India - 1991 Country economic memorandum (Vol. 2 of 2) : Agriculture : challenges and opportunities

Report No. 9412-IN India 1991 Country Economic tAemorandum (In Two Volumes) Volume II: Agriculture--Challenges and Opportunities August 23, 1991 Agriculture Operations lDivision India Country Department Asia Region FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY AND OTHER EQUIVALENTS Average Exchange Rates The currency is the Indian rupee. Average exchange rates per US$1: 1985/86 Rs.12.237 1986/87 Rs. 12.787 1987/88 Rs. 12.968 1988/89 Rs. 14.476 1989/90 Rs. 16.663 1990/91 Rs. 17.949 End March 1991 Rs. 19.620 End June 1991 Rs. 21.142 End July 1991 Rs. 25.773 In this report, a rate of Rs 17.90/US$1.00 was used for 1990/91. The Indian fiscal year runs from April 1 through March 31. The Indian agricultural year runs from July to June. Kharif is the wet season that runs from June to mid October. Rabi is the dry season that runs between mid October and mid February. Weights and Measures 1 hectare (ha) 2.47 acres 1 kilometer (km) = 0.62 miles 1 metric tonne (mt) = 1.1 short ton 1 kilogram (kg) = 2.2 lbs 1 quintal = 100 kilograms =220.5 lbs 1 lakh - 100,000 I crore = 10 million  FOR OFFICIAL USE ONLY AckuQ4g MCnts This volume was written by Pamela Cox, based on contributions by Deepak Ahluwalia, Sajitha Bashir, Lynn Bennett, Kanchan Chopra (consultant), A. Desai (consultant), Robert Evanson (consultant), Mario Fischel, David Greene, Stephen Haggbladde (consultant), Peter Hazell, Kundhavi Kaderisan (consultant), Cheik Kane, Richard Lacroix (consultant), Samuel Lieberman, Manoshi Mitra, Keith Oblitas, Kirit Parikh (consultant), Carl Pray (consultant), R. Radhakrishna (consultant), V.J. Ravishankar, H.L.S. Tandon (consultant) and Ravi Venkataraman. Kundhavi Kaderisan provided support for research and statistical analysis. Suggestions and helpful comments from David Greene, Oktay Yenal, Jan Wijnand, John Joyce and Hans Binswanger, as well as many staff of the India Department, are gratefully ack-awledged. The team is also grateful for the comments provided by GOI and state government officials, and participants in a series of seminars held on the report in India. The analytical work for this report has been undertaken over a period of eighteen months. The most up-to-date statistics available were used. But in many instances, lags in statistical reporting in India, especially collection of disaggregated data, has meant that the report has had to rely or, data three or four years old. The report was completed and discussed with the Government prior to the recent budget presentation. The Government of India has made several important policy changes in the last few months. To the extent possible, these have been reflected. I This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  DMIA COLIRY ECONOMIC MEMORANDUM TakLe of Contents ExecutiveSummary .............. i-vi I. AGRICULTURE'S PERFORMANCE AND POTENTLAL 1. A. Agriculture's Role in the Economy . Poverty and Employment ........................., 3 Ruml - Urban Growth Linkages .9 EnvironmentalIssues .10 B. Perfonnance and Sources of Growh.. 14 Aggregate Production Growth .14 Performance of Principal Crops .18 Sources of Growth .21 C. Future Growth Oppounities and Chalenges ...................... 22 Constrants ........................................... 22 Govemnment Development Strategy ........................... 22 OpporAnitiesandAdjustments .............................. 23 II. KEY ISSUES FOR GROMTH .................................... 26 A. Pricing and Trade Policies .................................. 26 Price and Market Intervention Progmms ........................ 26 Input Subsidies ........................................ 27 Agriculturl Trade ....................................... 32 Inpact and Policy Issues .................................. 33 B. Investent .49 Public Investment .49 Private Investment and the Role of Agricultural Credit .55 C. Technology . ........................................... 62 Fertizer ............................................. 64 Iinproved Seeds ............. ........................... 66 Looking to the Future: The Avaiiability of New Technologies ..... ..... 69 India's Research System and the Generation of New Technologies .... .. 70 Extension . ........................................... 73 Land Issues and Technology Adoption ......................... 73 D. Infrastrture ............................................ 74 Irrigation ............................................. 74 Rura Infrastructure ...................................... 83  III. STRATEGIC CHOICES FOR LONGTERM AGRICULTURAL DEVELOPNEW1T ............................................ 86 A. StategicFramework ....................................... 86 B. An Adjustment Program for Agriculture ........ ................. 88 Pricing and Trde Policy . .................................. 88 Investment ............................................ 95 Technology Adoption and Innovation .98 Irrigation .99 Timeframe for Adjustnent .101 MATRDI ................................................... 