed about 15 per cent in each of the two periods, but with different signs, so the net contribution was nil for the period as a whole. Overall, the contribution of the price spread was negligible. POLICY IMPLICATIONS The purpose of this analysis is to gain an understanding of underlying processes, which will enable us to evaluate roles for positive policies. At the aggregate level of the analysis, we can assess !growthand income distribution. It is an underlying fact that major technological advances changed agricultural production during the study period. In addition, there was an important development in non-agriculture in all three countries, at least in part of the study period. The input requirements of the new technologies were skewed in the direction of capital inputs, mainly irrigatedland, fertiliser and other forms of capital. By definition, capital is scarce, thus the implementation of these new technologies stretched over a long period of time. This relates to the supply side. On the demand side, the three countries had to expand their exports in order to supplement growing domestic demand and absorb growing supply. The pace of growth was determined largely by the flow of resources to agriculture, as reflected in the weights DETERMINANTS OF AGRICULTURAL GROWTH 485 these inputs receive in accounting for output growth. The message for the future is clear-for growth to continue, available technologies must continue to grow. Without such growth, the impact of input growth will eventually decline. We already see some evidence to this effect in the later years of the study period, but this is not the only determinant of future gowth. For agricultural producers to be able to take full advantage of new techniques, there must be a smooth flow of the required resources to agriculture. In retrospect, it would have been far more productive to respond without delay to the jump in fertiliser demand generated by the Green Revolution by allowing imports rather than relying on home production. The output of grain forgone because of the anti-import bias would have paid nicely for imported fertiliser. The state variables indicate that public goods are important in facilitating the implementation of new technologies. Physical infrastructure integrates areaswith major markets and reduces the costsof transactions.Other variables such as electricity, which did not enter the analysis because of the high correlation with roads, have their own important impact. Investment in such projects is not immediately connected to agricultural programs, but nevertheless has a strong impact on agricultural growth, and of course on the welfare of the rural population. This is also the case with health and schooling. Investmentin such programsis constrainedby resourceavailability, and it is in this sense that capital scarcity plays an important role in determining the pace of growth. Assuming that changes in the available technology facilitate growth, the focus should be to allow the inputs in demand to flow into agriculture and avoid a gap between their shadow price and the long-run supply price. This has several consequences. Growth will be fastest and the benefits will be directed mainly to farmers rather than the distribution channels that always benefit from shortages. Not independently, the contribution of TFP will increase relative to total factor. The statement on the removal of obstacles to the flow of resources is meant here to be a road signal and not a detailed road map of an elaborate program. The elimination of obstacles has many aspects related to the distribution system, bureaucratic standards, and elimination of monopolistic lacunae along the way. It may not sound like a dramatic program, but its importance cannot be exaggerated. On the whole, the new technologies are labour saving, and, along with natural population growth in agriculture, generate an oversupply of labour in agriculture. The excess supply is directed to non-agriculture, but the 486 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH ability of non-agriculture to absorb labour has to develop at a rather fast rate, the reason being that the more productive techniques in many industries are labour saving and more profitable even in countries with low wages. Lowagriculturalwages are one outcome of thisgap.That said,however, as is shown in Butzer, Mundlak and Larson (Chapter 12, this volume), the same type of investments in education and health services that spur productivity gains on the farm also facilitate the flow of agricultural labour resources to other sectors. Some of the country chapters suggest that rural poverty reduction was not progressing well, if at all. While aggregate poverty rates have fallen, much of the reduction has been urban poverty rates.'* Such lack of progress can be attributed to inadequate transfer policies. The more hndarnental question, however, is why the gowth that occurred did not lead to any decline in the incidence of poverty. This is another aspect of the issue of new technologies discussed above. Because the technologies are labour saving and wages are kept relatively low, labour income remains low. The wage rate did improve in some countries, but the big unknown is the average on- farm employment to which the daily wage rate is applied. In this situation, the welfare of landlesslabour is not improvingand may even be deteriorating. On the other hand, the situation of land and capital owners improves as the demand for the resources in their possession increases and with it their returns. Over and above this effect, landowners have a natural advantage of being able to work more days on the farm and thereby increase their annual wage income even when they receive the same daily wage rate. Aside from transfer programs run for humanitarian purposes, the alleviation of rural poverty depends largely on development of employment opportunities outside agriculture. This can still occur in rural areas, but it is a separate issue related to the geography of development. The trade terms of agriculture play several roles, some of them backstage. The flow of resources to agriculture depends on the relative profitability of agriculture, and this in turn depends on real product price. Similarly, the choice of new techniques is sometimes justified only in a good price environment, which helps to offset initial setup costs, as well as risk. The real price is determined by input prices and also by prices of non-agricultural ~roducts.Such prices are determined in the economy at large, which generates the economic environment within which agriculture operates. Even though the macro environment is not part of agricultural policy, it can still hurt agriculture. Finally, world agricultural prices affect domestic DETERMINANTS OF AGRICULTURAL GROWTH 487 prices and thereby the profitability of agriculture. The challenge here is for the countries under study to form an economic environment that will allow them to make the same progress that has led to dedining prices in the rest of the world. AGRICULTURAL PRODUCTIVITY THAILAND Background This short summary of events is based on Shigetomi (Chapter 8, this volume), Siamwalla (1996), and Coxhead and Plangpraphan (1998), among others. Agriculture. Thailand experienced agriculture-led growth at a fast rate from the 1950s through the 1970s, and at a slower rate thereafter. In the process, the share of agriculture in GDP declined from about 38 per cent in 1951 to about 10 per cent in 1995. Over the same time period, the share of manufacturing rose from 13 per cent to 32 per cent. The growth of agricultural production outpaced that of demand. This expansion facilitated the growing domestic supply of food at relatively low prices and growing exports of agricultural products that served as an important source of foreign exchange. In addition, agriculture contributed the labour needed to develop the non-agriculturalsector. In the 1970s and 1980s, agricultural employment increased; in 1989 it started to decline. Its share in the labour force, however, has declined all along, a trend accompanied by growing migration from agriculture to the cities. The share of the total active population in agriculture declined from 83 per cent in 1961 to 57 per cent in 1999. The economic boom caused a rise in wages in both agriculture and in non-agriculture. Agricultural policy underwent some major changes over the years. Until the mid 1970s, agriculture was taxed. This policy changed in the mid 1970s to income taxes and price supports, and agribusiness promotion. Starting in the early 1980s, policies shifted to agricultural protection, production diversification, and control. The economy. Prior to the 1960s,Thai agriculture relied largely on rice and rubber production. The 1960s and 1970s were a period of high economic growth. Modern rice varieties were introduced in the late 1960s and started to gain in relative importance in the 1970s. During the 1970s, agriculture benefited from the economic growth and fivourableworld prices for agricultural products that were a featureof the decade. Exportsexpanded, 488 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH as did cultivatedland.This expansion was associated with crop diversification, including the expanded production of export-oriented upland crops. The share of exports in total agricultural output rose fiom 31 per cent in 1971 to 58 per cent in 1982, and thereafter fluctuated at around 50 per cent. The early 1980s were painfd for the economy as a whole, with a recession lasting until 1986. This period was followed by an industrial boom and accelerated economic culminating in the economic crisis of 1997. Demand. Per capita rice consumption declined fiom about 145kgin 1971- 75 to about 105kg in 1995-97. Output increased at a faster rate than consumption, and the surplus grew from roughly 2 million tonnes in 1961- 75 to 6 million tonnes in 199697. In view of the impressive growth of exports, it is tempting to assert that demand is not a constraintto agricultural production (Martin and Warr 1993; Punyasavatust and Coxhead 2001). This assertion ignores the fact that global demand is a constraint to global agricultural output, and that when supply growth exceeds demand growth, prices decline, as has been the case for the past few decades. In fact, for some time Thailand tried to use its marketing power and export taxes to control exports in order to prevent world prices from falling. Still, falling prices do not prevent countries from exploiting their comparative advantage by exporting certain products so as to alleviate the constraints of domestic demand. Even then, output growth in most cases does not deviate much from growth in demand. We can see this if we estimate a pseudo-Engel curve by regressing per capita agricultural output on per capita total output. The values for the resulting elasticity are 0.3 for 1961-95, 0.26 for 1971- 95, 0.51 for 1961-71, 0.28 for 1971-81, and 0.25 for 1981-95. This elasticity expresses the proportional growth in per capita agricultural output associated with a proportional growth in per capita total output. Output here is GDE which is a good proxy for income. These values are not unreasonable. As indicated by Honma and Hagino (Chapter 10, this volume), Thailand's export gowth rate for a ricultural products averaged 9.7 per g cent per year from 1961-63 to 1995-97. The gowth rate during the commodity boom of the 1970s was particularly impressive-20 per cent per year in US dollars and 10 per cent per year in volume. Thailand maintained this export expansion until the currency crisis in 1997. The main export crops are rice, rubber, cassava, sugar and maize. The export of maize and cassava was important in the 1970s, but waned later when the targeting quotas of Japan and the European Economic Community were DETERMINANTS OF AGRICULTURAZ. GROWTH 489 terminated. This suggests that such exports were not a pure manifestation of comparative advantage and that the implicit social price received was not identical with the actual price. Export levels were sustained by expansion of natural rubber and sugar exports. Estimation. The reader interested only in the final results is invited to skip this discussion and move directly to Tables 11.8 and 11.9. The purpose of the discussion leading up to Table 11.8 is to explain considerations leading to the final results.The main issues are the choice of the principal component estimator to overcome the strong multi~olinearit~,the role of public inputs as carriers of technology, and the role of prices. We begin the estimation of the Cobb-Douglas production function with inputs alone. This is followed with a gradual introduction of state variables, starting with public capital (human and physical), and incentives. In the search we inspect sum input elasticities, DW statistics, and of course the sign of the coefficients. Table 11.6 presents a production function with inputs only. The statistical rank of the principal components estimation is 2, reflecting the high correlation among inputs. The sum elasticities are high, at 1.47 and 1.68 for the OLS and principal components estimates respectively.This reflects a rise in inputs that confounds the technicalchange and other state variables. The last column contains the normalised principal components elasticities, obtained as the ratio of the individual elasticities to their sum. As we will see below, these values are close to the final results of the analysis. In what follows, we present only the principal components results. Table11.6 Restricted production function: Thailand, 1971-95, withoutstate variables b OLS t OLS Norrnalised t PC Norrnalised Constant -3.461 -0.4 -8.3 Irrigated land 0.319 2.6 0.217 43.9 0.130 Rain-fed land 0.399 1.2 0.272 9.1 0.394 Fertiliser 0.070 1.4 0.048 35.3 0.049 Capital 0.438 2.8 0.298 12.2 0.288 Labour 0.243 2.2 0.165 12.5 0.139 Sum 1.469 1 .ooo 1.000 Rank DW RZ Source: Authors' calculations. 490 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH The next move (Table 11.7) is to introduce the infrastructure variables, or public goods. Because this group is highly intercorrelated, the size of the group was reduced. For schoolingwe use the measure of no schooling because it displays more variability around the trend line than the average level of schooling. The degree of infant mortality is chosen to represent the level of health, and road length represents physical infrastructure. The introduction of these three state variables does not change the statistical rank, and two principal components still summarise all the information contained in the regressors. The allocation of the explanation, however, is different, and the sum elasticities are now close to 1. This is consistent with the assumption that the level of public inputs is correlated with changes in the implemented technology. The last column presents the normalised elasticities. In Table 11.8 we present the results with two price measures added to the regression-the real farm price of rice and the inflation rate.15The statistical rank is still 2, even though the prices are not highly correlated with the other regressors. The signs of the respective coefficients are in line with conventional expectations, even though this expectation stands on weaker gound within the choice of techniques framework. The impact of the introduction of incentives on the input elasticities is rather weak, but sufficient to reduce the sum elasticities to 0.91. The weakness of the impact Table11.7 Production function: Thailand, 1971-95, with state variables t PC Normalised Constant 8.1 Irrigated land 39.3 0.146 Rain-fed land 3.7 0.271 Fertiliser 24.3 0.058 Capital 9.3 0.413 Labour 5.9 0.111 Sum 1.000 No schooling -24.0 Roads 31.1 Infant mortality -23.1 Rank DW R2 Source: Authors' calculations. DETERMINANTS OF AGRICULTURAL GROWTH 491 may reflect the fict that the variabilityin the incentivesis ratherlowcompared to the big trend changes in the other regressors and the output. The low marginal contribution of the incentives also occurs in other combinations of the incentives. This outcome is very different to that obtained for the Philippines where the price variability was larger and the trend of the regressors was weaker. The normalised elasticities of this regression are presented in the last column. There is a great deal of resemblance among the orders of magnitude of the normalised elasticities in the three tables. As indicated above, this is interpreted to mean that the state variables are strongly correlated with the technology carriers. This is not to claim that the public inputs included are identical with the technology variables. It simply indicates that, because of the strong correlation between the public inputs and the chan es in the g implemented technology, a subset of these variables can represent the disturbances that caused the sum elasticities in the naked regression to exceed 1. In the discussion that follows,we concentrateon the normalisedelasticities in Table 11.8. The sum elasticities of the irrigated and rain-fed land are 0.38. It is convenient to identify the estimated elasticities with factor shares. In that case, we can say that land accounted for about 38 per cent of output. The share of rain-fed land was about twice that of irrigated land. Table11.8 Production function: Thailand, 1971-95, base model t PC Normalised Constant 9.2 Irrigated land 31.0 0.132 Rain-fed land 2.7 0.248 Fertiliser 16.2 0.061 Capital 7.5 0.415 Labour 10.0 0.144 Sum 1.ooo Price 1.1 Inflation -14.5 No schooling -32.9 Roads 27.5 Infant mortality -16.5 Rank DW RZ Source: Authors' calculations. 492 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH The fertiliser elasticity is 0.06, which is in line with the value obtained and discussed for the two other countries, and in other studies of Thailand (Mundlak 1993). To judge the reasonableness of this value, we note that the proportion of fertiliser and lime in total farm household income in Thailand in selected years was 2 per cent in 1963, 5 per cent in 1970, 1977 and 1981, 4 per cent in 1971, and 6 per cent in 1983 (Mundlak 1993). The elasticity of capital is 0.41, which is quite high, and that of labour is 0.14, which is quite low. These values are consistent with the hypothesis that new techniques are capital-intensive and that capital was scarce. Shadow prices We now move directly to the results in Table 11.9 and Figures 11.5-11.9. The ratio of the marginal productivity of irrigated land was roughly 2.5 t i e s that of rain-fed. This ratio declines from 3.5 at the beginning of the Table11.9 Thailand: productivity,prices and shadow prices A. Productivity (1993 US$) Elasticity Avg. Marginal Avg. Marginal Avg. Marginal Irrigated land ($/ha) 0.132 2,670 352 2,745 362 2,611 344 Rain-fed land ($/ha) 0.248 559 138 456 113 632 156 Fertiliser ($lmt) 0.061 8,760 538 12,749 783 5,869 361 Capital (rate of return) 0.415 0.47 0.20 0.42 0.17 0.52 0.22 Labour ($/year) 0.144 548 79 502 72 576 83 B. Reported prices (1993 US$) Wage rate ($/year) 311 285 328 Fertiliser price ($/mt) 873 1,128 696 Fertiliser, distortion rate 0.62 0.69 0.52 C. Marginal rates of substitution Irrigated for rain-fed land 2.54 3.21 2.20 Irrigated land for labour 4.47 5.01 4.15 Irrigated land for wages 1.13 1.27 1.05 Irrigated land for labour adjusted 1.55 1.74 1.44 Irrigated land for capital 1,784 2,095 1,599 D. Derived prices (1993 US$) Irrigated land ($/ha) 2,346 2,411 2,294 Irrigated land-capital base ($/ha) 2,373 2,787 2,127 Source:Authors' calculations. DETERMINANTS OF AGRICULTURAL GROWTH 493 period to 2.2 at the end of the period. The decline reflects an increase in the productivity of rain-fed land, probably as a result of the introduction of new crops and overall practices. On the other hand, expansion of irrigated land might have caused a decline in its productivity on the margin, since new land brought under irrigation may be of lower quality than prevailing irrigated land. Similarly, expansion due to an increase in irrigated land may lead to lower-value crops. The productivity differential of the two types of land in Thailand is similar to that in the Philippines, but much smaller than that in Indonesia. There is a considerable discrepancy between estimated labour elasticity and labour share. This discrepancy points to a profound data problem. The labour share was computed by multiplying the daily wage by the labour data, assuming 150 working days per year. The agricultural labour share, calculated in current prices, fluctuated between 0.4 and 0.7, with an average of 0.58 (Figure 11.11). This is roughly four times the estimated labour elasticity. More striking, however, is the low value of the labour share in Figure11.ll Labour share of agricultural GDP in Thailand 80 Per cent Source:See Mundlak,Y.; Larson, D.F. and Butzer, R, 2002. Detminants ofAgrirltura1 Growth in Indonesia, the Philippines,and ThaiLznd,Policy Research Working Paper No. 2803, World Bank, Washington, DC. for discussion of data sources. 