102 ANNEXES: I. Issues in Input Supply and Distribution .110 II. Land Issues and Technology Adoption .115 List of Background Papers ................. .8 Bibliography ................ 119 Stitistical Appendix ................ 126 List of Tables in the Main Text Table 1.1 Growth Rates of Output, Inputs and Total Factor Productivity, 1956-1983 Table 1.2 Labor Use for Major Crops Table 1.3 National Agricultural Output and Income Multipliers for Irrigated versus Rainfed Agriculture Table 1.4 Trend Growth Rates of GDP Originating in Agriculture Table 1.5 Relative Importance of the Main States in India Agriculture and the Relative Importance of Agriculture within Each State Table 1.6 Trend Growth Rates of Production of Principal Crop Groups at the All-India Level Table 1.7 Trend Growth Rates in Area, Production and Yields at the All-India Level Table 1.8 Sources of Growth of Principal Crops Table 1.9 Demand Projections for Food Items Table 2.1 Agricultural and Food Subsidies, 1989-90 Table 2.2 Average Economic Subsidy on Fertilizers Table 2.3 Real Economic Cost of Input Subsidies in India Agriculture, 1980/81-1986-87 Table 2.4 Irrigation Charges, Gross Return and Net Return for Canal Irrigation Table 2.5 Barter and Income of Trade for Agriculture Table 2.6 Impact on Cereals Prices of Uncovered Shortfalls in Requirements Table 2.7 All India Consumer and Food Subsidy Table 2.8 Access to PDS by Income rGroup Table 2.9 Distribution of Subsidies by States, 1980-81 to 1986-87  Table 2.10 Protection Coefficients of Selected Commodities Table 2.11 Distribution of Foreign Exchange Loss Implied by NPCs, 1980/81 - 1985/86 Table 2.12 Divergence between Domestic and Effectist Incentives (AU India) Table 2.13 Agriculture's Share in Total Plan Outlays Table 2.14 Central and State Spending for Agriculture in 1988/89 Table 2.15 Grm -*h of Real Expenditure in Agriculture, 1974/78 to 1987/88 Table 2.16 Cost Recovery Rates in Four States Table 2.17 Estimates of Components of Private Fixed Capital Formation in Agriculture Table 2.18 Composition of Term Loans for Agriculture Table 2.19 Relative Shares of Different Regic;as in Direct Finance of Commercial Banks and Co-operative Societies, 1982-83 Table 2.20 Distribution of Credit by Commercial Banks and Co-operatives According to Size of Holdings by States, 1985 Table 2.21 Real Investment Credit Per Thousand Hectares of Net Sown Area Table 2.22 Interest Rate Structure for Loan Advances of Commercial Banks Effective September 1990 Table 2.23 Indicators of the Spread of Improved Agricultural Technology for India, 1960 to Present Table 2.24 Level of Yields Obtained by Paddy and Wheat Farmers Table 2.25 Reported Fertilizer Production, Imports, Movements and Stock Levels: 19.0- 1986 Table 2.26 Coverage of HYVs by Region 1987/88 Table 2.27 Seed Requirements and Commercial Supply for Major Field Crops, 1985 Table 2.28 Real Growth in Agricultural GDP and TFP of Crops and Allocation of Research Expenditure Table 2.29 Development of Irrigation: 1950-1985 Table 2.30 Changes in IrTigated Area and Cropping Intensity During 1960-63/1980-83 in Major States Table 2.31 Road Length by Road Category and Surface Type, 1988 (kan) Table 2.32 Roads by Region in India Table 3.1 Input Subsidies and Farmer Incentives Table 3.2 Estimated Fertilizer Price Increwae to Reduce Sul List of Fig=es in the Main Text Figure 1.1 Weights of Major Crops in Agricultural Production in India Figure 1.2 Land Use in India, 1950/51 - 1986/87 Figure 1.3 GDP from Agriculture and Rainfall Figure 1.4 Sectoral Shares in GDP, All India Figure 1.5 Growth of Area, Production and Yields of Foodgrains, 1950/88 Figure 2.1 Real Economic Cost of Input Subsidies (at 1980-81 Prices) Figure 2.2 Change in Real Wheat and Rice Prices Figure 2.3 Comparison of Domestic & International Rice and Wheat Prices Figure 2.4 Central and State Spending for Agricultre in 1988/89 Figure 2.5 Growth of Real Spending on Agriculture, 1975/76 to 1987/88  Lst of es in the Main Tx Box 1.1 The Experience of Punjab Box 1.2 Problems of LIgging Growth in Eastern India Box 1.3 Women and Agricultural Employment Box 1.4 What is the Growth Rate of Agriculture in India Box 2.1 Agricultural Subsidies Box 2.2 Gross Domestic Capital Formation in Agriculture Box 2.3 Rural Roads and Agricultaual Development  CONR EOO MEMORAb0DUM Abbreviations ACRC Agricultural Cedit Review Committee CACP Commission on Agricultural Costs and Prices CAD Commz-- Area Development CAN Calcium Ammonium Nitrate CCI Cotton Corporation of India CSO Centra Statistical Organizaiion EPC Effective Protection Coefficient ESC Effective Subsidy Coefficient ERR Economic Rate of Return FCI Food Corporation of India FAI Ferdlizer Association of India GDP Gross Domestic Product GDCF Gross Domestic Capital Formation GOI Govemment of Tndia HYV High Yielding Variety ICAR Indian Council of Agricultural Research IRDP Integrated Rual Develo-pmnt Pogamme IRRI International Rice Research Insitutute ISMA Indian Sugar Mills Association K Potash MOP Murate of Potash N Nitrogen NABARD National Bank for Agricultume and Rural