494 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH non-agriculture (Figure 11.12), which declined from a level of 0.16 in the early 1960s to a level of 0.06 in the late 1980s,with an average of 0.098. It is very likely that the sectoral composition of the labour data exaggerates labour in agriculture and understates labour in non-a riculture. For non- g agriculture, doubling the working days per year would give a more realistic labour share, but we have no basis on which to change the data and will therefore continue using the data we have, keeping in mind the limitation. The ratio of marginal productivity of irrigated land to labour declined from nearly 1.74 labour years in 1971-81 to 1.44 in 1981-95 (line C4, Table 11.9). This decline may reflect a decrease in the productivity of irrigated land or a rise in labour productivity. The average for the whole period is 1.55 years. Capitalising this by a discount rate of 0.15, it would require nearly 10 years of work to acquire one hectare of irrigated land. The value was higher in the 1970s and declined in the later period. Multiplying the value by the annual wage of $311 (in 1993 dollars), the derived value of one hectare of irrigated land varied from $3,306 in the first period to Figure11.12 Labour share of non-agriculturalGDP in Thailand 20 Per cent Source: See Mundlak,Y; Larson, D.F. and Butzer, R., 2002. Determinants ofAgngnmltural Growth h Indonesia, the Philippines,and Thailand, Policy Research Working Paper No. 2803, World Bank, Washington, DC. for discussion of data sources. DETERMINANTS OF AGRICULTURAL GROWTH 495 $3,145 in the later period, with an average of $3,214 for the whole period (these values are not reported in Table 11.9). Alternatively, the value of irri ated land can be derived by discounting the marginal value product of g land by 15 per cent. These values are reported in line Dl of Table 11.9 and are lower than the wage-based values. The marginal productivity of irrigated land in terms of capital ranged from $2,095/ha during the first period to $1,599/ha in the later period (line C5). Ths is the ratio of rent to user cost of capital. Assuming that the depreciation rate accounts for one-third of the interest rate, we obtain capitalised value of land that varied across the two periods from $2,787 to $2,127 (line D2). The ratio of the marginal productivity of labour to capital is about $400 for the whole period. This is the estimate of the ratio of the wage to the user cost of capital. The agricultural wage rate was $311, from which we can solve for the user cost of capital, which is quite high at 78 per cent for the whole period with small variations over time. This is a result of either a high marginal productivity of capital or-what is more likely-a low marginal productivity of labour, which in turn may be related to ambiguity in the content of the labour data. To evaluate this impact, we can assume, for instance, that the reported labour force in agriculture is twice as high as the actual force. A correction for this would reduce the labour-capital ratio, double the shadow price of labour, and reduce the shadow interest rate without changing the elasticities. A reduction of the elasticity of capital compensated by the elasticityof labour would also reduce the shadowinterest rate. Such calculations highlight the fact that our conclusions are sensitive to the assumption relating to the labour force. Growth accounting Table 11.5 presents calculations of TFP for the period as a whole and for the two sub-periods, 1971-81 and 1981-95. For the whole period, output grew at an average rate of 3.4 per cent, of which 67.6 per cent was due to the growth in total fictor and the remaining 32.4 per cent to TFl? In the earlier period, output grew at an average rate of 3.8 per cent, and the division to total factor and TFP was similar to that of the whole period. In the later period, the output growth rate was 3.2 per cent, of which 73 per cent was due to growth in total factor and only 27 per cent to TFl? In other words, the growth rate of TFP declined from 1.3 per cent in the first period to 496 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH 0.87 per cent in the second period, a decline of nearly 50 per cent. At the same time, total factor gowth changed only slightly. This movement is consistentwith changesthat take place duringthe transition to moreadvanced techniques that are intensive in scarce resources. The gowth rate of capital in the second period was 3.2 per cent, cornpared to 1.0 per cent in the first period. Meanwhile, the gowth rate of fertiliser was quite high in both periods. In terms of the contribution of individual inputs to gowth, labour contributed 14.3 per cent in the first period, only 1.9 per cent in the later period, and 8.6 per cent for the period as a whole. This rather small contribution is consistent with the fict that the initial endowment of rural labour exceeded the need, and that output growth was not in labour- intensive techniques. Fertiliser accounted for 19 per cent of growth, which is indeed substantial, but also alarming, since this source of growth cannot go on forever. The contribution of irrigated land declined from 13.3 per cent in the early period to 10.7 per cent in the subsequent period. At the same time, the contribution of rain-fed land declined from 8.9 per cent to a mere 0.7 per cent.This means that the contribution of total land expansion was practically coming entirely from the expansion in irrigated land. Finally, the contribution of capital was substantive, and increased drastically from 11 per cent to 40.6 per cent. Turning to state variables, they accounted for 37.5 per cent of output growth for the whole period, compared to 32.4 per cent for TFI? For the period as a whole, the weight given to state variables slightly exaggerated their importance as carriers of the implemented technology. For the sub- periods, the situation was reversed. As already stated, the elasticities used in the calculations are the same for the whole period, and it is therefore natural that there will be over and undershooting for shorter sub-periods. Most of the contribution of state variables is due to roads (physical infrastructure), education (a decline in the percentage of agricultural workers with no schooling), and health (a decline of infant mortality rates). To support the statement that roads are representative of physical infrastructure in general, we ran the regression in Table 11.8 again with electricity added (Table 11.10). The various coefficients changed very little, except for roads, which declined to 0.08, while electricity was 0.045. The growth rate of electricity was about twice that of roads, hence their contribution to gowth was nearly the same (Mundlak et al. 2002). DETERMINANTS OF AGRICULTURAL GROWTH 497 Discussion There are claims in the literature that agricultural production in Thailand increased largely because of land expansion. This might have been the case in the earlier years, but not in the study period, when land expanded at an average annual rate of 1.1 per cent while other inputs expanded at higher rates. Specifically, fertiliser increased at an average annual rate of 10 per cent, whereas irrigated land and capital increased at average annual rates of 3.5 and 1.8 per cent respectively. Consequently, total factor accounts for 67.6 per cent of output gowth. When the contribution of rain-fed land (4.5 per cent) is subtracted, total fictor contributed to 62 per cent of growth. This is similar to the experience of other Asian countries, and consistent with the proposition that, in general, shocks that improve profitability cause land expansion and a positive change in the intensive margins (Mundlak2000).The rise inTFP reflects improvedcrop varietiesand changes in output composition. Output growth was also influenced by growth of livestock production, which automatically increased output per hectare. The growth of agricultural production was associated with a remarkable growth in public goods-roads, electricity, health and education, all essential Table 11.10 Production function: Thailand, 1971-95, augmented base model b OLS t OLS t PC Normalised Constant 12.545 0.9 12.7 Irrigated land 0.238 0.8 27.3 0.129 Rain-fed land -0.025 -0.1 3.1 0.286 Fertiliser -0.053 -0.6 17.9 0.059 Capital -0.012 -0.0 7.4 0.377 Labour 0.293 1.5 10.1 0.149 Sum 0.442 1.000 Price 0.022 0.3 0.8 Inflation -0.003 -0.0 -11.3 No schooling -0.003 -0.4 -29.5 Roads -0.105 -0.5 34.0 Electricity 0.297 1.G 16.0 Infant mortality 0.003 0.8 -13.8 Rank DW R2 -- Source: Authors' calculations. 498 RURAL DEVELOPMENT AND AGRJCULTURAL GROWTH for the effective implementation of improvements in available technology taking place in Thailand as well as other Asian countries. Finally, incentives did not play an important role in the estimated production function, which indicates that growth conditions were generally favourable and not seriously darnaged by declining prices. Note, however, that this analysis does not cover the macro shocks associated with the 1997 financial crisis. AGRICULTURAL PRODUCTIVITY INDONESIA Estimation The reader interested only in the final results is invited to skip the current discussion and move directly to Tables 11.14-1 1.16. The purpose of the discussion leading up to Table 11.14 is to explain considerations leading to the final results. The main issues are choice of the principal components estimator to overcome the strong multicolinearity, the role of public inputs as carriers of the technology, and the role of prices. We have analysed data for three periods, 1961-98, 1971-98 and 1980-98. To avoid long technical discussions, however, we discuss only the last two periods. We begin estimation of the Cobb-Douglas production function with inputs alone. Next, we introduce public capital (human and physical), followed by incentives. In the search, we examine the sum input elasticities, DW statistics, and of course the sign of the coefficients. The tables contain the principal components estimates obtained at the 5 per cent significance level. Table 11.11 also presents the OLS estimates as a background to Table11.11 Productionfunction:Indonesia,withoutstatevariables 1980-98 b OLS t 01,s b PC t PC Constant -29.255 -4.5 -35.817 -5.7 Irrigated land 0.868 3.0 1.389 9.2 Rain-fed land 0.029 0.3 0.115 1.4 Fertiliser 0.028 0.6 0.108 2.9 Capital -0.060 -1.3 -0.088 -1.8 Labour 1.580 5.2 1.398 4.4 Sum 2.923 Rank 4 DW 1.957 1.850 R2 0.996 0.995 Source: Authors' calculations. DETERMINANTS OF AGRICULTURAL GROWTH 499 illustrate the impact of strong multicolinearity.In general, the DW statistics do not flag serial correlation. The R2 is high in all cases, but this does not mean much in view of the strong trend in the variables. Table 11.11 presents a production function with inputs alone. The statistical rank is 3 for the whole period and 4 for the shorter period. The order of magnitude of the estimates is not sensible, and the sum elasticities for the shorter period are excessively high, reflecting a rise in inputs that confounds the technical change and other state variables. The next step is to introduce infrastructure variables, or public goods (Table 11.12). From here on we present only the principal components estimates. The initial set of ~ublicgoods includes no schooling, infant mortality, and length of roads. We prefer the use of no schooling because it displays more variability around the trend line than the average level of schooling. The degree of infant mortality is chosen to represent the level of health, and road length represents physical infrastructure. The introduction of these three variables to the initial set of Table 11.11reduces the statistical rank to 2, which means that two principal components summarise all the information contained in the regressors. The estimates, however, are quite different from those in Table 11.11, and the sum elasticities are now close to 1. This is consistent with the assumption that the level of public inputs is correlated with changes in the implemented technology. The last column in each block presents elasticities that are normalised so that they sum to 1. Table 11.12 Productionfunction: Indonesia, with state variables 1971-98 t PC Normalised t PC Normalised Constant Irrigated land Rain-fed land Fertiliser Capital Labour Sum No schooling Roads Infant mortality Rank DW RZ Source: Authors' calculations. 500 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH In Table 11.13 we present the results with two price measures added to the regression-the real price of agriculture lagged one year, taken as the wholesale price ratio, and its spread. The main impact of the introduction of prices is to change the statistical rank to 4 for the longer period and to 1 for the shorter period. There is some similarity, however, in the order of magnitude of the estimates for the two periods. Compared to Table 11.12, the change in coefficients caused by price variables is not substantial. There are two possible explanations for this. First, price does not matter at all. This explanation is doubtful since the simple correlation between output and wholesale price is 0.73 (reported in Mundlak et al. 2002). Input prices are also highly correlated with price. It is therefore likely that much of the contribution of prices is channelled through inputs, and it is the net direct effect of price that is weak. The main outstanding result in Tables 11.12 and 11.13 is the very high elasticity for land, particularlyirrigated land. The sum elasticities of irrigated and rain-fed land are 0.7, meaning that the two categories of land account for roughly 70 per cent of output. A possible explanation for this is that common shocks affect output and irrigated land. To test this hypothesis and overcome its consequences, we estimated the average irrigated land productivity function where output and inputs were expressed as ratios to Table 11.13 Productionfunction: Indonesia, basemodel Model A 1971-98 Model B 1980-98 b PC t PC b PC t PC Normalised Constant -6.344 -6.3 Irrigated land 0.583 27.4 Rain-fed land 0.080 1.1 Fertiliser 0.066 16.4 Capital 0.035 40.0 Labour 0.227 59.2 Sum 0.990 Wholesale price ratio 0.057 1.7 Price spread 0.069 0.5 No schooling -0.003 -55.0 Roads 0.084 43.6 Infant mortality -0.002 -39.3 Rank 4 DW 1.883 R2 0.996 Source: Authors' calculations. DETERMINANTS OF AGRICULTURAL GROWTH 501 irrigated land. In this equation output and inputs (in logarithms) are expressed as differences from irrigated land, thus the common shocks are likely to disappear (Table 11.14). The table presents two regressions, without and with prices.The estimates in the 'irrigated land' row are values needed to bring the sum input elasticities to one. In both regressions, the sum land elasticities are roughly 0.5. This reduction is consistent with the above hypothesis. The reduction in land elasticities is compensated by an increase in labour elasticity. The correlation coefficient of labour with irrigated land and with capital is high, which may cause the variability in the estimates (also reported in Mundlak et al. 2002).Another strikingdifference from the results for the other two countries is the low capital elasticity. The fertiliser elasticity is 0.05, which is in line with the value discussed for the Philippines and Thailand. The sign of price elasticity is positive and negative for price-spread. Shadow prices The magnitudeof new opportunitiesfrom the Green Revolutionin Indonesia is illustrated by the change in paddy yield from 1.76 metric tonneslha in 1965 to about 4.5 metric tonneslha in 1996 (Kawagoe, Chapter 6, this Table 11.14 Production function: Indonesia, 1971-98, alternative specifications for land Model C Model D b PC t PC b PC t PC Constant Irrigated land" Rain-fed land Fertiliser Capital Labour Price Price spread No schooling Roads Infant mortality Rank DW R2 'Calculatedfromhomogeneityconstraint. Source: Authors' calculations. 502 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH volume, Table 6.5). This change is reflected in the rise in marginal productivity of irrigated land, as well other factors (Figures 11.5-11.9, Table 11.15). There was a continuous dramatic increase in the marginal productivity of irrigated land, measured in 1993 dollars per hectare, from $1,200 in 1961 to nearly $3,000 at the end of the 1990s. This is high relative to the level in the other countries, but no less impressive is the fact that growth continued relentlessly at a high pace. The new technology used fertiliser intensively, and its introduction generated a big jump in demand. Instead of importing fertiliser to meet this new demand, Indonesia relied on home production, which was far from adequate. As shown by Kawagoe (Chapter 6, this volume, Table 6.9), fertiliser production started to gain momentum in the late 1970s, but it was not until around 1985 that production reached one-half its 1995 level.16 Consequently, excess demand arose, reflected in a gap between the marginal productivity of fertiliser and its official price used in the national accounts Table11.15 Indonesia: productivity,prices and shadow prices Productivity (1993 US$) Elasticity Avg. Marginal Avg. Marginal Avg. Marginal Irrigated land ($/ha) 0.457 5,004 2,288 3,903 1,785 5,644 2,581 Rain-fed land ($/ha) 0.230 602 138 460 106 686 158 Fertiliser ($/mt) 0.084 17,793 1,493 27,591 2,314 11,544 968 Capital (rate of return) 0.031 3.07 0.09 5.01 0.15 1.91 0.06 Labour ($/year) 0.198 544 108 474 94 586 116 Reported prices (1993 US$) Wage rate ($/year) Fertiliser price ($/mt) Fertiliser, distortion rate Marginal rates of substitution Irrigated for rain-fed land Irrigated land for labour Irrigated land for wages Irrigated land for labour adjusted Irrigated land for capital Irrigated land for capital adjusted Derived prices (1993 US$) Irrigated land ($/ha) Irrigated land-capital base ($/ha) Source: Authors' calculations. DETERMINANTS OF AGRICULTURAL GROWTH 503 to compute value added. This was reflected in actual domestic prices, and as a result fertiliser had to be heavily subsidised (Kawagoe, Chapter 6, this volume, Figure 6.12). The gap, measured in 1993 dollars per metric tonne, was very high in the early 1960s,at the beginning of the Green Revolution, and even rose to a ~ e a kexceeding $9,000 in 1965 (Figure 11.5). From then on, it declined, falling to about $1,000 in the mid 1980s and remaining there for the rest of the study period. The distortion rate declined from a value of 4 in 1971 to about 1.5 in 1998. It can only be expected that under such a situation there were opportunities to gain from privileges granted under various government programs. There is no question that reliance on home production of fertiliser to meet the new demand was very costly in terms of agricultural output and farmer income. The marginal productivity of rain-fed land also increased over the same period by a factor of 3 (Figure 11.9), but its level was only around 6-7 per cent of that of irrigated land (cf. Figure 11.6). This rise is due to the improvement and expansion of non-rice food