Development NCAER Nadonal Council of Applied Economic Research NPC Nominal Protecdon Coefficient NSS National Sample Survey OGL Open General License O&M Operadons & Maintenance P Phosphorus PAC Primazy Agriculture Cooperative PDS Public Distribudon System R&D Research & Development RBI Reserve Bank of India SSC State Seed Corporadon TFP Total Factor Producdvity  CUTIE UK, AgSdbuMftd=ance and Problems 1. Indian agriculture has achieved impressive success in raising foodgrain output. The introduction of high-yielding seed varieties (HYVs) for major foodgrains, expanded use of ferilizer and extensive investment in irrigadon (the "Green Revolution') has made India largely food self-sufficient. This grawth has been achieved despite :onsiderable discrimination of the uhde and price regime against agriculture. Yet overall, agricultural GDP growth, 2.6% p.a. since 1965, has been modest, compared both to performance in other Asian countries and to India's population growth rate of 2.1% p.a. in the same period. Despite major technological breakthroughs, the all-India agricultural growth rate has not accelerated in the past two decades. Growth remains concentrated in a few major crops (mainly rice, wheat, sugarcane and rapeseed/mustard), in a handful of regions (mainly in the north), and in irrigated areas. There is incrasing concern that without shifting Indian agriculture to a higher growth path, India will not be able to meet many of its development objectives. 2. For many reasons, agriculture policy and strategy need revision. The current macroeconomic situation has heightened awareness of the verv significant subsidies to agriculture, but even in the absence of fiscal constraints, it is evident that agricultural policies need adjustment. Even moderate growth in the sector has come at a high cost, both to the budget and to the economy. As the limits to growth from Green Revolution technologies are approached, there are no new technologies on the horizon. The scope for relying on expansion of irrigation is smaller than in the past: much of the high potential and easily irrigable land has been developed, and new schemes are cosdier and more difficult to implement. The gains from growth have not been spread evenly, and there has been a relative neglect of rainfed agriculture, diversificadon and trade opportunities. These will be especially important, as rising incomes and rapid modernizadon in India are shifting the structure of demand toward a wider array of commodities. Serious physical resource constraints dictate that future growth will come from intensifying producdon, increasing efficiency and shifting to higher-value crops. Government's development strategy, which has been heavily dependent on extensive public direction and involvement in the sector and concentration of resources on selected crops and regions, provided a weak foundation for growth. 3. The choices India makes in revising agricultural strategy and policy in response to the current crisis have long-term implications for the pace and patterns of future development. Cutting subsidies and restraining public investment at the margin will not address the underlying and fundamental problems constraining growth. These center on the nature and extent of Govemment's role in the sector, on increasing efficiency in India's resource-poor environment, and on improving equity and distribution of the benefits of growth. A framework for sectoral policy reform will include four main elements: (i) improving the structure of incentives; (ii) mcreasing the efficiency of public investment and promodng greater private investment; (iii) encouraging technology adoption and generation; and (iv) improving the performance of irrigation. Issuesand OpQns 4. Structure of Incendves. The strategy of agricultural development in India and the related policies which have evolved over the past decades have reached a stage which necessitate a fundamental review. This necessity is underlined by the clear realizadon that the heavy fiscal subsidies to agriculture are not sustainable. Seemingly less urgent but more important for long- tm growth is the viability of the sector as an efficient and dynamic engine of growth in the India economy. Analytically, the pattern of incentives and their impact on resource allocation ;ie at the heart of both questdons. Reducing agricultural incentives through general policies and trying to compensate through fiscal transfers is both ineffective and excessively costly.  