crops and export crops, as described by Kawagoe (Chapter 6, this volume). The new technology was capital-intensive at the farm level as well as in terms of infrastructure requirements. We detected the importance of the infrastructure-physical and human-in our estimates of the production hnaion. Initially, the capital level was low and the rate of return was very high. It was only as late as 1985 that the ratio of investment to agricultural output started to rise above the 5 per cent level (Figure 11.13). The rise in this share can be seen as a response to the high rate of return. This rise in investment led to a subsequent decline in the rate of return and an increase in the rent of land. Public programs such as Bimm financed some of the capital flowing to agriculture. Whether knowingly or not, the same forces that determined the high shadow interest rate might have also affected these programs. In any case, the flow was rationed and costly to obtain. Plantationsalso sometimes benefited fiom subsidised credit. In reality, then, some investments were made with subsidised credit, which may bias downward our calculation of the shadow rate of return. The changes in technology are clearly reflected in the shadow prices of land, which are obtained by capitalising the shadow rent (Figures 11.6 and 11.9) by a discounting factor of 15 per cent. Using 1993 dollars, the value of rain-fed land increased fiom roughly $44O/ha in 1961 to over $1,300 in 1998. At the same time, the shadow price of irrigated land increased fiom $7,80O/ha to nearly $2O,OOO/ha. 504 RURAIJ DEVELOPMENT AND AGRICULTURAL GROWTH Figure11.13 Ratioof approved investmentto agricultural GDPin Indonesia 0.25 , Ratio Source:See Mundlak, Y.; Larson, D.F. and Buner, R., 2002. DeterminantsofAgrrgrrculturd Growth in Indanesia, the Philppines, and ThaiLznd,Policy ResearchWorking Paper No. 2803, World Bank, Washingon, DC. for discussion of data sources. Growth accounting The growth accounting presented in Table 11.5 shows that in the period 1971-98 factor accumulation accounted for 56 per cent of total growth, leaving 44 per cent for changes in TFP Because the various alternative regressions differ in estimated elasticities,it is desirable to check how sensitive growth accounting is to the choice of regression. Table 11.16 presents calculations for two alternative elasticities taken from Tables 11.13 and 11.14. The main difference between them is the order of magnitude of the land and labour elasticities. The results are fairly similar, with factor accumulation accounting for 59 per cent and 55 per cent of total output gowth. The contribution of state variables practically exhaustsTFP growth, supporting the conclusion that state variables serve well as carriers of technology shocks. The relative contribution to output was about 12-13 per cent for each of the three public goods-schooling, roads, and health- and 5 per cent for prices. The contribution of the price spread was negligible. DETERMINANTS OF AGRICULTURAL GROWTH 505 Table 11.16 Sourcesof growthfor Indonesia Model A Model B (1980-98) Model D 1971-98 (1971-98)" normalised' (1971-98)b Change Share of Share of Share of (%/year) Elasticity growth Elasticity !growth Elasticity growth (Yo) (%I ("/.I Output 3.39 Inputs Irrigated land 0.80 0.583 13.8 0.457 10.9 0.524 12.4 Rain-fedland 0.52 0.080 1.2 0.230 3.5 -0.003 -0.0 Fertiliser 8.18 0.066 15.8 0.084 20.3 0.050 12.0 Capital 11.59 0.035 11.9 0.031 10.5 0.038 13.0 Labour 1.88 0.227 12.6 0.198 11.0 0.391 21.8 ' State variables Price 1.36 0.057 2.3 0.127 5.1 0.123 4.9 Pricespread 0.10 0.069 0.2 0.161 0.5 -0.198 -0.G Noschooling -1.30 -0.003 13.1 -0.003 11.5 -0.003 12.3 Roads 5.71 0.084 14.2 0.073 12.4 0.073 12.4 Infantmortality -2.79 -0.002 14.8 -0.002 13.2 -0.001 11.5 Factor accumulation Total factor productivity State variables Portion of TFP due to state variables " Models A and B are from Table 11.13. Model D is from Table 11.14. Source: Authors' calculations. Discussion Indonesian agriculture is dichotomousthere are both many smallholders and big plantations.The backbone of small farms is rice farming, but there are smallholders of tree crops as well. On the whole, rice farms are small and do not provide U employment for the fsmily, which forces families to seek off-farm work. The ease of finding such opportunities varies geographically. Off-farm employment is relatively easy to find in Bali, a small island with a highly developed tourist industry, where some complain that tourism competes with agriculture. In other areas, limited on-farm and non-farm opportunities result in relatively low wages. Rural poverty is 506 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH of concern, and has influenced policies aimed at improving the profitability of rice production. Because rice provides only a fraction of family income, however, this measure by itself is not an efficient way to overcome rural poverty. In the long run, if and when labour demand from non-agriculture expands, labour will leave agriculture. When farmers are asked about career priorities for their children, the universalanswer is non-agricultureif possible. This is revealed by the fast and pervasive growth of schooling. Farmers are poor, but is that an indication of the w&e of agriculture?The answer is no if we look at land prices. The ratio of land price to wage rate is very high by internationalstandards. This can be easily seen by dividing the price of land by the wage rate in other countries, specificallyin the big grain- exportingcountries.Thisisalso reflectedin thefactorshareofland in contractual arrangements-50 per cent of rice in crop sharing. In part, agriculture is profitable because labour costs are low. For this reason, we can think of the labour cost of land as a good indicator of the hture evolution of the sector: it will decline when other employment alternatives develop and wages rise. The question still looms-why are land prices so high? And, not independently, why are farmers willing to tie their capital to land?Again, the answer is lack of alternatives. If a farmer sells his 0.2 hectares of land, what can he do with the money? Move to the city and seek shelter and work, with all the risks involved?A better strategy is for family members to shift to non-farm work in the city while maintaining family roots in the village. Eventually the family may sell out of the village, but only after establishing roots in the city. This is the reason only a small fraction of the labour force leaves agriculture in any given year, a universal finding. Farming provides shelter and community, as reflected in the phrase 'farming is a way of life'. This attribute is built into land prices, and is applicable more to established farms than to frontier land or land for planting trees such as palm, coffee, cacao and rubber. In the case of fiontier land, labour is scarce and the pace of development is determined by labour supply. This implies a shadow price for labour higher than the wage rate in labour-scarce areas, and consequently lower land prices. Also, land price is strongly affected by proximity to roads and cities, reflecting lower transportation costs, but more so proximity to employment opportunities. Non-rice agriculture The treecrop sector has expanded rapidlybecause it is profitable.It responded favourablyto changes in the real exchangerate,generatinga boom in products priced by the world market such as cacao and palm oil. The sector includes DETERMINANTS OF AGRICULTURAL GROWTH 507 smallholders as well as big plantations run by corporations; some big plantations are owned by the government. This coexistence of small and big holdings raises the interesting question of economies of scale. All major tree crops (oil palm, rubber, coffee, cacao, coconut and tea) are harvested by hand and require continuous harvesting at frequent intervals almost year- round. Because harvest labour is such an important cost, the scope for scale economy is limited. Furthermore, the difficulty of monitoring large groups of hired harvest labour produces negative economies, which plantations try to minimise by innovations in their approach to labour management. Positive economiesof scale are derived from processing plants, which are not divisible and require a continuous product supply, which is achieved by joint ownership of plantations and processing plants. Beyond this, there seem to be economies of scale in acquisition and development of new land and the finances allocated to these activities. Resource constraints Evidence suggests a serious capital constraint on agricultural development, which is consistent with a relatively high shadow value for the user cost of capital.The level of capital was relatively low in the 1960s but subsequently grew rapidly. This is consistent with the revealed high shadow price. Still, in spite of this growth, the capital-utput ratio was relatively low for most of the period, as reflected in the low factor share of capital. Cre&t markets are not well developed; in !Act,they hardly exist for long- run investment in agriculture and are also poorly developed for short-term loans. This can be judged by the barter arrangements made between fertiliser suppliers and farmers, implying a relatively high interest rate, which results in under-utilisation of fertiliser. The cacao yield of smallholder farms in Sulawesi, for example, could be increased considerably if the fertiliser rate was increased. To sum up, considerable growth can take place in agriculture with the expansion of conventional !Actors.This is consistent with past performance, where our calculations show that total factor growth contributed 55-60 per cent of agricultural growth. AGRICULTURAL PRODUCTIVITY: THE PHILIPPINES Background Accordin to Bdisacan, Fuwa and Debuque (Chapter 7, this volume), the g study ~eriodcan be divided into two distinct periods-the 1960s and 508 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH 1970s, a period of good economic and agricultural performance, and the 1980s and 1990s, a period of volatility characterised by recessions, inflation, and political instabilityand changingpolicymeasures.The differencebetween the two periods is reflected in the fact that per capita income in 1996 was roughly the same as in 1981, indicating a waste of two decades of potential growth. In the 1980s and early 1990s, production growth rates for virtually all crops decelerated. This deceleration can be attributed to a decline in the expansion of cultivated area, a drop in world commodity prices, a series of natural calamities and droughts, the virtual completion of the Green Revolution in the early 1980s, and policy-related factors, including uncertainty about the Comprehensive Agrarian Reform Program (CARP), and a sharp decline in public investments in agriculture. Most growth in rice production is due to yield increases, which tapered off in 1980-97. This decline can be attributed to a drop in the world price of rice, stagnation in public investment in irrigation, exhaustion of the productivity potential of modern rice varieties, and soil erosion. From the mid 1960s to the early 1990s,the share of rice area harvested under irrigation expanded at 2.6 per cent per annum. Irrigated rice grew from 33 per cent of rice area in 1965 to 61 per cent at the start of the 1990s. In contrast to the weak performance of crop agriculture in the second period, poultry and other livestock showed a robust gowth of 5-6 per cent per annum. Estimation Table 11.17 presents the production function estimates with inputs alone. Irrigated land is represented here as a ratio to total land. The OLS estimates do not make sense. The signs of the principal components results are fine, but not the magnitude, and the DW statistics are low. For further reference, the last column shows normalised values of the elasticities derived by the principal components estimates. The elasticity of the share of irrigated land (0.092) is derived at the average value of this share (0.117). We explain this procedure below. Table 11.18 presents the principal components estimates with the state variables retained in the analysis. The addition of state variables does not significantly improve the DW statistics, and the estimate for the labour coefficient has the wrong sign. To overcome the low DW statistics, we compute autoregressive regression, from which we obtain the autoregressive coefficient (p) of 0.43. We then use this estimate to filter the variables by [~(t)- px(i-I)], and rerun the regression (Table 11.19). The coefficients of the principal components regression all have the right sign. This result is DETERMINANTS OF AGRICULTURAL GROWTH 509 Table 11.17 Production function: the Philippines, 1961-98, without state variables b OLS t OLS Normalised Constant 6.722 1.9 Irrigated land ratio -1.643 -0.7 Land 0.636 2.6 Fertiliser 0.129 3.8 Capital, machines 0.120 3.2 Capital, agricultural origin Labour sum Rank D W RZ Source: Authors' calculations. Table 11.18 Production function: the Philippines, 1961-98, with state variables Constant Irrigated land ratio Land Fertiliser Capital, machines Capital, agricultural origin Labour Education Inflation Price Price spread Rank D W for O U RZ for OLS D W for PC RZ for PC Source: Authors' calculations. obtained only after the introduction of the state variables, which serve as carriers of the implemented technology. The elasticity for the irri ated land ratio (this variable is not logged) is g the product of the ratio and its coefficient. The ratio varied approximately between 0.09 at the beginning of the period and 0.14 toward the end. The 510 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH mean value is 0.117, and at the mean the elasticity is 0.26. Note that the coefficient of the ratio is partial, indicatingthe marginal impact of increasing the share of irrigated land when other variables, specifically total land, are constant.Thus the elasticityof the ratio indicates the premium of converting a unit of land to irrigated land. The sum elasticities of the other inputs is 0.695 and, adding the elasticity of irrigated land evaluated at the mean, the sum is 0.955. When the ratio equals 0.14, the sum becomes exactly I, whereas at 0.09, the sum is closer to 0.9. The sum elasticities of the two types of land is over 0.5, meaning that over one-half of the value added can be attributed to land and irrigation. The elasticity of fertiliser is around 0.07, in line with that obtained in the other countries. The elasticity of machines is 0.054, and that of capital of agricultural origin (livestock and trees) is 0.093. For the period as a whole, capital showed the fastest growth after fertiliser. The growth of capital of agricultural origin extended over the whole period, whereas that of machines almost disappeared in the period 1980-98. The elasticity of labour is 0.165, but it is difficult to judge the realism of this figure. For part of the period we have daily wages. To compute the total wage bill in agriculture on the basis of this information, it is necessary to assume the number of working days in a riculture per year. To get some g Table 11.19 Productionfunction:Philippines,1961-98, with filtered data Filtered variables, p= 0.43 b OLS t OLS b P C Constant 5.024 0.8 7.311 Irrigated land ratio -0.549 -0.2 2.212 Land 0.414 1.5 0.312 Fertiliser 0.024 1.O 0.073 Capital, machines 0.101 2.3 0.054 Capital, agricultural origin 0.351 0.9 0.093 Labour -0.093 -0.7 0.165 Education 0.197 0.3 0.207 Price spread -0.786 -2.2 -0.657 Inflation -0.137 -2.1 -0.037 Price 0.451 3.5 0.196 Rank 3 DM 2.003 1.678 R2 0.973 0.958 Source: Authors' calculations. DETERMIh7ANTS OF AGRICULTURAL GROWTH 511 idea of the share of labour in value added we assumed 150 working days per year and multiplied the daily wage by 150 and the labour figure used in the regression. The product is divided by the value added to yield the labour share (Figure 11.14), which fluctuated greatly, from a minimum of roughly 0.25 in the early 1980s to a maximum of almost 0.4 in 1998. For the share to equal the elasticity, the number of working days would have to be cut by about one-half. Including the efficiencyor schooling of labour in the estimation seemed to offer a possible solution. To do this we defined eff~cienc~labour as the cross-product of labour and education. However, running the regression with this variable, without education as a separate variable, did not help to resolve the issue of the difference between labour share and estimated elasticity. The price elasticity is positive and sizable considering the hct that this is a net direct response with inputs held constant, that is, it does not represent the effect of the price on inputs. The spread has a negative effect, as does inflation. These results are fairly robust. This is extremely interesting, not Figure11.14 Labour share of agricultural GDP in the Philippines 0.4 Per cent Source: See Mundlak, Y.; Larson, D.F. and Butzer, R, 2002. DeterminantsofAgricultural Growthin Indonesia, the Philippines,and Thailand,Policy ResearchWorking Paper No. 2803, World Bank, Washington, DC. for discussion of data sources. 512 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH only because of the theory but also because these variables represent the non-trended part of the data while output is trended upward. The results for Thailand and Indonesia are weaker. The main difference between the countries is that there was much larger price and inflation variability in the Philippines, making it possible to capture the price effect with greater precision. Although the regression we have just discussed is reasonable, it is not identical to the regression used for the other two countries in the treatment of irrigation and the filtering of the variables. To compare with Thailand and Indonesia, Table 11.20 shows two additional regressions. We present only the principal components results, but also include the D W and R2 statistics for the OLS regressions. In both regressions, the price is lagged one year and the period of analysis is reduced by one year to 1962-98. In the first regression, the log of the irrigated land ratio replaces the irrigated land ratio in order to obtain the partial elasticity of irrigated land (note that the total land is held constant). The resulting elasticity is 0.239, compared with a value of 0.26 obtained at the mean of the ratio from the value in Table 11.19. The second regression separates between irrigated and rain- fed land. The sum of the normalised elasticities of the two types of land is 0.58, which is similar to the sum of irrigated land and agricultural land obtained in the other two regressions. The main difference is that the regression with irrigated and rain-fed lands completely separated (Model B) gives a lower elasticity to the irrigated land. The elasticities of all the other variables are very dose in aU three regressions. Note that the regressions in Table 11.20 are in actual values, and this shows that the results in Table 11.19 are not a direct outcome of the filtering. The D W statistics in the OLS regressions do not flag serial correlation. On the other hand, the constraints imposed by the principal components estimator cause a low D W statistic. Nevertheless, as indicated above, the elasticities are similar to those obtained in the filtered version and therefore provide a reasonable basis for the substantive discussion. Shadow prices Paddy yields in the Philippines roughly doubled between the early 1960s and the 1990s (Kawagoe, Chapter 6, this volume, Figure 6.5), indicating that the impact of the Green Revolutionwas less dramaticthan in Indonesia. This might be relatedto the behaviour of the marginal productivityof irrigated land in the Philippin-it was lower than in Indonesia, and did not change DETERMINANTS OF AGRICULTURAL GROWTH 513 Table11.20 Production function: Philippines, 1962-98, alternative specifications Model A Model B t PC Normalised t PC Normalised Constant 46.7 28.2 Irrigated land 40.8 0.249 45.8 0.155 Land 30.8 0.345 Rain-fed land Fertiliser Capital, machines Capital, agricultural origin Labour Sum Education Price Pricespread Inflation Rank D W for OLS RZ for OLS DW for PC RZ for PC Note: 'Irrigated land' is expressed as the log of the ratio of irrigated land to total land in Model A and as hectares in Model B. Source: Authors' calculations. much over the years (Figure 11.6). Average levels are given in Table 11.21. The level is affected by the choice of elasticity (we use Model B &omTable 11.20). Had we used the value in Table 11.19, the level would have been higher but still below that of Indonesia, and the time pattern would have remained the same. On the other hand, the marginal productivity of rain-fed land is higher than in the other countries and also showed the fastest growth (Figure 11.9). Consequently, the ratio of marginal productivity of irrigated to rain-fed land declined from 3.7 in 1961 to 2.3 in 1998. As for Thailand, the decline is likely to reflect an increase in the productivity of rain-fed land caused by the introduction of new crops and overall practices. On the other hand, the expansion of irrigated land might have caused a decline in its productivity on the margin because new land brought under irrigation may be of lower quality than existing irrigated land. Similarly, expansion due to the increase in irrigated land may lead to lower-value crops. 514 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Table 11.21 Philippines: productivity, prices and shadow prices 1961-98 1961-80 1980-98 Productivity (1993 US$) Elasticity Avg.Marginal Avg. Marginal Avg. Marginal Irrigated land ($/ha) 0.155 6,448 1,001 6,617 1,027 6,292 977 Rain-fed land ($/ha) 0.425 856 363 749 318 970 412 Fertiliser ($/mt) 0.077 10,985 842 13,556 1,040 8,238 632 Capital (rate of return) 0.101 1.53 0.15 1.72 0.17 1.34 0.13 Labour ($/year) 0.181 883 160 867 157 905 164 Machines (rate of return) 0.062 92.02 5.72 120.18 7.47 60.74 3.78 Reported prices (1993 US$) Wage rate ($/year) Fertiliser price ($/mt) Fertiliser, distortion rate Marginal rates of substitution Irrigated for rain-fed land Irrigated land for labour Irrigated land for wages Irrigated land for labour adjusted Irrigated land for capital Irrigated land for capital adjusted Derived prices (1993 US$) Irrigated land ($/ha) Irrigated land-capital base ($/ha) Note: The data distinguish between capital of agricultural origin (livestockand orchards) and machines. In this table, 'capital' is referring to capital of agricultural origin. Source: Authors' calculations. The gap between the marginal productivity of fertiliser and its official price, or the distortion, was high in the early 1960s and declined gradually thereafter (Figure 11.5). The distortion is lower than in Indonesia but higher than in Thailand from 1971 on. Distortion is related to excess demand evaluated at official prices. The decline in distortion is probably the result of a high growth rate of fertiliser use, which considerably exceeded the growth rates of other inputs.The pattern of fertiliser use would have looked completely different under a perfectly elastic supply of fedser throughout the sample ~eriod.An inspection of the relationship between the real price of fertiliser and the price of maize or rice supports this view. The behaviour of this price ratio is shown in Figure 11.15. The fertiliser is ammonium sulphate, the maize is white maize, and the rice is the special variety. It DETERMINANTS OF AGRICULTURAL GROWTH 515 Figure11.15 Price ratioof fertiliser tocropsin thePhilippines Source: See Mundlak,Y; Larson, D.F. and Butzer, R., 2002. Detemimnts ofAgprgPrml~ral Grmutbin Indonesia, the Philippines,and Thailand,Policy Research Working Paper No. 2803, World Bank,Washington, DC. for discussion of data sources. turns out that while prices fluctuated, the supply continuously moved upward. Furthermore, fertiliser use climbed as the real price rose. A logarithmic regression of fertiliser use on the real price of fertiliser (maize as a numeraire) gives a positive elasticity of 0.5 with an R2 of 0.78. This result is consistent with a continuous excess demand for fertiliser. The decline in the price of fertiliser in the 1980s resulted in a rise in the distortion rate. The marginal productivity of capital of agricultural origin was in the 0.16-0.18 range in 1961-81, then it gradually declined to 0.11 in 1998. Assuming a depreciation rate of 5 per cent, the shadow value of the interest rate declined from 13 per cent to 6 per cent. This decline may be attributed to a continuous rise in the capital stock at a high rate and a deterioration in the performance of a riculture after 1981. g A similar calculation for the user cost of capital in machines gives a very high value, which does not make sense. The problem can be detected by 516 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH computing the marginal productivity of the two kinds of capital, which indicates that the marginal productivity of a peso invested in machines was 29 times higher than that invested in capital of agricultural origin. We currently have no explanation for this result. Growth accounting The calculations in Table 11.5 of the TFP are based on Model B estimates in Table 11.20. For the period 1961-98, total factor growth accounted for 90 per cent of growth in output, leaving 10 per cent for TFP. The contribution of the individual inputs is similar at around 15-16 per cent each except for machines, which contributed about 11 per cent. The contribution of state variables exceeded the growth of TFC mainly because of schooling in the latter period. A broader interpretation can attribute some of this impact to changes in physical infrastructure, such as electricity and roads, which were highly correlated with education and were not supported by the regression. The price variables contribute to annual variability in output, but because their average gowth rate is nearly zero, their direct contribution in the period as a whole is basically negligible. A different picture is obtained when the exercise is conducted for the two sub-periods. The growth rate of output in the first sub-period is 3.8 per cent, of which 74per cent was accounted for by total factor growth and 26 per cent by TFP gowth. The main contributors to growth were rain- fed land (16 per cent), fertiliser (15 per cent), and irrigated land (13 per cent). The contribution of other inputs was in the neighbourhood of 10 per cent. In the second sub-period, output grew at an average rate of only 1.4 per cent, 91 per cent of which was due to total factor growth. The contribution came from almost all the inputs, with the exception of machines. Clearly the second sub-period is inferior in growth of total factor and output. There is a considerable difference between the two periods in the effect of the state variables, ~articularl~price. In the first sub-period, the favourable price accounted for about 15 per cent of growth in output, more than the 8 per cent of which was due to schooling. In addition, price volatility had a negative effect, which amounted to 2.6 per cent of output growth. In the second sub-period, the declining price subtracted around 16 per cent of the output, thus neutralising the positive effect of price in the first sub- period. This is the reason that for the period as a whole the price effect was negligible. This demonstrates the potential positive direct effect on output DETERMINANTS OF AGRICULTURAL GROWTH 517 of a favourable price environment. In addition, there is the indirect contribution made through the choice of durable inputs as well as the ratchet effect of the technique choice. The absolute effect of education was similar throughout, but its relative effect dominated the second sub-period because of low TFP growth. As a result, the state variables overestimateTFP g o d in the second period, while they underestimateit in the first period. This indicates that education alone only poorly represents the role of public goods as carriers of implemented technology. The regression did not sustain more variables of this group. The most striking result is the decline of the TFP growth rate from 0.98 per cent in the first period to only 0.13 per cent in the second one. The crucial questions concern the sensitivity of the results to the choice of elasticities used for the calculations. To answer this, we present in Table 11.22 results based on elasticities from Model A in Table 1120. The main Table 11.22 Philippines: sources of growth using alternative parameters Elasticity Change (%/year) Share of growth (%) 1961 1961 1980 1961 1961 1980 -98 -80 -98 -98 -80 -98 Output 2.55 3.82 1.38 Inputs Irrigated land share 0.249 1.44 1.60 0.84 14.1 10.4 15.1 Land 0.345 1.19 1.60 0.31 16.1 14.5 7.7 Fertiliser 0.076 5.36 7.35 4.90 15.9 14.6 26.9 Capital, agricultural origin 0.097 3.75 3.47 3.35 14.2 8.8 23.5 Labour 0.174 2.17 2.30 1.50 14.7 10.5 18.9 Capital, machines 0.059 4.55 6.64 0.28 10.5 10.3 1.2 State variables Price 0.354 0.00 1.75 -0.70 0.0 16.2 -17.9 Price spread -0.862 0.03 0.14 -0.01 -1.0 -3.2 0.6 Inflation rate -0.104 0.11 0.67 -0.64 -0.5 -1.8 4.8 Education 0.215 1.65 1.46 1.60 13.9 8.2 24.9 Factor accumulation 2.19 2.64 1.10 85.6 69.1 93.3 Total factor productivity 0.37 1.18 0.28 14.4 30.9 6.7 State variables 12.4 19.4 12.4 Portion of TFP due ro state variables 0.9 0.6 1.8 Source: Authors' calculations. 518 RURAL. DEVELOPMENT AND AGRICULTURAL. GROWTH difference is in the weights of rain-fed and irrigated land. The elasticities from Model A give more weight to irrigated land and less to rain-fed land. The main difference in the overall results is that TFP now accounts for 14.4 per cent, rather than 10 per cent, of output growth in the period as a whole, 31 per cent rather than 26 per cent in the first sub-period, and 6.7 per cent rather than 9.1 per cent in the second sub-period. The overall picture-specifically, the contribution to the slowdown of growth in the second period-remains the same. There are several possible reasons for this slowdown.To a large degree, this might be a reflection of the declinin g impact of new varieties, and the declining profitability as measured by real agricultural price. Because new varieties are intensive users of fertiliser and irrigation, these inputs immediately became scarce, in the sense that their shadow price exceeded the quoted market price. The inputs' scarcity generated an increase in their supply, so they expanded much faster in the first period than in the second. Fertiliser continued to grow at a fast rate in the second period, but less rapidly than in the first period. Exceptions were capital of agricultural origin and labour. The rise of capital of agricultural origin reflects a continuous rise in the demand for livestock products. The rise in the labour force reflectsa growth in thelabourforceassociatedwith the rate of population growth, and the hilure of the non-agricultural sector to absorb the growing labour supply. If this is indeed the explanation, the question is, then, why does TFP account for only 26 per cent (Table 11.5) or 31 per cent (Table 11.22) of output growth in the first period?The answer is that part of the impact of the technical change generated by the new varieties is reflected in the shadow price of scarce inputs, and this contributes to growth of total factor and not of total factor productivity. Other possible effects Demand and export. In a closed economy, output is affected by domestic demand. In an open economy, the world market constitutes another demand component. A plot of the export share in per capita production shows that early in the study period exports reached high values. Later on, this ratio shows a declining trend, converging to 10 per cent. To trace the behaviour of domestic demand, we ran a regression of per capita output in agriculture on total per capita output expressed in logs. The estimated elasticity is 0.6. Allowance for export had a negligible effect. This elasticity is the pseudo-income elasticity. This equation summarises DETERMINANTS OF AGRICULTURAL GROWTH 519 the population and income effect on demand for the agricultural product. Adding the price ratio to the regression had little effect. Food is not identical with the agricultural product, though it accounts for most of it. Thus, the elasticity is a rough estimate for the income elasticity for food. The point is that if production is oriented largely to the domestic market, then demand has an effect on production. Land tenancy. Census data point to two pertinent developments in the farm operation that may affect productivity (Figure 11.16). It appears that the proportion of farms operated by non-owners (tenanted or leased) declined during the study period fiom over 35 per cent to about 15 per cent. At the same time, there was a considerable decline in farm size. It is basically impossible to measure the impact of these changes within the empirical framework used above with any reasonable precision. Tenanted farms are sometimes believed to be less efficient, but if this is the case, productivity should have been higher in the second sub-period. This is not the case. The impact of farm size might be a partial explanation. Since Figure 11.16 Small farms and leased farms in the Philippines 40 , Per cent 30 Share of farms tenanted or leased 20 Source:See Mundlak,Y.; Lason, D.F. and Butzer, R., 2002. Determinants ofAgn'mltura1 Growth in Indonesia, the Philippines,and Thaihnd, Policy Research Working Paper No. 2803, World Bank, Washington, DC. for discussion of data sources. 520 RURAL DEVELOPMENT AND AGRICULIXRAL GROWTH much of the production takes place on small farms that are subject to increasing returns to scale, the increase in production concentration on small farms might have contributed to declining productivity. ACKNOWLEDGMENT The authors would like to Takarnasa Akiyama, William Cuddihy, Nobuhiko Fuwa, Yujiro Hayami, Masayoshi Honma,Toshihiko Kawagoe, and Shinichi Shigetomi for comments on earlier drafts. NOTES The actual period analysed was determined by data availability. For a more detailed description of data and sources, see Mundlak et al. (2002). For a discussion of country differences in physical environment and political histow see Hayami (Chapter 2, this volume); for a comparison of trade protection rates see Akiyama and Kajisa (Chapter 9, this volume). Rain-fed land is calculated as the difference between agricultural land and irrigated land. The coverageof fixed capital data is not well defined. For some comments on this subject, see Larson et al. (2000). The value data are re orted in local currency in constant prices1985 for the Philippines, p 1988 for Thailand, and 1993 for Indonesia. They are converted to US dollars using the archange rates for theseyears: 18.607,25.34, and 2087 for the three countries, respectively. The result is then adjusted to 1993 values using the US GDP deflator:1985 = 0.784, 1988 = 0.853, and 1993 = 1.00. All references to dollars in this paper are US dollars. A good example is the use of mechanical threshers that make possible a third crop for rice in some areas of the Philippines (Cuddihy 2002, personal communication). ' Nominal wage rates were deflated by the consumer price index to obtain real wage rates. These were converted to 1993 dollars following the procedure described in endnote 5. For Indonesia we deflated the nominal wages by the GDP deflator. It iswell recognisedthat rural householdsoften diversifytheir labouramongseveralactivities, some of which are off-farm. See Lanjouw and Lanjouw (1995). l oUsinghouseholdsurveydatafrom the Bicolregionof the Philippines,Larsonand Plessmann (2002) estimate an elasticity of 0.09 for fertiliser and find the estimate robust under alternative model specifications. 'This is considered here to be the main reason, but there may be others, such as a difference between the price of fertiliser used in the national accounts and [he farm gate cost. I ZSee Buaer, Mundlak and Larson (Chapter 12, this volume). l 3Note on the calculations in Table 11.5. Let E represent elasticities and g growth rates and the subscripts i, s, and y represent respectively, inputs, state variables and output. Then DETERMINANTS OF AGRICULTURAL GROWTH 521 factor accumulation,in percentageterms, is given by Gi=Z,&gj, where inputelasticitieshave been normalised so that ,Z.sj=l.Growth in state variables is given by G,=?&g. Shares of gowth due to factor accumulation, Pi, and due to changes in the state variables, q,are given byPi=Gi/gy,and J',=G,/& Theshareof output growthdueto totalfactorproductivity (TFP)is I-Pi, and the portion of productivitydue to changes in the state variables (SP) is P/TFP.Growth rates used in the calculationsare mostly obtained from trend regressions of the type 1nx=c+bxtime,where g=b*100. Exceptions are made for variables already represented as ratios: inflation, price spread, no schoolingand infant mortality, where the regression x=c+b%me is used. '*SeeWarr (2000)for actual dataon povertyratesin Indonesia,Thailandand the Philippines. ' The consumer price index was used in the calculation of these variables. ' Other countriesalso relied on domesticproduction.McGuirkand Mundlak(1991)discuss this issue for India. REFERENCES Coxhead, I. and Plangpraphan, J., 1998. 'Thailand's economic boom and but, and the hte of agriculture', TDRI Quarterly Review, 13(2):15-24. Gardner, B., 2002. American Apiculture in the Twentieth Century: how it flourished and what it cost, Harvard University Press, Cambridge, Massachusetts. Lanjouw, J. and Lanjouw, I?, 1995. Rural Non-fdm Employment: a survey, Policy Research Working Paper No. 1463, World Bank, Washington, DC. Larson, D.F. and Mundlak, Y., 1997. 'On the intersectoral migration of agricultural labour', Economic Development and Cultural Change, 45(2):295-319. Larson, D.F. and Plessmann, F., 2002. Do fdrmers choose to be in@cient? Evidencefiom Bicol, Philippines, Policy Research Working Paper 2782, World Bank, Washington, DC. Larson, D.F., Butzer, R, Mundlak, Y. and Crego, A., 2000. 'A cross-country database for sector investmentand capital', World Bank Economic Review, 14(2):371-91. Larson, D.F., Butzer, R, Mundlak, Y. and Plessmann, E, 2002. Do Famm Choose to be In@cient! Evidencefiom Bicol, Philippines, Policy Research Working Paper 2782, World Bank, Washington, DC. Martin, M. and Warr, PG., 1993. 'Explaining agriculture's relative decline: a supply side analysis for Indonesia', World Bank Economic Review, 7(3):381-401. 522 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH McGuirk, A. and Mundlak, Y., 1991. Constraints, Incentives, and the Tranrfomzationof Punjab Agriculture: 1960-1980, Research Report No. 87, International Food Policy Research Institute, Washington, DC. Mundlak, Y., 1981. 'On the concept of non-significant functions and its implications for regression analysis', Journal of Econometvics, 16(1):139- 49. --,1988. 'Endogenous technology and the measurement of productivity', in S.M. Capalbo and J.M. Antle (eds), Agricultural Productivity: mea-surement and explanation, Resources for the Future, Washington, DC:316-31. -, 1993. Land expansion and land augmentation with application to Thailand, International Food Policy Research Institute, Washington, DC (unpublished). -, 2000. Agriculture and Economic Growth: theoy and mea-surement, Harvard University Press, Cambridge. -, 2001. 'Production and supply', in B. Gardner and G. Rausser (eds), Handbook of Agricultural Economics, vol. 1, North Holland, Amsterdam:3-85. -, Larson, D.F. and Butzer, R, 2002. Determinants ofAgfmltural Growth in Indonesia, the Philippines, and Thailand, Policy Research Working Paper No. 2803, World Bank, Washington, DC. Punyasavatust,C. and Coxhead, I., 2001. On the decline of Agriculture in developing countries: a reinterpretation of the evidence, University of Wisconsin, Madison (unpublished). Siamwalla, A., 1996. 'Thai agriculture: from engine of growth to sunset statues', TDRI Quarterly Review, 11(4):3-10. Warr, I?,2000. 