5. The existing incentive structure has had both positive and negative effects on producers, but overall the impact has been negative. Heavy protection of industry and trade controls have led to considerable discrimination in the trade and price regime against agriculture in India. Agriculture has an effective protection coefficient of 0.87, compared to 1.4 for industry. Industrial protection has kept the costs of manufactured inputs and consumer goods high, and discriminates against agricultural exports. Export restrictions have prevented producers from benefitting from higher international prices. World prices for rice and cotton have consistently been higher than domestic prices, while domestic prices for oilseeds and sugar have tended to exceed international prices. On the other hand, farners who use improved inputs have benefitted from fiscal subsidies and producers of oilseeds and sugar have benefitted from import icstrictions and closed domestic markets. Protection and subsidies have had high fiscal and economic costs (misallocation of resources, losses of efficiency and foreign exchange, and foregone trade opportunities). Protection is high for some crops in which India has limited comparative advantage (oilseeds and sugarcane), while crops India can produce competitively (rice, wheat and cotton) are disprotected. Estimates of the net foreign exchange loss from import substitution for oilseeds and agricultural trade controls range up to US$3.4 billion per year. Government involvement in output marketing and trade controls has also absorbed large financial and administrative resources, and proved a disincentive to producers and exporters. Agricultural trade has declined sharply in the last decade, and now accounts for only 18% of exports. 6. Subsidies for food and agricultural inputs totaled Rs. 135 billion annually in 1989/90 (US$ 7.6 billion or 3.1% of GDP) and exceeded public investments in the sector (Rs. 110 billion). High input subsidies to agriculture - for irrigation, electricity, fertilizer and credit - have had little impact on protection, primarily because they are concentrated in certain regions, go mainly to wealthier farmers, and include transfer payments outside agriculture (half the fertilizer subsidy goes to the inefficient domestic fertilizer industry). Subsidies are crowding out productive investments in irrigation and power infrastructure essential for future growth. Electicity subsidies total Rs. 41 billion alone, and, since they are generally not covered by the budget, are a major factor in the poor financial performance of most state electricity boards. Food subsidies (through the Public Distribution system, PDS, and state programs) are also increasing rapidly. Although counted as agriculture subsidies, food subsidies are essentially consumer subsidies (about 20% goes towanrs FCI's stocking and procurement program). PJ'S neverthelcs3 L= aq important indirect influence on procurement programs: many of the more resjrictive marketing measures (levies and monopoly procurement) have been instituted in order to supply PDS. Most existing food distribution programs are not well targeted to the needy, as PDS is concentrated in urban areas and an estimated 40Wo of offtake goes to the richer 40% of the population. 7. Adjusting the structure of agricultural incentives requires both short and medium term actions. In the shor term, the fiscal crisis dictates immediate reduction in subsidies. Recommended actions include phased increases in electricity tariffs and fertilizer prices over three to four years. Food subsidies can be restrained through better targeting to the rural and urban poor, particularly through elimination of sugar and more distribution of coarse grains, but it is important that an adequate safety net be retained. Adjustments in industrial protection and trade policy, which have already begun, will also improve agricultural incentives to the extent that imbalances in protection are conrected and agriculture trade (both imports and exports) liberalized. 8. Reordering incentives also involves review and restructuring of the entire complex policy edifice of Government programs and interventions in the medium txjm. High levels of protection for oilseeds and sugarcane need to be reduced, and incentives increased for rice, cotton and wheat production. This will require in the first instance reorientation of public procurement and price support programs, to define objectives better, reduce coercive market and trade controls, improve tthe nutritional impact of PDS (while reducing administrative costs), and redefine the role of the Food Corporation of India (FCI). Second, restructuring of the fertilizer industry is vital, as it is unrealistic to expect Indian farmers to pay for high protection and inefficiency. Recommended changes include linking fertilizer prices to long-run international marginal production costs,  liii gradually opening the fertilizer industry to competition frem trade, closing inefficient plants, and reducing Government's role in fertilizer production and distribution. Third, removal of credit and irrigation subsides will require longer term programs to address the underlying causes of low cost recovery and poor financial performance (paras. 10 and 18 below). Finally, there are clear opportunities to expand agricultural trade in India (cotton and rice are two possibilities), but this will depend on immediate measures to reduce both trade and domestic market controls (quotas, trade bans and monopoly procurement). 