'Poverty incidence and economic growth in Southeast Asia', Journal of Asian Economics, 11(4):43141. INTERSECTORAL MIGRATION IN SOUTHEAST ASIA-EVIDENCE FROM INDONESIA, THAILAND AND THE PHILIPPINES Rita Butzer,Yair Mundlakand Donald F. Earson Beginning with Lewis (1954), the flow of labour resources from agriculture has been an important consideration in models of economic development, and well-formulated models have been available to guide empirical measurement of sectoral migration since Todaro (1969). In practice, rates of migration and rates of natural population gowth tend to be similar, thus the accumulated effects of migration only occur over decades. Long time-series measurements of sectoral migration pose special difficulties for researchers, however, since certain explanatory variables are trended and intercorrelated. Omitting correlated variables runs the risk of specification error, a problem formulated by Theil (1957). In this chapter, we describe the flow of labour resources from agriculture to other sectors of the economy in Indonesia, Thailand and the Philippines over three decades, and examine bctors that determine the rate of such migration. To mitigate the problems associated with intercorrelated variables, we employ principal components, using the algorithm given in Mundlak (1981), which imposes parameter restrictions as a substitute for the elimination of specific variables. The threestudy countries are geographicallyclose and share similar climate and other characteristics; their economic growth experiences, however, have diiered. In Chapter 11 (Mundlak et al., this volume), we discuss sources of growth and productivity in the three countries, where economic growth 524 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH and growth in agricultural income have been associated with out-migration of labour from agriculture. As with growth, however, country experiences differ in key ways. Figure 12.1 shows the share of agriculture in total employment in each of the three countries. In the 1960s and 1970s,shares were very high (over 50 per cent for Indonesia and the Philippines and over 70 per cent for Thailand); gradually, however, they fell. Nonetheless, even in the late 1990s, the agricultural sector still employed nearly 40 per cent of workers in the Philippines and Indonesia and 50 per cent in Thailand. In developed economies, these shares level off at 2-3 per cent (Larson and Mundlak 1997). While the shares in Indonesia, Thailand and the Philippines have decreased steadily over the past three decades, the pace of the reallocation of labour resources from agriculture to other sectors of the economy has been relatively slow, with indications of acceleration occurring only in the past decade. We examine the process of the reallocation of labour below. Figure12.1 Agriculture's share of totalemployment 1 90 Per cent of total errphyrrsnt Source: Authors' calculations. See appendix for a discussion of data sources. INTERSECTORAL IMIGRATION IN SOUTHFAST ASIA MIGRATION FROM AGRICULTURE Framework Change in the sectoral composition of the labour force can be regarded as the outcome of migration from agriculture conditioned by births and mortality rates. Below, we describe the analysis of this process, using the framework created by Larson and Mundlak (1997) and Mundlak (1979), who also provide a review of the literature relevant to the approach we take here, as summarised below. Intersectoral labour allocation is analysed within the framework of occupational choice. The postulation is that individuals tend to maximise their remaining-life utility by, among other things, choosing their occupation from a feasible set of occupations, as determined by the intersecting trajectories of income and uncertainty. Here, we concentrate on the binary choice between agricultural and non-agricultural work. The two sectors are not homogeneous; in most of this discussion, however, we ignore this fact. Because non-agricultural employment often requires migration to other areas, it involves other costs in addition to those attached to change of occupation alone. Data on off-farm migration in Indonesia, Thailand and the Philippines are not available, thus we must infer them from changes in intersectoral allocations of labour. We assume that, without migration, agricultural labour would grow at the same rate as total labour, thus we attribute deviations from this rate to migration. Migration was calculated as where Mis the number of migrants (outflow of workers from agriculture to non-agriculture), LA is agricultural labour, n is the rate of growth of the total labour force, and m is the ratio of migrants to agricultural labour. In practice, we calculate n from nr=(Lt-LC-J/L+,,where L is total labour. Similarly, we define the growth rate of the agricultural labour force as nAr=(LAt-LA+,)/LAt-,, then m=n-nA. Thus, the rate of migration out of agriculture is the difference between the growth rates of the total labour force and a ricultural labour. g Ideally, we would like to use data on labour to calculate migration rates as shown above. However, the countrysources report data not on agricultural 526 RURAL DEVELOPMENTAND AGRICULTURAL GROWTH labour, but rather on agricultural employment. Annual sectoral labour data are available from the Food and Agriculture Organization (FAO) and the World Development Indicators (WDI) of the World Bank, which are constructed from census data. Such data, however, are collected every 10 years, and the annual series are obtained by interpolating. Thus, migration rates calculated fiom these data do not actually measure the annual variation that we hope to explain. We have therefore chosen to use employment data from the country sources to calculate migration rates. This choice is the outcome of the data limitation, but it is not an ideal one. The constructed migration series are plotted in Figures 12.2-12.4. The migration rates are volatile to the point where it is difficult to discern any trends. It should be noted, however, that most of the observations are in the positive part of the graphs, which indicate positive migration over the entire period.The volatility is from two sources: first, the use of data on employment rather than data on labour brings the demand for workers (and hence shocks to that demand) into the constructed migration series; second, annual data tend to be more erratic. Nevertheless, the trend still prevails. This is seen in Table 12.1, which presents decade averages of the migration rates. The Figure 12.2 Migration from agriculture in Thailand Wr cent of employment V Source: Authors' calculations. See appendix for a discussion of data sources. INTERSECTORAZ. MIGRATION IN SOUTHMT ASIA 527 Figure12.3 Migrationfrom agriculturein thePhilippines 10 1Per cent of agricultural Source: Authors' calculations. See appendix for a discussion of data sources. Figure12.4 Migrationfromagriculturein Indonesia lo Wr cent of agricultural enploym n t 5 0 -5 -10 1 Source: Authors' calculations. See appendix for a discussion of data sources. 528 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Table 12.1 Averagemigration rates for selected periods (percent per annum) Decade averages 1960s 1970s 1980s 1990s Period of analysis Period averages Thailand 0.61 0.89 0.55 3.09 1962-99 1.32 Philippines 1.32 0.35 1.39 1.45 1962-98 1.11 Indonesia .. 1.72 0.39 2.27 1972-99 1.44 Asia 1.07 1.40 1.80 n.a. Source: Asia decade averages taken from Larson, D. and Mundlak, Y., 1997. 'On the intersectoral migration of agriculturallabour', Economic Developmentand CulturalChange, 45(2):295-319. migration rates are relativelylow by internationalstandards and are consistent with the slow change in the share of agriculture in employment discussed earlier and seen in Figure 12.1. The average migration rates are greater than 2 per cent in only two cases (Indonesiaand Thailand in the 1990s). Comparing these rates with those for Asian countries from the study by Larson and Mundlak (1997), we see that migration was somewhat slower in Indonesia, Thailand and the Philippines than in other Asian countries.' We should also note from this table the considerably lower migration rates in Thailand and Indonesia in the 1980s. We now turn to identification of the factors that determine these rates of migration. Formulation The analyticframework is summarisedin the form of the migration equation where m is defined above, RI = IJIAis the income ratio (the ratio of GDP per worker in the non-agricultural sector to that in the agricultural sector), RL = LJLAis the ratio of employment in the non-agricultural sector to that in the agricultural sector, and S represents other attributes and exogenous state variables. We isolate the income and labour ratios from the other variables for the following reasons. The income ratio is assumed to be the major incentive faced by the potential migrant. In applying the formulation to the market, we need to consider the size of the labour force in the two sectors. The IN'TERSECTORAL MIGRATION IN SOUTHEAST ASIA 529 number of migrants depends on the size of the labour force in agriculture (the origin); the larger the labour force in agriculture, the more migrants can be expected in a given economic environment. The actual number may also depend on the ease of finding work in non-agriculture. In the absence of strong unemployment in non-agriculture, it is expected that the larger the labour force in the destination, the easier it will be to get a job. The formulation assumesthat the migration rate depends onlyon thecomposition of the labour force (the ratio of labour in non-agriculture to agriculture), and not on the size of the total labour force. The variables covered by S are discussed below. Variables and data The decision to migrate or not migrate is strongly influenced by incentives, the main one being sectoral income. The basic idea is that labour, like other resources, flows fiom low to high-income occupation^.^ As such, this is a qualitative property, and to test it would require a comparison of data with and without a sectoral income gap. The data do not appear in this dichotomous form, and instead show variability. The empirical analysis, as formulated in Equation 12.2, is designed to examine the impact of such variability on the migration rate. It thus tests a much stronger proposition. Proposition: The larger the income gap between the sectors, the stronger the migration rate. In other words, the income gap affects the pace of the resource allocation. We measure the incentive by income and not by wages. When it comes to long-term decisions such as migration out of agriculture, income is thought to be a more informative measure of future prospects than wages. Wages tend to be the most important component of income, but they are not the only component.The non-wagecomponent of income (for example, rent on land and returns to capital, both physical and human) may be influential in the migration decision. In addition, in our particular case there is a basic problem with the wage data. As indicated by Mundlak et al. (2002), the published wages are daily rates. To convert them to annual figures, it is necessary to know the number of working days for both sectors, but this information is not available. The intersectoral income differential is measured by the ratio of income in non-agriculture to income in agri~ulture.~Evidence fiom cross-country studies shows that, as countries develop, the income differential decreases. 530 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH 'In middle and high income countries, the [income] ratio is almost equal to 1 and, as the data show, this statement was as true in 1950 as it is today' (Larson and Mundlak 1997). This finding summarises a long-run process whereby migration of labour from agriculture to non-agriculture will lessen the gap in productivity. As the supply of labour in agriculture decreases, the shadow price of agricultural labour will rise, leading to investment in labour-saving techniques in agriculture. Increased stocks of capital (human and physical)enhance the productivityof the labour remainingin agriculture. Our current analysis deals with annual variations, and in this sense it differs from the cross-country analysis based on period averages, which are subject to less variability. Figure 12.5 shows that the productivity in non- agriculture has been much higher than in agriculture. In part this may reflect the problem of the labour data, which may overstate labour in agriculture and understatelabour in non-agriculture (for details see Mundlak et al. 2002). We continue here to review the data, ignoring this possible problem.An increase in the income differentialoccurs when non-agricultural Figure12.5 Ratioof averageincomeperworker,non-agricultureto agriculture 0 1 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 Source: Authors' calculations. See appendix for a discussion of data sources. INTERSECTORAL MIGRATION IN SOUTHEAST ASIA 531 productivity grows at a faster rate than agricultural productivity. For most of the study period agricultural ~roductivityincreased, but at a slower pace than non-a ricultural productivity. It is important to keep in mind that g the productivity measure is an outcome of the effect of the economic environment on output and labour input. Because the economic environment varied across the three countries and over time, their income gap patterns were not uniform. In Indonesia, the difference in the sectoral incomes actually increased throughout the 1970s and 1980s, peaking in 1990 before falling to the levels of the early 1980s,in spite of consistent positive growth in agricultural GDE Agriculturalproductivityin Indonesia increased by 25 per cent between 1971 and 1980, but remained stagnant throughout the 1980s. Agricultural GDP increased, but at the same rate as agricultural employment. After 1985, non-agricultural sectors grew rapidly, widening the existing income differential. From 1990 onward the gap decreased,as agriculturalproductivity increased by over 40 per cent (due to decreases in agricultural employment), before the crisis in 1997. The sectoral income gap in Thailand is extremely high and has remained so throughout the past three decades, although in the late 1990s it began to decline steadily. Agricultural productivity in Thailand increased by 17 per cent in the 1970s and only 12 per cent in the 1980s,as the commodity boom ended and prices stagnated. After the influx of foreign direct investment in the late 1980s, the non-agricultural sector grew rapidly. From 1990 to 1999, agricultural employment decreased, and agricultural productivity increased by over 55 per cent, contributing to the decline in sectoral income differentials. In the Philippines, the income ratio increased over part of the study period (1975-84) and stayed above 2.5. Much of the movement in the income ratio was due to slow growth in non-agricultural productivity, particularly since the early-mid 1980s when the Philippines experienced a series of political and economic crises. Agricultural productivity in the Philippines has increased by 25 per cent over the past three decades, but most of this growth occurredin the 1960sas a result of the Green Revolution. In the early 1980s, agricultural GDP actually decreased. Agricultural employment increased continuously throughout the study period. Still, from 1983 onwards the gap declined. To sum up, in more recent years there has been a tendency for the income gap in the three countries to decline. This was not sufficient to affect the 532 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH decade averages of the gap as shown in Table 12.2, but it is expected that these countries will follow the development that took place in other countries and the gap will eventually decrease substantially. We can compare these trends to the trends for Asian countries described by Larson and Mundlak (1997). The slow rate of migration from agriculture to non-agriculture plays a significant role in these trends. Migration has not yet been sufficient to close the income gap between the sectors. Conversely, given that the income gaps are still rather large, it is not surprising that migration rates have not decreased in the 1990s. The vector S in Equation 12.2 consists of other variables that may affect the migration decision. We think of two groups, one related to incentives and the other including variables representing infrastructure, which may bear on the cost of migration. The latter group contains trended variables, whereas the migration rates show high variability. Thus, the onus of the explanation is on the former group. We therefore list several variables that might have affected migration, and leave it for the data to determine their relevance in our case. As for the incentives, the income ratio may not fully summarise the opportunities and their stability. We therefore consider other variables that reflect the attractiveness of the two sectors. It is worth noting at the outset that several of the variables we consider have a direct role in determining agricultural income and productivity, which we attempt to measure in Mundlak et al. (2002). For reasons given below, however, we expect these hctors to play an additional role in migration as well. A natural variable is the terms of trade of agriculture as measured by the ratio of sectoral prices, agriculture to non-agriculture! The price ratio affects Table 12.2 Decadeaverageincome ratios, per worker, non-agriculture to agriculture 1960s 1970s 1980s 1990s Thailand 9.47 9.63 10.69 10.57 Philippines 3.29 3.29 3.19 2.79 Indonesia 3.88 4.07 4.44 Asia 3.37 3.31 3.57 Source: Asia decade averages taken from Larson, D. and Mundlak,Y., 1997. 'On the intersectoral migration of agricultural labour', EconomicDevelopmentand Cultural Change, 45(2):295-319. INTERSECTORAL MIGRATION IN SOUTHEAST ASIA 533 the relative profitabilityof agriculture, thus a decline in the price is expected to encourage migration. The condition of agriculture can also be viewed in the ability of the a ricultural sector to support its expanding labour force. g For this purpose we use the ratio of agricultural population to agricultural land as a measure of agricultural density. It is expected that the more densely populated the agricultural sector, the greater the incentive to migrate out of agriculture. This can be viewed as a measure of the ~ u s hside. On the pull side, we try several measures related to the prospect of finding work in non-agriculture. Unutilised capacity in non-agriculture is a roximated by the difference between the peak of previous per capita pp output and current per capita output, divided by the peak value. If the current value exceeds the peakvalue, unutilisedcapacityis zero. An alternative measure is the growth rate of output in the non-agriculturalsector. Periods of fast growth are presumably more attractive for migration. A more direct measure is the growth rate of the non-agricultural labour force. We also tried the difference between the gowth rates of non-agricultural and agricultural output. Turning to the second group of variables, the rate of migration is also affected by the cost of migration. This cost is negatively related to rural areas' degree of integration with labour markets, which depends on the state of the physical infrastructure, such as roads and telecommunications. Education may also reduce the cost of migration, probably through the ability to obtain and digest information, thus we expect a positive effect on migration. The education variable also reflects the demand preference for skilled labour. Labour and technology are not homogeneous, and changes in technology cause demand for skilled labour to increase; this supplements the effect of education on the migration cost. Certain public health issues, such as the spread of infectious diseases, can be viewed in terms of the cost of migration and incentives to migrate. The prevalence of roads, telecommunications,education and health depends on investments in such activities. This investment originates mostly in the public sector, thus we refer to this group as the policy variables. The impact of public goods on the migration rate, however, might be ambiguous. On the one hand they reduce the migration cost, but on the other hand they increase labour productivity in agriculture, as well as in non-agriculture, and thereby might reduce the income gap. In that case, the net effect is uncertain. Moreover, the empirical scope for these variables in time-series 534 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH analysis of the migration equation is rather limited. Migration is subject to annual variations, whereas these explanatory variables are strongly trended and highly intercorrelated. Regression results The migration equation was estimated separately for each country (Tables 12.3-12.5). The time period varied depending on data availability. The explanatory variables were lagged one period and expressed in natural logs, except unutilised capacity, inflation, and variables measuring rates of growth. As we saw in earlier figures, migration rates for some years were negative, therefore the dependent variable is the migration rate (m). Recall that the migration rate was calculated from the employment data, rather than labour force data, and the resultingseries are ~olatiie.~Since some of the explanatory variables are trended and intercorrelated, the explanatory power of the regressors is not high. The regression did not sustain all the contemplated variables and the exercise amounted to a search for the relevance of the various variables in explaining the data. We report results obtained by ordinary least squares (OLS),and by principal components. As mentioned Table 12.3 Principal componentsregression resultsfor Thailand, 1962-99 Estimate t-score Estimate t-score Intercept -28.59 Income ratio 13.59 Employment ratio 1.64 Unutilised capacity -9.13 R-square 0.26 DW statistic 2.17 Statistical rank Means Migration rate Unutilised capacity Elasticities Income ratio Employment ratio Unutilised capacity Note: Elasticities were calculated at mean migration rates and means of the variables which were not expressed in natural logs. Source: Authors' calculations. INTERSECTOW MIGRATION IN SOUTHEAST ASIA 535 earlier, the latter were used to impose restrictions on the parameter as a substitute for the elimination of specific variables. Such a procedure reduces the specification error due to omission of regressors (Theil 1957).The statistical rank reported in the tables is obtained as the difference between the number of regressors and the number of restrictions imposed on the parameter^.^ Throughout, the Durbin-Watson statistics did not flag a serial correlation problem. The most importantand robust iesultis the positiveimpact of the income ratio on the migration rate. The numerical value of the income-ratio Table12.4 Principal components and ordinary least squares regression results for the Philippines, 1962-98 Estimation method Ordinary least Ordinary least Principal squares squares components Estimate t-score Estimate t-score Estimate t-score Intercept 4 . 4 5 -0.91 -18.21 -1.93 -17.20 -1.87 Income ratio 4.31 0.99 8.50 1.74 5.58 1.25 Ratio of GDP deflators -2.84 -0.77 -5.32 -1.47 4.79 -2.19 Growth rate of industrial employment Mean years of schooling Growth rate differential R-square D W statistic Statistical rank Significance level Means Migration rate Growth rate of industrial employment Growth rate differential Elasticities Income ratio Ratio of GDP deflators Growth rate of industrial employment Mean years of schooling Growth rate differential Note: Elasticities were calculated at mean migration rates and means of the variables which were not expressed in natural logs. Source: Authors' calculations. 536 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Table 12.5 Principal components and ordinary least squares regression results for Indonesia, 1972-99 Estimation method Ordinary least Ordinary least Principal squares squares components Estimate t-score Estimate t-score Estimate t-score Intercept -13.78 -1.60 -11.55 -1.64 -18.42 -5.66 Income ratio 10.22 1.73 2.88 0.45 10.80 6.17 Unutilised capacity -35.96 -2.46 -35.89 -2.69 -12.12 -2.18 Inflation rate 0.11 1.96 0.09 2.21 0.05 1.85 Mean years of schooling 3.88 1.61 0.90 2.60 Growth rate of non- agricultural employment R-square DW statistic Statistical rank 2 Means Migration rate 1.442 1.442 1.442 Unutilised capacity 0.021 0.021 0.021 Inflation rate 13.495 13.495 13.495 Growth rate of non- agricultural employment 5.170 5.170 Elasticities Income ratio 7.09 2.00 7.48 Unutilised capacity -0.52 -0.52 -0.17 Inflation rate 1.03 0.85 0.42 Mean years of schooling 2.69 0.62 Growth rate of non- agricultural employment 1.96 2.01 Note: Elasticities were calculated at mean migration rates and means of the variables which were not expressed in natural logs. Source: Authors' calculations. coefficient varied somewhat across the experiments, but on the whole the coefficient was significant. This is empirical validation of the proposition stated above and is consistent with other studies with a similar specification. The elasticity with respect to the income ratio (computed at the mean value of the migration rates) varied in the ranges of 10-12 in Thailand and 4-8 in the Philippines, and was 7 in Indone~ia.~These are high values, but it should be noted that the impact of a 10 per cent change in the migration rate on the sectoral employment is far smaller. We turn now to a review of the role of the other variables. INTERSECTORAL MIGRATION IN SOUTHEAST ASIA 537 The labour ratio was practically irrelevant with a few exceptions, one of which is shownfor Thailand.Even there, however,the importanceis marginal. Instead, what seems to be important is the absorbing capacity of non- agriculture. This is revealed by several indicators. The unutilised capacity in non-a riculture was important in both Indonesia and Thailand. For g most of the period, the measure of unutilised capacity was zero; as a result the mean value is low, even though in some years the unutilised capacity was quite high. The peak value for this variable in the study period was 0.12 in Thailand and 0.16 in Indonesia. The elasticities reported in the table are calculated for the mean values. In years of high-unutilised capacity in non-agriculture, the elasticities would be considerably higher. For the extreme values reported above for the study period, the elasticities would be ten times higher. In the Philippines, the non-agricultural sector did not develop kit enough to attract and absorb labour from the agricultural sector, an issue raised by Balisacan et al. (Chapter 7,this volume). 'To take account of this fact, we used data on the growth rate of employment in the industrial sector. This variable has a strong positive effect on migrati~n.~We did not have a similar measure for the other countries. We therefore used the growth rate of non- agriculturalemployment.This variableenters the calculation of the migration rates, therefore its effect should be interpreted with caution. Thii was done for Indonesia, where the coefficient is positive and significant. Note, however, that its introductiondid not significantlyaffect the coefficient of the income ratio, and its effect on the other coefficients is not substantive. Our interpretation is that the variable picked up some of the noise in rn coming from the use of the employment data in the calculation of the migration rates. The results for Thailand were less meaningful. We also tried the - difference between the growth rates of non-agricultural and a ricultural g output. This variable had a marginal positive effect only for the Philippines. The terms of trade for agriculture, measured by the price ratio, were relevant in the Philippines, where the variable varied greatly over the study period, unlike in Thailand and Indonesia. The sign of the coefficient is negative, meaning that the migration rate is higher in years when the terms of trade for agriculture are low. We tried a population density variable (the ratio of population to agriculturalland); the variabledid not have explanatory power. Inflation was empirically relevant in Indonesia with a positive coefficient, suggesting that the heating up of the economy was supportive of migration. The role of the inflation rate can be interpreted in various 538 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH ways, however the finding is not strong enough to justify diving into a discussion of this subject here. Finally, the public good, or policy, variables discussed above had little impact. These variables are trended, thus they had only weak correlations with the migration rate. The most pronounced effect is education in the Philippines, obtained in the principal components regression. It was possible to force some of the policy variables in other regressions using the priicipal components procedure, but at the cost of reducing the level of explanation. This means that the contribution of these variables to the explanation, conditional on the other regressors, is negligible or negative. This result may reflect the data problem, but it may also be due to the fact, discussed above, that the impact of public goods on the migration rate might be ambiguous. In a cross-countrystudy (Larson and Mundlak 1997),education showed a robust positive impact on the migration rate. However, that study did not include measures of physical infrastructure, such as road length, or health conditions, thus it is not strictly comparable to the current study. It does, however, contain year and geographic dummies that may confound someof the impact of infrastructure.Toconclude, we refrainfrom generalising our empirical results on the role of the policy variables. CONCLUSION Labour migration from agriculture is related to the dynamics of the sectoral allocation of labour. This chapter has examined the process in Indonesia, Thailand and the Philippines. The migration rates from agriculture to non- agriculture in these three countries are relatively low compared to those of other countries, thus labour surpluses have not been reallocated very quicldy to other sectors of the economy. The effect of these low migration rates on the persistence of intersectoral income differentials is obvious. Even so, migration rates have been responsive to each country's income ratio, and were also affected by the absorbing capacity of non-agriculture, as indicated by several measures. The policy variables consisting of indicators of physical and human capital had little impact on the migration rate apart from that captured by relative incomes. Unfortunately, the migration series contains considerable variability for which we have not been able to account. This variability may be due to data problems or perhaps to economy-wide shocks. This remains part of the puzzle of migration in Indonesia, Thailand and the Philippines. INTERSECTORAL MIGRKTION IN SOUTHEAST ASIA 539 NOTES Migration in Indonesia,Thailand, and the Philippines was also far slower than in Latin Americawhere rates were close to 2 per cent in the 1960s,and over 2 per cent in the 1970s and 1980s (Larson and Mundlak 1997). This is the essenceof the dual economy model of Lavis (1954) and Ranis and Fei (1961). Income is calculated as GDP (in constant prices) per worker, where non-agriculturalGDP is the hifference between total GDP and GDP in agriculture (and similarly for non- agricultural employment). The price ratio is the ratioof the GDP deflator in agricultureto that of non-agriculture.The GDP deflator is derived from the ratio of GDP in nominal prices to GDP in constant prices. The data problems are discussed in Mundlak et al. (2002). The principal components method used followed the algorithm in Mundlak (1981). The levelofsignificancewas10 percent.Astatisticalrankof3meansthat threelinearcombinations of the variables summarise the information contained in the regressors. Let m = b Inx, then the elasticityof m with respect to x is dlnm/dIm = blm.The results reported here are obtained at the mean value of m. The labour data in the Philippinesis reported for three sectors-agriculture,industty and services.Thus, industrialemployment is not the same as non-agriculturalemployment. In 1960, approximately 50 per cent more peoplewere employed in servicesthan in industry. By 1971, employment in services was double that in industry; by the late 1990s, it was nearly triple. REFERENCES Coxhead, I. and Plangpraphan, J., 1998. Thaihndj Economic Boom and Agricultural Bwt: some economic qzlestions and policy puzzles, Department of Agricultural and Applied Economics Staff Paper Series No. 419, University of Wisconsin, Madison. Larson, D. and Mundlak, Y., 1997. 'On the intersectoral migration of agricultural labour', Economic Development and Cultural Change, 45(2):295-319. Lewis, W.A., 1954. 'Economic development with unlimited supplies of labour', Manchester School of Economic and Social Studies, 22(2):139- 91. Mundlak, Y., 1979. Intersectoral Factor Mobiliq and Agricultural Growth, Research Report No. 6., International Food Policy Research Institute, Washington, DC. 540 RURAL DEVELOPMENT AND AGRICULTURALGROWTH ----, 1981.'On theconceptofnon-significantfunctionsanditsimplications for regression analysis', Journal of Econometrics, 16:13949. -, Larson, D. and Butzer, R ,2002. Detminants ofAp'irulturzd Growth in Indonesia, the Philippines, and Thaihnd, Policy Research Working Paper 2803. World Bank, Washington, DC. Nehru, V;, Swanson, E. and Dubey, A., 1993. A New Database on Human Capital Stock: sources, methodology, am! results, Policy Research'Working Paper No. 1124, World Bank, Washington, DC. Ranis, G. and Fei, J.C.H., 1961. 'A theory of economic development', American Economic Review, 51(4):533-65. Theil, H., 1957. 'Specification errors and the estimation of economic relations', Review of International Statistical Institute, 28(1):41-5 1. Todaro, M., 1969. 'A model of labor migration and urban unemployment in less developed countries', American Economic Review, 59(1):13848. APPENDIX-DATA SOURCES Thailand Gross Domestic Product (agricultural and total). The agricultural GDP series includes forestry and fishery but does not include simple agricultural processing products. National income accounts were obtained in constant and current market prices (baht) horn the Office of the National Economic and Social Development Board (NESDB). Employment (agriculturaland total). Data on employment come fiom the National Statistical Ofice, Rtport of the Labor Force Suruey. Round 2 Uuly- September) was used for 1969-83, and Round 3 (August) was used from 1984 on. No data were reported for 1970, so straight-line interpolation was used to estimate the data. For 1961-68, data obtained fi-om Coxhead and Plangpraphan (1998) were used. Philippines Gross Domestic Product (agricultural and total). The agricultural GDP series includes forestry and fishery. National accounts were obtained in constant and current market prices (pesos) fi-om the Economic and Social Statistics Office, National Statistical Coordination Board (NSCB). Employment(agricultural,industrialand total). Data on employment come from the Labor Force Survey, National Statistics Office. When available, data fiom the October survey were used. Seaoral data were not reported in INTERSECTORAL MIGRATION IN SOUTHEAST ASIA 541 1964, 1969 and 1979. For these years, the ratios of sectord employment to total employment were estimated using straight-line interpolations. Agricultural and industrial employment figures were then calculated from these estimates. Education. Economy-wide human capital is proxied by the mean school years of education of the total labour force. This data series was constructed by Nehru et al. (1993) from enrolment data using the perpetual inventory method and is available up to 1987. Data for 1988-98 are forecast by fitting the series using an OLS regression of human capital on time. Indonesia Gross DomesticProduct (agridmd and total). The GDP series in current and constant prices were obtained from various issues of Statistik Indonesia (the Statistical Yearbook of Indonesia), Badan Pusat Statistik (BPS). The agricultural GDP series includes forestry and fishery. Employment (agricultural and total). The National Labour Force Survq, contains data on population aged 10 years and over who worked, by main industry. These were obtained from various issues of Statistik Indonesia, BPS. Data for the missing years of 1972-75, 1979 and 1983-84 were estimated using s~rai~ht-lineinterpolations of the ratio of agricultural employment to total employment, as well as the total employment series. Agricultural employment was then calculated from these estimates. Employment data reported for 1998-99 were for population aged 15 and over, so the annual growth rates in this series were calculated and applied to the previously mentioned employment series (for population aged 10 and over) to obtain estimates for these years. Consumer Price Index. Data on the consumer price index of 17 capital cities in Indonesia are reported in the Intemtioml Financial Statistics of the International Monetary Fund. Data were converted to a base year of 1993. Education. Economy-wide human capital is proxied by the mean school years of education of the total labour force. This data series was constructed by Nehru et al. (1993) fiom enrolment data using the perpetual inventory method and is available up to 1987. Data for 1988-98 are forecast by fitting the series using an OLS regression of human capital on time. INDEX Page numbers in italic1 refer to Figures and input-output ratios 456-60 Tables. marginal productivity of capital 464-65, Abbreviations (Id), Up), (Ph) and (Th) refer 472, 473 to ~ndonesia;Japan, Philippines and Thailand marginal productivity of fertiliser 464, 466- respectively. 