9. Investment. The rate of capital fornation in the agricultural scctor has been falling, a worrisome trend for future growth. Much of this decline has been due to declining real public investment, which fell by 18% in the 1980s. Private investmnent, which accounts for two-thitds of capital formation in agriculture, has also stagnated in recent years, even in the face of growing credit subsidies. Pattems in public investment at both the central and state level (states account for over two-thirds of agricultural investnent) show falling overall allocadons to agriculture, growing recurrent expenditures, a crowding out of productive investment by subsidies and transfer payments, and falling real investment. In the 1980s, recurrent expenditures in the states grew 3% p.a. in real terms, while capital expenditures f=l by 1.8% p.a. Declining cost recovery and burgeoning personnel expenses have led to runaway rect;._ .t expenditures. Private investment, concentrated in groundwater development and machinery, complements public investment, and thus slowdowns in public capital formation are contributing to stagnating real private investment. 10. Problems in the agricultural credit system may also be constraining investmnent. Policies designed to expand the coverage and disbursement of agricultural credit are now beginning to undermine the agricultural credit system, causing very real concerns about its future viability. Large loan write-offs, mandatory lending, inadequate interest rates and margins and the unviable spread of bank branches have strained the system to the point of near collapse. Despite Government policies, credit use remains concentrated in certain groups of farmers and in certain regions. Most farners use no formal credit at all: only 27% of India's fanners use cooperative credit, and a mere 4% use commercial bank credit (although indications are that informal credit flows remain substantial). Credit subsidies, stemming both from low controlled interest rates and from high loan default rates, have primarily gone to larger and wealthier farmers. Performance of credit institutions is weak: overdues are high (over 40% of debt service), loan recovery is poor, many institutions are financially unviable, and there is a high level of political interference. 11. Arresting the decline of investmnent in agriculture requires broad changes in patterns of public investment, and resolution of the agricultural credit crisis. In the shAr term, the most urgent need is to reduce subsidies and increase cost recovery, although total elimination of subsidies will require adjustments in related programs (para 8). Review of the composition of agricultural investment at both the central and state levels is also needed as part of the process of rationalizing public expenditures, eliminating the spreading of scarce resources over myriad small programs, and targeting priority activities. Areas in which Government has a clear role and which merit priority attention include research, irrigation (consistent with a shifting of priorities within the subsector, para 18), resource management (forestry, land reclamation and management, drainage, etc), rural infrastructure, and training and education. The third short-term priority is to start to resolve the agricultural credit crisis, within the context of general reforms to the financial system. This will involve consolidation of rural banking, reducing interference, changes in interest rates to eliminate subsidies and make credit more viable, reduction in targeted lending, and institutional rehabilitation (which will be a long-term process). In the miium tenrn, recurrent expenditures must be controlled both through improvements in cost recovery and through devolving activities to the private sector (seed supply, animal health, tubewells are examples). Changes in the structure of incentives coupled with on-going programs to strengthen the rural financial system will help promote private investment.  iv 12. IechnogI . India's Green Revolution is eloquent testimony to the impact of technology on agricultural growth and development. There remains scope for gains from currently available technologies. Although fertilizer use has increased rapidly since 1970, per hectare consumption (55 kg/ha) remains low by regional standards, and fertilizer use is heavily concentrated in irrigated areas and on four crops (rice, wheat, sugarcane and cotton). Government policies to encourage fertilizer use (high subsidies and extensive involvement in production and distribution) have had less impact than investment in irrigation, and in the case of marketing, may actually be restricdng distribution and supply of appropriate fertilizer types. The Government has also been extensively involved in the production and promotion of improved seeds, although there is no evidence to confirm that this has led to wider adoption. Improved seeds now cover over 90% of wheat area, 50% of rice area and one-third of coarse grains area. Lack of H YVs tailored to India's vastly different agroclimatic and production conditioii- is constraining adoption of rice and coarse grains varieties, underscoring problems in the research system in vnendting suitable varieties. 