67, 471, 474, 520 marginal productivity of inputs 46243, 464 abaca industry (Ph) 257, 259 marginal productivity of labour 464, 465- Abrenica, M. 278 66, 470, 472 ABRI (Angkatan Bersenjata Republik Indonesia) marginal productivity of land 46344, 470, 174 471, 473 Abueva, J. 137 marginal rates of substitution 464, 467-70 AFMA (Agricultureand Fisheries Modernization political determinants 4 1 4 3 Act) (1997) (Ph) 254, 280 production function estimation 474-80 African countries, protection policies 406 production function summary results 46042 AFTA (ASEAN Free Trade Agreement) 279 shadow prices evaluation 462-63, 464 Agrarian Land Reform Code (1963)(Ph) 39 state variables for 479-80, 481-82 agrarian structures agricultural growth accounting agricultural growth and 4, 3641, 44-45 comparative dara 480-84, 520-21 comparative data 19-20, 28-29 Indonesia 504-07 ecological bases 3-4, 15-16, 17 Philippines 516-20 impact on poverty reduction 65, 67 Thailand 495-98 '480, 481-82, 483-84, Indonesia 4, 33, 35 520-21 land pre-emption and tenancy 4, 32 agricultural labour (see labour) Philippines 4, 34-35 agricuIturaL Labour productivity (see labour plantation system 4, 29-32, 37-38, 41, productivity) 44-45, 65, 201 agricultural land, defined 71 (see also land) Thailand 4, 35-36, 44 agricultural performance 62, 63, 380-82 agribusiness (Th) 345-46 agricultural productivity agricultural commodities (see commodities) determinants of 87 agricultural development, economic growth Indonesia 12, 493, 498-507, 512-13, 531 driven by 309 Philippines 11-12, 491, 493, 501, 507- agricultural GDP 520, 531 economic growth and 62, 63 Thailand 12, 487-498, 501, 512, 513, 531 global data 50, 51-52 agricultural productivity growth growrh rates 50, 51-52, 53 capital accumulation determining 58-59 agricultural growth (see also agricultural consequences of 5, 62-68 productivity) determinants of 55, 57 agrarian structure and 4, 3641, 44-45 initial conditions derermining 57-58 comparative data 21-22, 23, 44, 455-56, 457 investment determining 58, 68 comparative evaluation of 11-12, 484-87 protection policies and 59-60, (2,67, 68, derived prices 464, 467-70, 480 380-81, 390 growth accounting 480, 481-82, 483-84, Agricultural Rehabilitation Plan (ARP) (Th) 520-21 308, 357 INDEX 543 Agriculture and Fisheries Modernization Act Philippines' production 223-24, 250 (AFMA) (1997) (Ph) 254, 280 production systems 29-30 ago-industrial commodities 315 (see also Bangladesh 138, 153, 320 pineapple industry; poultry industry) Bank Desa 146-52 almost ideal demand system (AIDS) 43941, Bank of Agriculture and Agricultural 442 Cooperatives (BAAC) (Th) 133, 154, 308, Anderson, Benedict 84 343-44, 355, 357-58 Angka&n Bmenja~Rqublik Ilndonesia (ABRI) 174 BAPPENAS (National Development and animal feed (see also maize industry; tapioca Planning Board) (Id) 83 industry) barangays 135-37, 154 Indonesia 98 Basic Agrarian Law (Undang-Undung Pokok Philippines 225, 250, 254 Agraria) (Id) 185 animal husbandry (see livestock industry) Bates, R. 380-81 Aquino, Benigno 219 Benedicro, Roberto 110 Aquino, Corazon Cojuangco 86, 269-70 Bimas program (Id) 190-91, 200, 202, 503 Aquino administration Bina Swadaya (Id) 14546, 149 accession to power 107, 269 Board of Investment (BOI) (Th) 303, 34546 agricultural policies 110, 250-53, 270, 280 Board of Investments (BOI) (Ph) 240, 279 elitism in 281-82 Boeke, J. 162, 172 land reform 249, 251-52, 255, 275-76, 280 Booth, A. 173 problems during 255 Bowring Treaty 25-26 reform programs 241-43, 269-70 Bresnan, John 113 Arndt, H. 210 Bridgestone Corp. 319 ARP (Agricultural Rehabilitation Plan) (Th) Britain (see Great Britain) 308, 357 Buddhist temples 130-3 1, 153-54 Asahi Broiler up) 331 BULOG ( B d n Uman LogistiklState Logistics AS&V (Association ofSoutheast Asian Nations) Agency) (Id) 279 assessment of 100-01 A S W Free Trade Agreement (AFTA) 279 creation of 96 Asian currency crisis 221 expansion of 97--99 Assembly of the Poor (Th) 295 financing of 99 AssociationofSoutheast AsianNations (ASEAN) pricing policies 384 279 revenue sources 100 Athukorala, P. 427 rice policies 101-03, 105, 188, 384 authoritarian regimes 5-6, 74-75, 111-13 role of 95-97 (see also Indonesia; Philippines) sugar policies 103-05 Ayala Corp. (Ph) 266 trading operations 82-84 Burma (see Myanmar) BAAC (see Bank of Agriculture and Agricultural BUUD ( B d n Unit Usaha Desa) (Id) 191, 203 Cooperatives) Badan Unit Usuha Desu (BUUD) (Id) 191, CALF (Comprehensive Agricultural Loan 203 Fund) (Ph) 250 Badan Urusan Logistik (Id) (see BULOG) Cane Intensification Progcun (TRI) (Id) 103-04 banana industry capital marker share data 21-22, 23 accumulation of, in ~rod~ctivitygrowtb 58-59 Philippines' exports 422, 423 growth of agricultural 457, 458-59 Philippines' NPRs 256, 257, 259 marginai productivity of 464-65, 472, 473 544 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH CARD (Center for Agricultural and Rural copra production, Phiiippines 246, 256, 257 Development) (Ph) 140 copra protection rates 256, 257, 383 CARP (see Comprehensive Agrarian Reform Indonesia 194,195, 197 Program) Philippines' policies 108, 246, 248, 250, 259 cassava industry Philippines' production 224, 224, 226-27 Indonesia 192-93, 195 production systems 30 Philippines 106 Coconut Industry Stabilization Fund (Ph) 385 Thai crop-income share 360 coconut oil industry (see also coconut industry) Thai exports 296, 423, 424-25 market share data 23 Thai production 295, 298, 301-02, 321 Philippines 246, 256, 257, 422, 423 Thai yield trends 341 Thailand 360 Catholic Church 266 (see also religion) coffee bean industry cattle feed (see animal feed) Indonesia 193, 194,195, 196, 197,421, 422' CDA (Cooperative Development Authority) market share data 21-22, 23 (Ph) 137 Philippines 223-24 Cenraco Farm Co., Lrd. (Th) 332 Thailand 301 Center for Agricultural and Rural Development Cojuangco (Ph) 266 (CARD) (Ph) 140 colonial policies Center for Strategic and International Studies British free trade system 25, 26-28 (CSIS) 175, 202 Dutch 4, 32, 33, 35, 103, 104 Chao Phraya River delta Spanish 4, 34, 35, 135-36, 266 agrarian systems 35-36 USA 32, 266, 267 canal consrruction 25, 301 commodities (see aiJo speczfic industries) Chainat Dam 301 export performance 421, 42243, 434-35, location 24 436, 437-38 rice production 44, 299-300 FA0 classifications 194 Charoen Pokphand Co., Ltd. (Th) 331, 345 Indonesian 168-69 Chartchai Choonhawan 307 determinants 87 Chavalit Yongchaiyudh 295 Thai income shares of 360 chickens (see poultry industry) types of 23 China, People's Republic of 315, 316, 426 commodity-specific policies Christensen, S. 80 Indonesia 88, 95-105, 112 Chulalongkorn (Rama V ) 44, 78 Philippines 88, 105-1 1, 112 clientelism Thailand 87-95, 114 Indonesia 82-83 communal organisations 125-26 (see also Philippines 84 microfinance organisations) CMSA (constant market share analysis) 427-29 comparative advantage cocoa bean industry determinants of 449 Indonesia 194, 197, 421, 422 national resources bases for 44 production systems 30 traditional crops 21-22 coconut industry competitiveness, of arports 427-28, 435, 449, coconut oil market share data 23 450 coconut oil, Phiiippines 246, 256, 257, 422, Comprehensive Agrarian Reform Program 423 (CARP) (Ph) 227, 251, 273, 275-76, 508 coconut oil, Thailand 360 Comprehensive Agricultural Loan Fund copra production, Indonesia 193, 194, 385, (CALF) (Ph) 250 450 constant market share analysis (CMSA) 427-29 INDEX 545 continental ecological zones 15-16, 16 17, 24 employment-educational level ratios 355 contract farming 38, 45 desa 142-44, 151 Cooperative Development Authority (CDA) Desa Banks 146-51, 152 (Ph) 137 development organisations 125-26 (see also cooperatives (rural) microfinance organisations; rural Indonesia 175, 178, 191, 200, 203 organisations) Philippines 138, 137, 140 development projects for poverty reduction 124 policy implications for 152-53 copra industry (see also coconut industry) rural organisations for 123-24 Indonesia 193, 174, 385, 450 development trilogy xxii, xxiii-xxiv, miii Philippines 246, 256, 257 de Wolff van Westerrode, P.D. 155 protection rates 256, 257, 383 Dillon, H. 195-96 corn industry (see also maize industry) direct nominal protection rate (DPR) 433 Indonesia 98 distribution systems 448 Philippines 105, 106, 108, 256, 257 Dole Food Company, Inc. (US) 333-35 CBte d'Ivoire 334, 335, 336 Dowling, J., Jr. 304 Coxhead, I. 487 Dutch colonial era (see colonial policies, Dutch) Cropper, M. 339 Dutch East Indies (see Indonesia; Netherlands) CSIS (Center for Strategic and International Studies) 175, 202 East Timor 188 ecological zones 15-16, 16, 17, 24 Datt. G. 65 econometric models 391 David, C. 258, 381, 390-91, 402, 407 economic growth Deaton, A. 439 agricultural drivers 309 de Dios, E. 279-80 agricultural linkages xxiii-xxiv definitions 71-72 agricultural sector performance and 62, 63 deforestation (Th) 338, 339 comparative data 455-56, 457 de Gorder, H. 407 determinants of 449 Deininger, K. 57 in development trilogy xii, miii de Janvry, A. 281 poverty reduction due to 65-67 Del Monte Corp. (US) 333, 335 pro-poor strategies xxv democratic regimes 5-6, 75, 111-1 3 (see also economic indicators (Id) 165 'Thailand) economic performance, agricultural sector 62, demographics (Indonesia) 63, 380-82 (see also productivity growth) demographic and social indicators 16647, 215 economies, World Bank classifications 201 ethnic groups 170, 201 education, and migration rate 533, 538 population 201 educational level rural-urban employment 169 agricultural growth and 481, 482 demographics (Philippines) and employment in Thailand 355 demographic and social indicators 167, 214, and income in Thailand 36445 215, 236-38 income related to 353 Ifugao ethnic group 154-55 EEC (European Economic Community) 325, rural poor 218, 219, 227-30 326, 365, 425 rural-urban changes 290-93 effective rate of protection (ERP) (Ph) 260, 281 demographics (Thailand) Ellis, F. 100 demographic and social indicators 167, 215, employment 295, 346-47 agriculture's share of total 524, 524 546 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH educational level in Thailand 355 Facoma (Farmers Cooperative Marketing Indonesian sectoral distribution 169-70 Associations) (Ph) 138 entrepreneurship 172-73 Family Income and ExpendituresSurvey (FIES) environmental conservation (Th) 313 (Ph) 230-38, 278, 290-93 equilibrium exchange rate 433 family planning programs (Id) 167 error correction model (ECM) 429, 450 FA0 (Food and Agriculture Organization), Esguerra, E. 248 commodity classifications 194 estate crops 201 (see also plantation system) Farmers' Aid Fund (Th) 90, 91 European Economic Community (EEC) 325, Farmers Cooperative Marketing Associations 326, 365, 425 (Facoma) (Ph) 138 Evans, Peter 74 farm size, distribution of 19-20, 448-49 exchange rates Fei, J. 62 in export performance 433 fertiliser Philippines 260-61, 281 Indonesia 190-92, 198, 200 export performance marginal productivity of 464, 46647, 471, by commodity 421, 422-23, 434-35, 436, 474, 520 437-38 Philippines 248, 514-15 comparative 416-19, 446, 447, 449, 450 ratio to land 459-60 constant market share analysis of 427-30 Thailand 342-43 determinants of 427-28, 447-49 fibre crops (XI) 360 (see alio kenaf industry m) effects of government interventions 433-34, FIES (Family Income and Expenditures Survey) 438, 439 (Ph) 230-38, 278, 290-93 exports-imports price transmissions 10, financial liberalisation (Ph) 241, 243, 276 441-46, 447-48 fiscal reforms (Ph) 242-43, 245 growth accounting analysis of 431-33 fishing industry product differentiation and price competition Philippines 222-23, 224, 225 in 438-41, 442 Thailand 306, 306 Southeast Asia 5, 63-64, 67 Five-Year Plans (Id) 187, 188, 190, 191, exports (agriculd) (see also export performance) 198, 202 benefits of 415 flour industry comparative performance of total 416-19, Indonesia 97, 99 446, 447, 449, 450 Philippines 250 competitiveness of 427-28, 435, 449, 450 flower industry (Th) 360 demand structure for 10, 415-16, 438-41, Food and Agriculture Organization (FAO), 442 commodity classifications 194 destinations of 425-26, 442, 447 food crops (see also specific industries e.g. rice determinants of 363, 427-28, 447-49 industry) Indonesia 10, 11, 168-69, 421, 422, 446, Indonesia 192-93 447 Thailand 296, 301 Indonesian crop development for 193-96, food security 197, 200 bases of xxii-xxiii performance by commodity 421, 422-23 in development trilogy xxii, xxiii-xxiv Philippines 10, 11, 226-27, 421-24, 446, Indonesian crises 99 447 food self-sufficiency, and poverty xxii share of total exports 419-20 food supply, growth rates 55, 56 Thailand 10, 11, 296, 305-06, 423, 424- forestry sector 25, 446-47 historical exploitation of Thai rain forem 2526 Philippines 224, 225 development of infrastructure 381 Thailand 296 economic development 50, 62 Foster-Greer-Thorbeckeindex 278 intra-village cooperation 137 Four-Pronged (Si Praran) Project (Th) 306, 308 political entrepreneurs 407 fruit industry (Th) 296, 296, 301, 360 (see rural incomes 236, 264, 278, 281 also pineapple industry) Herdt, R. 235 Furnivall, J. 35 High Performing Asian Economies (HPAE) 164-65 Gallup, J. 65 Hill, H. 188 Garcia-Garcia, J. 100, 104 Honma, M. 451 Gardner, B. 407 HPAE (High Performing Asian Economies) GDP (see gross domestic product) 164-65 Geertz, C. 162, 170, 172-74, 191 Huang, J. 381, 390-91, 402, 407 Glassburner, B. 176 Huke, E. 44 glossary of terms 71-72 Huke, R. 44 GNP (see gross national product) Hull, T. 167 Golkar party (Id) 187 Hutchcroft. P. 271-72 governance Indonesia 77, 80, 81-83 IFAF (Integrated Farmers' Aid Fund) (Th) Philippines 77, 80, 84-87 308, 356 political economy of sectoral 5-6, 75-77, Ifugao ethnic group 154-55 111-13 IGGI (Inter-GovernmentalGroup on Indonesia) Thailand 77-8 1 186 Grameen Bank 128, 138-39, 140-41, 153 IMF (International Monetary Fund) 102-03, 110 Great Britain, free trade system 25, 26-28 import-substitution industrialisation (ISI) Grifiths, C. 339 East Asia 271, 282 gross domestic product (GDP) Philippines' maintenance of 270-73, 275- agricultural sector 50, 51-52, 53, 62, 63 77, 281, 282 Asian per capita data 352 policy instruments associated with 272-73 comparative data 215, 216 in post-war period 265 Indonesian regional data 210-1 1, 212-13 import-substitution policies gross national product (GNP) adoption of 267, 268, 281 comparative data 22, 164 in Philippines 239-42, 265, 267-68, 281 and poverty in Philippines 232 income groundnut industry (Id) 98, 192 agriculture protection related to 389 growth accounting calculation of 539 comparative data 480-84, 520-21 sectoral ratios 530-32 Indonesia 504-07 income distribution Philippines 516-20 Indonesia 203, 210-11, 212-13 Thailand 495-98 migration rate and 529-33 Guided Economy policy (Id) 184-85 poverty reduction and 65 gunnysacks 320-21 Thailand 9, 351-53, 362, 364 incremental output-capital ratios (IOCRs) (Ph) Hamengkubowono IX 186, 202 261-62 Hara, Y. 42 indirect nominal protection rate (IPR) 433 Hatta, Mohammed 177 Indonesia (see also entries for Indonesia under Hayami, Y. specific industries e.g. rice industry) 548 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH agrarian systems 4, 33, 35 protection policies 9-10, 382-84, 386-87, agricultural productivity 12; 493, 498-507, 404-06 512, 513, 531 regional development and agriculture 210- agriculture-GDP ratio 220, 274 11, 212-13 area of 170, 201 regional dimensions 170-71, 203, 210- commodity-specific policies 88, 95-105, 112 11, 212-13 community structures 151-52 rural administrative system 142-44, 155 comparative growth data 455-56, 457 rural pessimism and optimism 172-74, demographics 16647, 169, 170, 201, 215 178, 199 development chronology 179, 180-82, rural sector importance 169-70 183-86 sectoral governance 75-76, 77 development policies 7, 170-71, 186-201, social organisations 144-45, 155 265 sources of growth 481, 505 ecological zones 24 structural adjustment 196-99, 200-01 export crops 193-96, 197, 200 trade partners 425, 442 export performance 10, 11, 168-69, 416- trade structure 168-69 20, 422, 446, 447, 450 industrial sector export performance (comparative) 421 Indonesia 167-68, 196-99, 210-11, 213 farm size distribution 448-49 Philippines 218, 238-41, 242, 244, 264- food-trading parastatah 87, 90 (see also 65, 276-77 BULOG) Thai exports 420 governance and polity 77, 80, 81-83 infrastructure, and agricultural growth and history 2-3 481-82 independence of 183-84 Inmas (Intenszj5kasi MnssaL) (Id) 202 industrial growth 196-99, 200-01 inputs industrial sector 167-68, 196-99, 210- growth of agricultural 457, 458-60 11, 213 marginal productivity of 462-63, 464 input-output ratios 456-60 institutions, purpose of 124-25 labour migration 171, 527-28, 528 insular ecological zones 15-16, 16, 17, 24 land endowments 17, 18, 19 Intal, l? 108, 240 macroeconomic growth 164-68, 215, 216 integralism, doctrine of 174 microfinance organisations 6, 128, 141, 145- Integrated Farmers' Aid Fund (IFAF) (Th) 51 308, 356 military power 82-83, 113, 174 Intenszfikasi Massal (Inmas) (Id) 202 New Order fertiliser policies 190-92 interest groups, influence of 390, 407 New Order rice policies 7, 186-91, 199-200 Inter-Governmental Group on Indonesia nominal protection rates (NPRs) 259, 387 (IGGI) 186 non-rice foodcrops 192-93 International Monetary Fund (IMF) 102-03, non-traditional commodities 430, 450 110 paddy area 44 International Rice Research Institute (IRRI) plantation system 44-45 188, 340 political arena 174-77 invesrmenr, in productivity growth 58, 68 polirical ideology 162-63, 174-76, 178- Investment Incentive Acts (Ph) 240, 279 79, 183, 200, 202 IOCRs (incremental output-capital ratios) (Ph) poverty 203, 236, 274 261-62 INDEX 549 Irian Jaya 185, 202 Labor Force Survey (LFS) (Ph) 230-38 IRN (International Rice ResearchInstitute) 188, labour 340 agricultural, defined 71 IS1 (see import-substitution industrialisation) data sources 526 Islam 176-77 growth rate and 456, 457 Ivoiy Coast (see CBte d'lvoire) marginal productivity of 4a,465-66, 470, 472 Japan Philippines' agricultural GDP share 256, animal feed imports 321-23 510-11 exports to 426 ratio to land 456, 457 FDI in Thailand 307 Thai agricultural GDP share 493, 493 frozen chicken imports 331-33 Thai non-agricultural GDP share 493-94 maize imports 424 labour migration (see entries beginning with occupation of Indonesia 183, 184 migration) Java (see Indonesia) labour productivity 50, 51-52,53 Jocano, F. 137 Ladejinsky,W.190 Johnson, D. Gale xxiv land Johnston, B. 62 agricultural, defined 71 Johnston-Mellor Linkages xxiv comparative endowments 17, 18, 19 Jones, G. 167 cost of Indonesian 506 Jorgenson, D. 62 growth rates 456, 457,458 jute industry 320-21 marginal productivity 46364, 470, 471,473 ownership in Java 148, 155 Kano, H. 144 pasture, defined 71 Kanshoku Co., Ltd. Up) 331, 332 ratio of fertiliser 459-60 kenaf industry (Th) 295, 296,298, 302, ratio to labour 456, 457 320-21, 341 land distribution Kenya 44 colonid concessions 4, 31-32 Kikuchi, M. 137, 236, 264, 278, 281 export expansion and 448-49 Kishi, K. 144 regional 57-58 KOLOGNAS (Komando Logistik Nariona~(Id) land reform 96 (see aLo BULOG) implementation of 273 Komando Logistik Nacional (KOLOGNAS) (Id) in Indonesia 185-86 96 (see aLo BULOG) land reform (Philip ines) p Koperasi Unit Desa (KUDs) (Id) 102, 103, under Aquino 249, 251-52, 255,275-76, 145, 191, 203 280 Kravis, I. 427 under Marcos 249, 255 Kriangsak Chomanan 349-50 overview 39-41 KUDs (Koperasi Unit Desa) (Id) 102, 103, under Ramos 25455, 273-74, 282 145, 191, 203 land tenure Kukrit government (Th) 349 comparative tenancy distribution 19-20 Kuznets, S. 65 institutional development 4, 32-33 Kuznets' Curve hypothesis 57, 65 Philippines 519-20 Kwik Kian Gie 177 Thailand 337-38 languages, Indonesian 170, 201 550 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Larson, D. 60, 520, 525, 528 NFA financing 106-07 Lewis, W. Arthur xxiv, 523 policy reforms 269 Lewis Linkages xxiv rice industry 108-09 LFS (Labor Force Survey) (Ph) 230-38 sugar industry 105, 109-10, 279-80 Lindert, P. 389, 390-91 Marketing Organization for Farmers (MOF) Lipton, Michael xxiv (Th) 90-91, 343, 355 livestock industry Masagana 99 program (Ph) 42, 86, 108, 245- Indonesian meat-supply policies 98 46, 247-48, 279 pastureland share of agriculture 55, 56 MAV (minimum access volume) system 25554 Philippines 224, 225, 250, 256, 257, 258 meat supply (see livestock industry; poultry Thailand 296, 310 indusrry) Logic of Collective Action, The (Olson) 390 Medalla, E. 260 Love, J. 427 Mellor, John m, 62 lumbung Asa (village rice banks) 146, 147, 155 microfinance organisations analytical framework 125-26 Magna Carta of Small Farmers (Ph) 251, 253, Bangladesh 153 280 Indonesia 6, 128, 141, 145-51 maize industry (see also animal feed; corn Philippines 6 , 128, 135, 137-41, 150- industry) 51, 154-55 Indonesian production 192, 192, 193 Thailand 6, 127, 129, 131-34, 150-51, market share data 23 154, 349, 350 Philippines' policies 254, 258, 384 typology of 127-28 Philippines' production 223-24, 225, 226 migration protection rates 383 Indonesian rural-urban 527-28, 528 Thai exports 296, 423, 424, 425 Indonesian 'transmigration' programs 171 Thai income share 360 Philippines' rural-urban 290-93 Thai policies 314, 385 Thai rural-urban 303, 307, 353, 354 Thai production 295, 298, 302, 321-23 migration (intersectoral) Thai R&D 341-42, 345 analysis conclusions 12, 538 Thai yield trends 341 analysis results 534-38, 539 Malari riots (Id) 187 analyrical framework 525-29 Malaysia analytical variables 529-34 agricultur+GDP ratio 220 data sources 539, 540-41 governance and polity 80 military power palm oil industry 194, 335 Indonesia 82-83, 113, 174 pineapple industry 334, 335 Philippines 1 13 plantation system 44-45 Thailand 113 rubber industry 318, 319, 320 minimum access volume (MAW system 25554 Malthus, T. 62 Mizuno, K. 155 Mangkusuwondo, Suhadi 176 models, economerric 391 Mani, M. 339 Moertopo, Ali 174, 202 Marcos, Ferdinand 84, 86, 108, 240, 269 MOF (Marketing Organization for Farmers) Marcos, Imelda 106 (Th) 90-91, 343, 355 Marcos administration moneylenders agricultural policies 245-50 Indonesia 148 IS1 policies 240-42, 279, 281 Thailand 133-34 INDEX 551 Mongkut (Rama IV) 25, 44, 78 NEDB (National Economic Development mongo bean industry (Ph) 106 Board) (Th) 94 