13. It is not evident that Indch? s research system is gearing up for the technical challenges of the coming decade. The,re is a growing pcrception, shared by many scientists in India. that the agricultural research system is lagging and in particular not responding to the increasinglv varied and complex needs of the agricultural economy. Returns to research remain high (although there have been some declines in research productivity). Problems relate to the content of the research system, primarily its excessive focus on plant breeding for foodgrains and relative neglect of other crops and farm management; its quality; imbalances in allocation of funds (eastern India in particular has received very low allocations) and failure to set priorities; and poor organization and management. There is a dearth of good and relevant messages coming from the research system, as evidenced in lagging adoption rates for new seeds. There is, however, considerable scope for private research to expand and fll gaps in the research system, not the least in high technology and basic science. 14. Adjustments needed to encourage technology adoption and innovation focus largely on medium term programs to revitalize the research system, exploit the unrealized returns from existing technologies through wider adoption, increase attention to resource and environmental issues, improve extension and address land issues in eastern India. The first priority is to implement far-reaching changes in the research system. Resources need to be reallocated away from a concentration on foodgrains and plant breeding, and towards a wider range of commod;ties and issues (particularly land management, farming systems and agroforestry, among others). The scope of research must be broadened to include more basic resexrch on one har,. and more on-farm research on the other, and to shift resources toward programs relevant for neglected regions. Changes must be made in the organization and management of the research system to increase quality and incentives. One of the most important needs is to open up the research system to international exchanges and technologies, and to encourage technology imports, particularly in basic science. Policy changes are also needed to encourage more private research, for example through protecdon of intellectual property rights and immediate relaxation of controls on technology imports. More attention must be paid to ensuring sustainable use of India's physical resources, through changes in the structure of incentives (to prevent overexploitation), introduction of techniques to reduce land degradation, increased investment in forestry, and attention to social issues contributing to environmental problems. Here more work needs to be done (and is already und.rway in forestry) on devising clear strategies and action plans. Land issues (tenure status, land fragmentation) are constraining investment and innovation in eastern India, and steps should be taken to increase access to land, consolidate holdings and improve security of tenure. 15. Further adopdon of existing technologies will depend on several factors, most importandy expanded investment in inrigation and development of seed varieties and recommendations tailored to agroecological conditions. Govemment's extensive role in input production and marketing also needs to be considerably reduced to promote long term development of efficient input supply channels and encourage private sector involvement in research and marketing. Consistent with programs to restructure the fertilizer industry, Goveriment control of distribution should be eliminated, particularly fertilizer allocations, licensing and price controls. The recent reduction in  v subsidy and removal of piice and movement controls on low analysis ferdlizer is a first step in this direction. In seed supply, involvement of state seed companies in commercial seed production should end, and they should concentrate on producing foundation and breeder seed, and on building up seed supply where technologies are new and market imperfections exist (pasture and fodder seed is one example). AU seed subsidies should be eliminated. 16. Ijrigation. The future of Indian agriculture is inextricably linked with the development of irrigation, but in many ways the sector is a paradox. On the one hand, irrigation development has been the major determinant and source of past agricultural growth, and future growth and diversification will depend on existing and expanded irrigation infrastructure. Irrigation development has had important secondary effects on protecting the environment and increasing rural incomes and development. On the other hand, the subsector has severe problems in almost every aspect of implementation and performance: resource planning is weak, design is often poor, construction quality is a major problem in most states; operation and maintenance are below acceptable standards; and cost recovery is virtually non-existent. Potential for new irrigation is dwindling (India has developed nearly two-thirds of estimated potential), costs are escalating, and rates of return on projects have been below expectations. There is considerable scope for further exploitation of groundwater, however, and private tubewells have high rates of return, even without the large electricity subsidies. 17. Major efforts are needed in four key areas: (i) forging a coherent water policy; (ii) setting investment priorities and regaining control of public expenditures; (iii) improving system productivity and assuring sustainability; and (iv) building institutional capacity within both the public and private sectors to manage irrigation more efficiendy and effectively. 18. First, planning and coordination of water use must be done on the basis of the natural hydrological unit, the river basin. As river basins cut across states in nearly every instance, this will require resolution of difficult interstate issues, as well as conflicting demands for water use. The existing National Water Policy needs to be applied widely, complementary state water policies drafted, and river basin commissions formed. Second, the investment focus for irrigation should concentrate on, in order of priority, improving the performance of existing systems, finishing the large backlog of uncompleted projects, and investing cautiously in new construction (including flood control, drainage, storage, reclamation and rehabilitation). Runaway public expenditures need to be reined in, especially for recurrent costs, together with measures to fund operations and maintenance (O & M) fully, increase water rates, and improve cost recovery. Third, improving water management should be the principal technical goal for irrigation managers, focusing on low- cost, command-specific actions. Construction quality also needs improvement, particularly through much dghter supervision and reducdon in corruption. Greater attention must also be given to environmental and resettlement and rehabilitation issues in planning and implementing irrigation investments. Finally, there is scope for harnessing the tremendous energies and capital - both human and investment - of the private sector, through devolvement of activides to farmer groups, nongovernmental organizations and contractors. At the same time, the Government will stll play an important role in guiding and often implementing irrigation, and thus irrigation institutions must be strengthened, and their accountability to their clients - farmers - increased. 19. Conclusion. Our dialogue with Government about constraints to the accelerated development of India's agricultural sector has identified a significant number of desirable policy adjustments. Given the importance of agricultural sector development for the adequacy, equity and sustainability of the Government's overall development effort, adjustments in agricultural sector policy deserve to be pursued with priority in the context of the Government's broader program of stabilization and reform. 20. Postscrip This report was completed and discussed with the Government of India prior to its presentation of the 91/92 budget to Parliament. The following are the main policy changes arising  vi frm the budget and from associated policy statements which are likely to have a significant impact on agriculture. They are expected to have positive net effects on agricultural incentives and also to be significant in correcting skewed intersectoml incentive patterns: Fertilizer prices will be raised by an average of 30%, except for poor and marginal farmers. Price and movement controls are to be abolished for low analysis fertilizer. Farmers are to be compensated by higher procurement prices, which presumably would be passed on to consumers. This goes some way towards reducing the very large subsidy bill. * The new industrial policy, which has gone a long way towards deregulating entry and opening the economy to compedtion, will benefit agricultural industries over the medium to long term. * The new trade policy which, alongside the changes in exchange rate, includes abolishing export subsidies, and expanding and simplifying of the replenishment license scheme, should also benefit agriculture over the medium to long term and will significandy improve the incentive for agricultural exports. e The subsidy on sugar has been eliminated thereby correcting some of the distordons oudined in paragraph 3.24 of the main reporL * The establishment of a high level coanmittee with the mandate to review all relevant aspects of structure, organization, functions and procedures of the financial system bodes well for the strengthening of the agricultural credit system recommended by the Agricultural Credit Review Commission.  