Mosher, Art xxv NESDB (National Economic and Social Muellbauer, J. 439 Development Board) (Th) 303, 306, 307 Mundlak, Y. Netherlands agricultural productiviry 58, 459, 479 colonial policies 4, 32, 33, 35, 103, 104 algorithm 478, 523 reoccupation of Java 183-84 government intenrention evaluation 60, 380 rural development of Indonesia 179, 180, migration from agriculture 525, 528, 529, 532 183, 190 mungbean industry (Id) 98 Newberry, D. 386 Muslims 176-77 newly agro-industrializing countries (NAICs) Myanmar (formerly Burma) 25, 315, 316 297, 306-07, 310 Myint, H. 64 New Order regime (see Indonesia) NFA (see National Food Authority) Na Chiangmai, C. 359 NFAC (National Farmers' Aid Committee) NAICs (newly ago-industrializing countries) (Th) 356 297, 306-07, 310 NGA (National Grains Authority) 106, 246 NASUTRA (National Sugar Trading NGOs (see non-government organisations) Corporation) (Ph) 109, 110 nominal protection rates (NPRs) National Council of Sugar Producers (Ph) 111 defined 433 National Development and Planning Board Philippines 256-59, 280, 387 (BAPPENAS) (Id) 83 non-government organisations (NGOs) (see aho National Economic and Development Authority rural organisations) (NEDA) (Ph) 85 programs in Indonesia 14546, 149-50 National Economic and Social Development programs in Philippines 138-39, 140-41, Board (NESDB) (Th) 303, 306, 307 150, 151, 154-55 National Economic Development Board NRDP (National Rural Development Program) (NEDB) (Th) 94 (Th) 350 National Farmers' Aid Committee (NFAC) (Th) 356 Ockey, J. 79 National Food Authority (NFA) (Ph) oil crops (see coconut oil industry; palm oil evaluation of 105 industry) functions of 246 oil palm (see palm oil industty) under Marcos 106-09, 246, 281 oil (petroleum) sector (Id) under Ramos 254 economic significance 167, 168, 169, 188, 198 role of 85, 86, 384, 404 Pertamina 83, 202 National Grains Authority (NGA) (Ph) 106, 246 regional earnings 201, 210 National Grains Industry Development Fund terms of wade 196 (Ph) 106 oligarchies 265-66, 271-72, 273 National Rural Development Program (NRDP) Olson, M. 390 (Th) 350 organisations, rural (see rural organisations) National SugarTradingCorporation (NASUTRA) (Ph) 109, 110 paddy production natural rubber (see rubber industry; rubber trees) areal extent 44 NEDA (National Economic and Development comparative shares 19 Authority) (Ph) 85 paddy, defined 71 552 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Palat Cooperative (Ph) 140 export performance 10, 11, 226-27, 421- ~ a l moil industry 24, 446, 447 Indonesian exports 421, 422 export performance (comparative) 41620,450 Indonesian introduction of 183 farm size distribution 448-49 Indonesian policies 382, 384 food-trading parastatas 87, 90 Indonesian production 195-94, 195, 196, 197 governance and polity 77, 80, 84-87 Malaysia 194, 335 government intervention 41, 42 market share data 21-22, 23 history 2-3 protection rates 383 impact of commodity-specific policies 256- Thailand 301, 360 65, 280 paneas&, doctrine of 174, 202 incremental output-capitd ratios 261-62 pastureland industrial sector 218, 238-42, 244, 264- defined 71 65, 276-77 share of agriculture 55, 56 input-output ratios 456-60 patrimonial states 271-72 labour migration 290-93, 527, 528, 528 Peasant Federation of Thailand 89 land endowments 17, 18, 19 peninsular ecologicalzones (see insular ecological land reform (see land reform (Philippines)) zones) macroeconomic growth 7-8, 214, 215, 216- pepper industry (Id) 194, 195 21, 274 performance, agricultural 62, 63, 380-82 macroeconomic policies 238-39 Pertamina (Id) 83, 202 Marcos administration 105-11 , 240-41, petroleum (see oil (petroleum) sector) 242, 245-50, 269, 279-81 Philex (Philippine Exchange) 109-10, 246 microfinance organisations 6, 128, 135, 137- Philippine Exchange (Philex) 109-10, 246 41, 150-51, 154 Philippine National Bank (PNB) 246 military power 113 Philippines (see also entries for Philippines nominal protection rates 256-59, 280 under specific industries e.g. rice industry) paddy area 44 administrative ladder 135 political economy of delayed reforms 264- agrarian systems 4, 34-35 74, 276-77, 281 agricultural policies 245-55, 275-76, 279-80 poverty (see poverty (Philippines)) agricultural production 222-27, 275, 277-78 protection policies 9-10, 382, 383, 384- agricultural productivity 11-12, 491, 493, 87, 404, 405-06 501, 507-20, 531 Ramos administration 243-45, 249, 253- Aquino administration 241-43, 249-53, 56, 273-74, 282 255, 269-70, 275-76, 280-82 regional crop diversity 222 British control of 26 rural admiistrative system 135-36, 141, 154 commodity-specific policies 88, 105-12 rural development 221-27 community structures 13637, 141, 151, 154 rural organisations 273-74 comparative growth data 455-56, 457 secroral governance 75-76, 77 demographics 167, 214, 215, 218, 219, sources of growth 481 227-30, 236-38, 290-93 trade partners 426, 442 development policies 238-65 wage rate trends 262-65, 281 development strategies 238-39 pineapple industry ecological zones 24 global trade data 333, 334, 335 economic performance 164 Malaysia 334, 335 exchange rate movements 260-6 1, 281 market share data 21-22, 23 Philippines' exports 333, 334, 422, 423 wage rate trends 262-65, 281 Philippines' NPRs 256, 257,259 poverty (Thailand) Philippines' production 223-24, 335, 336 agricultural price policies to reduce 353, 355- production systems 38 56 Thailand 294, 333-36, 341, 345, 360 changed policy-targets to reduce 358-59 Plangpraphan, J. 487 diversification and differentiation to reduce plantation crops 201 359-61 plantation system production control policies to reduce 357-58 conditions of 4, 29-32, 44 reduction of 236, 295 Malaysia 4 4 4 5 reduction of absolute 305, 3 4 6 4 8 negative aspects 37-38, 41, 4 4 4 5 , 65 reduction strategies 349-51, 364 plantation crops, defined 201 rural-urban disparities 9, 304, 351-53, Plessmann, F. 520 362, 364 pluralism (Th) 5, 79 trends over time 274 PNB (Philippine National Bank) 246 poverty gap index 278 political pressure groups (see interest groups) poverty reduction polity in development trilogy xxii, xxiii-xxiv in Indonesia 77, 80, 81-83 export expansion for 448 in Philippines 77, 80, 84-87 through productivity growth 65-67, 68 in Thailand 77-81 Power, J. 108, 240 poor, the (see entries beginning with poverty) Prem Thinsulanonda 350 population data (see entries beginning with pressure groups (see interest groups) demographics) price index, Stone's 440 populations, organisation of nuaJ 124, 152-53 prices, effect on productivity growth 60, 61 pork (see livestock industry) price stabilisation 10, 38687, 405-07 poultry industry privatisation (Ph) 244, 245, 276 Philippines 224,225, 250, 256, 257 ProductionSystem RestructuringProject (Th) 308 Thailand 294, 296, 306, 330-33, 345 productivity, agricultural poverty determinants of 87 incidence of 486 Indonesia 12, 493, 498-507, 512, 513, 531 measures for 227, 278 Philippines 11-12, 491, 493, 501, 507- poverty gap index 278 20, 531 'shared' 173 Thailand 12, 487-98, 501, 512, 513, 531 poverty (Indonesia) productivity growth poverty line 203 capital accumulation determining 58-59 reduction of 236 consequences of 5, 6 2 4 8 trends over time 274 determinants of 55, 57 poverty (Philippines) initial conditions determinin g 57-58 characteristics of rural poor 227-30 investment determining 58, 68 incidence of 2 4 4 4 5 protection policies and 59-60, 61, 67, 68, policy biases against 248 380-81, 390 reduction policies 255-56, 280 protection policies regional profiles of income poverty 230 agricultural secror performance and 59-60, rural-urban population changes 290-93 61, 67, 68, 380-81, 390 trends over time 218, 230-38, 250, 274- analysis methodology 391-92 76, 278-79 analysis results 392, 393-98, 399-406 554 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH analytical framework 382, 387-91, 411-14 Philippines 252-53 comparative impact of 433-34, 438, 439 Thailand 339-42, 345 determinants of 406, 407 Restructuring and Production System (RAPS) effect on productivity growth 59-60, 61, (Th) 308, 357-58 67, 68 Ricardo, D. 62 Indonesia 9-10, 382-84, 386-87, 404, Rice and Corn Administration (RCA) (Ph) 105 405-06 rice area, defined 71 Philippines 9-10, 382, 383, 384-87, 404, rice industry 405-06 HYVs 188, 190, 340 for price stabilisation 10, 386-87, 405-07 Indonesian Five-Year Plans 187, 188, 190, protection rates 433 191, 198, 202 Thailand 9-10, 382, 383, 385-87, 404-07 Indonesian policies 42-43, 95-97, 99, 1 0 C protection rates, defined 433-34 03, 112, 163, 184, 1 8 0 1 , 199-200, 384 PRRM (Philippine Rural Reconstruction Indonesian production 188-91, 202 Movement) 154-55 land share 53, 54, 55 Public Warehouse Organization (PWO) (Th) market share data 21-22, 23 90, 355 Philippines' NPRs 256, 257, 258, 280 PWO (Public Warehouse Organization) (Th) Philippines' policies 105, 107, 108-09, 90, 355 112, 245-48, 250, 254, 384 Philippines' production 189, 223-24, 225, Quezon, Manuel 281 226 production of rice, defined 71 rai, defined 44 productivity growth 54, 55 rain forests, historical exploitation 25-26 protection rates 383 Ramage, D. 176 Thai exports 294, 423, 424 Rama IV (see Mongkut) Thai income share 360 Rama V (see Chulalongkorn) Thai policies 87-91, 297, 315-18, 385 Ramos administration 243-45, 249, 253- Thai price policies 353, 355-56 56, 273-74, 282 Thai production 189, 295, 296 297, 310, Ranis, G. 62 312 RAPS (Restructuring and Production System) Thai R&D 339 (Th) 308, 357-58 Thai regional production 299-302 Kavallion, M. 65 Thai yield trends 341 reform implementation, and regime type 5-6, rice production, defined 71 (see also rice 74- 75, 111-13 industry) religion Rodrik, D. 271, 272, 281, 282 Indonesia 144, 145, 149, 170, 176-77 rubber industry Philippines 136-37, 141, 266 Indonesian exports 318, 319, 421, 422 Thailand 130-31, 153, 154 Indonesian policies 382 rent-seeking Indonesian production 193-94, 195, 197, East Asia 271, 282 318, 319 Indonesia 83, 100, 105, 202 Malaysia 318, 319, 320 Philippines 84, 85, 107, 165, 254, 266, market share data 21-22, 23 269, 277 Philippines 194, 203 Thailand 78, 326 protection rates 383 research and development (R&D) Thai exports 294, 423, 424 Thai income share 360 Schultz, T.W. xxiv, 380 Thai policies 297, 318-20, 386 seafood industry (see fishing industry) Thai production 194, 295, 296 298, 301 sectoral governance, ~oliticaleconomy of 5-6, Thai R&D 339, 342, 345 75-77, 111-13 Thai yield trends 341 self-sufficiency, and protection rate 403 tubber trees, Indonesian introduction of 183 services sector (Ph) 244, 265 rural organisations Setboonsarng, S. 89 analytical framework 125-26 Siam Canals Land and Irrigation Company 25 Bangladesh 153 Siamwalla,A 89, 314, 342, 344, 359, 365, 487 contribution to development 123-24 Sicac, Gerardo 85, 107 defined 125 Si Praran (Four-Pronged) Project (Th) 306, 308 Indonesia 6, 141, 145-52 SN (Samabang Nayon) (Ph) 138 Philippines 6, 135, 137-41, 150-51, 15& Social Action Center of Panpangas (SACOP) 55 (Ph) 140, 154 and Philippines' local politics 273-74, 282 social indicators policy implications for 152-53 Indonesia 166, 215 purpose of 124-25 Philippines 214, 215, 236-38 Thailand 6, 129, 131-34, 150-51, 154, Thailand 215, 295, 346-47, 355 349, 350 Social Reform Agenda (SRA) (Ph) 256, 280 types of 125 Soedjono Hoematdani 175-76 typology of 6, 127-28 Soo, D. 304 rural pessimism and optimism (Id) 172-74, Soriano (Ph) 266 178 soya (see soybean industry) rural populations, organisation of 124, 152-53 soybean industry rural-urban migration (see also migration Indonesia 98, 99, 101, 192 (intersectoral)) Philippines 106 Indonesia 171, 527-28, 528 Thailand 301 Philippines 290-93, 527, 528, 528 Spain Thailand 303, 307, 353, 354, 526 528 colonial administrative units 135-36 Ruttan, V. 407 colonial policies 4, 34, 35, 266 Squire, L. 57 Sachs, J. 202 SRA (Social Reform Agenda) (Ph) 256, 280 SACOP (Social Action Center of Panpangas) Sriswasdilek,J. 340 (Ph) 140, 154 Sta. Monica Farmers' Cooperative (Ph) 140 Saha Farms Co., Ltd. (Th) 332 staple theory 64 Salim, Emil 202 State Logistics Agency (Id) (see BULOG) Samabang Nayon (SN) (Ph) 138 Stiglia, J. 386 Samudavanija, C.-A. 78, 79 Stone's price index 440 sanitation (Ph) 237-38 subsidies, effect on productivity !growth 60, Santa Monica Farmers' Cooperative (Ph) 140 61 (see also protection policies) Sar Desai, D. 78 Suchinda Kraprayoon 298 Sarit Thanarat 303 Sugar Code (Th) 328 savings groups (Th) 6, 131-34 Sugar Industrial Aid Fund (Th) 93 Schiff, M. 380, 382, 390-91, 433 sugar industry Schiller, J. 155 comparative market shares 22 Schmitter, Philippe 78 global production costs 365 556 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH global trade data 327, 328, 365 Thailand 294, 323-27 Indonesian policies 97, 99-01, 103-05, 384 tariff reform programs (see trade liberalisation Indonesian production 183, 202 (Ph)) marker share data 21-22, 23 taxation, effect on productivity growth 60, 61 Philippines' exports 421, 422, 423 (see also protection policies) Philippines' NPRs 256, 257, 258 TDRI (Thailand Development Research Philippines' policies 105, 108, 109-1 1, Institure) 339 245, 246, 248-50, 254, 279-80, 385 tea industry Philippines' production 224, 224, 226-27 Indonesia 193, 194, 195, 197, 421, 422 producrion systems 30, 31 production systems 29, 44 protection rates 383 Thailand 301 Thai exports 249, 296, 423, 425, 450 tenancy (see land tenure) Thai income share 360 terminology 71-72 Thai policies 91-95, 314-15, 385 terms of trade, agricultural 532-33, 539 Thai production 294, 295, 298, 301, 302, Thailand (see also entries for Thailand under 310, 327-30, 365 specific industries e.g. rice industry) Thai R&D 339, 345 agrarian systems 4, 35-36, 44 Thai yield trends 341 agribusiness policies 345-46 Sugar Market Development Fund (Ph), agricultural development 295-99, 301-13 proposed 111 agricultural productivity 12, 487-98, 501, Sugar PriceStabilization Bureau (SPSB) (Th) 328 512, 513, 531 Sugar Research Foundation (Ph) 111 agriculture-GDP ratio 274 Suharto British control of 26 accession to power 81-83, 186 canal construction 25, 301 agricultural policy 42 capitallcapital goods S&D 342-44 BULOG's role 96-97 commodity-specific policies 87-95, 114 land reform 185-86 comparative grovnh data 455-56, 457 loss of power 98 demographics 167, 215, 295, 34647, 355 rice self-suficiency policy 101 development policies 265, 349-51, 353, rural origins 163, 178 354-61, 364 Sukarno domestic food supply 310, 312 'Guided Economy' policy 184-85 ecological zones 24 ideology of 202 economic performance 164 overthrow of 81 environmental conservation 313, 338 Sukarno administration 175 export commodities 310, 311, 362, 363 Sumitro Djojohadikusumo 98 export performance 10, 11, 296, 305-06, Sutowo, Ibnu 202 423, 424-25, 446-47 sweet potato crops (Id) 192 expott performance (comparative) 416-20,450 Swinnen, J. 407 farmer welfare 313 farm size distribution 448-49 Taiwan 44, 334, 335 governance and polity 77-81 Tambon Administrative Organizations (TAOs) government intervention 41-42, 43 (Th) 130 history 2-3 tapioca industry (see aLo animal feed) industrialisation 9, 303, 307 global trade data 324, 325 industrial sector 420 Indonesia 324, 325 infrastructure 338-39, 349, 350-51 input-output ratios 456-60 Tolentino, V. 281-82 labour migration 303, 307, 353, 354, 526, Tongpan, S. 339 528 total factor productivity (TFP) land endowments 17, 18, 19 in economic growth xxiv land resources 337-38 Indonesia 504 macroeconomic growth 215, 216, 294-95, Philippines 516-18 303, 309-10 state variables 479-80, 481-82, 483-84, microfinance organisations 6, 127, 129, 131- 520-21 34, 150-51, 154, 349, 350 Thailand 495-97 military power 113 TPEA (Thai Produce ExportersAssociation) 323 nominal protection rares (NPRs) 259, 387 trade liberalisation (Ph) non-traditional commodiries 430, 450 AFTA hastening 279 paddy area 44 under Aquino adminisrration 24143, 250- political transformation 297-98, 303, 304- 51, 270 05, 308, 349-50 under Marcos administration 241 poverty (see poverty (Thailand)) under Ramos administration 253-54, 260, protecrion policies 9-10, 382, 383, 385- 276, 280 86, 404, 405-06, 407 trade protecrion (see protection policies) regional differences 299-301 trader's dilemma, theory of 173-74 tesearch and extension 33942, 345 tree crops (see also specific industries) rural administrative system 129-30, 136 comparative shares 19 rural organisations 6, 129, 131-34, 150- Indonesia 50607 51, 154, 349, 350 production systems 30-3 1 sectord governance 76-77 Thailand 360 sources of growth 482 types of 507 trade development policies 314-15 TRI (Cane Intensification Program) (Id) 103- trade partners 426, 442 04 village communities 130-31, 151 TSC nhailand Sugar Corporation) 93, 94, 329 Thailand Development Research Institute TSTC (Thai Sugar Trading Corporarion) 94 (TDRI) 339 Tsur, Y. 407 Thailand Sugar Corporation (TSC) 93, 94, 329 Thai Maizs Importers' Council (TMIC) Up) 323 Undang-UndangPokok A p r i a (Basic Agrarian Thai Pineapple Canning Industry 334 Law) (Id) 185 Thai Produce ExportersAssociation (TPEA) 323 Unger, D. 80 Thai Sugar Trading Corporarion (TSTC) 94 United Kingdom (see Great Britain) Thaksin Shinawatra 295 United States (US) Thanin government (Th) 349 colonial policies 32, 266, 267 Thanin Kraivixien 305 pineapple trade 333, 334, 335 Theil, H. 523 protection policies 407 Timmer, C.P. 50, 65, 381, 382, 389 rice trade 315, 316, 317-18 TMIC (Thai Maize Importers' Council) Up) 323 trade with SE Asia 425, 426 tobacco industry urban areas (Ph) Indonesia 194, 195, 197, 450 migration to 290-93 Philippines 256, 257, 259, 423 poverry in 228, 234, 278-79 Thailand 301, 360 social indicarors 237-38 Todaro, M. 523 urban bias xxiv 558 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH Valdes, A. 380, 382, 390-91, 433 Thailand 346 Valsan, E. 137 Watkins, M. 64 van Niel, R 173 West Irian (see Irian Jaya) vegetable crops wheat flour (see flour industry) Indonesia 192 wheat industry Thailand 360 Indonesia 97, 99, 100, 101 vent-for-surplus development Philippines 106, 250 bases of 16, 24-28, 43 Why Poor People Stay Poor (Lipton) xxiv impact on agrarian structures 28-29 Widjojo Nitisastro 98 Thailand 64 Wolff van Westerrode, l?D. 155 theory of 64 World Bank Vietnam 25, 315, 316 classification of economies 201 village rice banks (lumbung desa) 146, 147, East Asian 'economic miracle' 381 155 education-income relationship 353 village unit cooperatives (Koperasi Unit Desa) fertiliser policy 191 145 on Indonesian economic growth 164 minimum access volume (MAV) system wage rates 253-54 comparative data 464, 465-66, 486 poverry measurement 236 Philippines 262-65, 281 R&D programs, Thailand 303, 340 Thailand 495 on Thai governance 80 Wardhana, Ali 98 World Trade O H t i o n (WTO) 243, 253, 280 water facilities Philippines 237-38 Yoshihara, K. 268 RURAL DEVELOPMENT AND AGRICULTURAL GROWTH in Indonesia, the Philippines and Thailand Understanding econoniic growth is central to the study of development. Rural economic growth is an important aspect of economic growth. Historically,rural agriculture has employed most people in most countries, and continues to do so today. Nevertheless, the causal relationships between economic growth and growth in agriculture remain poorly understood. This volume focuses on economic growth in the agricultural sectors of Indonesia, the Philippines and Thailand. Starting from similar positions, the agricultural sectors of these economies ave diverged considerablyover the last 40 years. '1 The volume investigatesthe ways in which policy, institutions, investments, resource constraints and the reallocation of agricultural labour have driven this divergence. The volume documents the interplay of endowments, technology, the accumulation of productive factors, policy, and advocacy in the rural sectors of these three countries. Each contributes in its own way to an explanation of the past. Good policy rests on an understanding of successes and failures in the past. This book is a critical contribution to such an understanding. Takamasa Alayarna 8cDonald I? Larson(eds)