CHAPTER I. AGRICULTURE'S PERFORMANCE AND POTENTIAL 1.1 For the majority of people in India, what happens in agriculture directly shapes their daily lives, and their hopes and prospects for a better life. This has been true for centuries, and is likely to remain so well into the next decades, even with India's notable progress in industrialization. What happens in agriculture also influences how India will tackle the problems of poverty, employment and environment which dominate the development agenda. Faster economic growth presents new opportunities and new challenges for Indian agriculture. Rising incomes and rapid modemization are shifting the structure of demand for agricultural commodities and opening up new possibilities for domestic and international trade and linkages to the rural non- farm economy. Yet at the very time when agriculture is called upon to make a greater contribution to economic growth, there is growing concem that this challenge cannot be met. What is the scope for future rapid growth of the economy if agriculture continues to grow only moderately? And even if India could raise its GDP without accelerating agricultural growth, what would the impact be on the distribution of the benefits of growth? 1.2 Agricultural policy is in crisis. The current macroeconomic crisis has heightened awareness of the very high and unsustainable levels of subsidies to the agricultural sector, but even without the short-term immediacy of restraining Government spending, it is evident that agricultural policies need adjustment. Overall agricultural growth has remained moderate, and India has yet to witness the accelerating agricultural growth rates that accompany structural transformation, despite major technological breakthroughs in the Green Revolution. Past sources of growth are petering out. Moreover, the gains from growth have not been spread evenly, but have been concentrated in certain regions and certain crops. Thus, although famines have largely been eliminated, hunger remains a persistent problem for nearly 20% of Indian households. There has been a reladve neglect of rainfed agriculture, diversification and trade opportunities. Official development strategy has been based on heavy and extensive Government involvement. The legacy of this is now haunting the sector, as evidenced in the high and unsustainable subsidies, significant resource misallocations, and declining agricultural trade. At the same time, as populatdon continues to grow at 2% a year, agriculture faces formidable resource constraints: land and water supplies are limited and environmental pressures are increasing on both marginal and cropped lands. 1.3 There is an urgent need for a fundamental review of agriculture strategy, to address both the short-term exigencies, and the medium-term issues of growth and efficiency. This report examines the potential of agriculture to make a larger contribudon to India's development, and the factors currently constraining sectoral performance. It is not a comprehensive overview of the agriculture sector in India, and thus does not cover every subsector and issue. Rather, the report focuses on four areas where adjustments are critical for growth: pricing and trade policies, investment (public and private), technology, and infrastructure. This chapter looks at agriculture's contribution to solutions of selected development problems, and draws lessons from past performance for future strategy. The second chapter reviews key issues, and the final chapter articulates a strategy and oudines the adjustments needed for accelerated agricultural growth. A. Agriculture's Role in the Economy 1.4 India's agricultural sector is as rich and diverse as any continent's. Its climatic zones vary from the humid subtropical in the south to the semi-arid in the north. Production systems span a condnuum from tradidonal peasant agriculture to intensive drip irrigadon of high- value fruits. Average technical levels are low by intemational standards, although use of modern  2 inputs is growing rapidly. Although less than one-third of the cropped land is irrigated (about two- thirds of the estimated potential), irrigation development has made a major contribution to growth. Small farms predominate and fann sizes are declining (average farm size was 1.84 ha in 1980). There is a growing number of landless workers. Foodgrains dominate production, accounting for 63% (Figure 1.1); India also produces a wide range of nonfood commodities. The take-off of the Green Revolution after the mid- 1960s has allowed India to become self-sufficient in foodgrains in normal years and recently to accumulate sizeable grain stocks. Agricultural imports are now marginal, and thus the sector feeds and clothes over 800 million people. Growth patterns have differed sharply among states and regions, in particular the booming northwest in contrast to the lagging east. The sheer diversity of producdon systems, technical levels, agroclimatic conditions and infrastructure throughout the country make it difficult to generalize, and mean that the national picture of what is happening in agriculture often obscures much of the story at the farn level. Figure 1.1: Weights of Major Crojs in Agrcultural Production in India Oilseeds 12.64% Rice 29.74% Others 11.27% Pulses 7.94% Cotton 4.37% Sugarcane 8.11% Wheat 14.45% Coarse Cereals 10.79% Jute & Mesta 0.69% 1.5 Agriculture's share of GDP (33%) masks the extent to which India's people depend on the sector. India remains a primarily rural society: about 75% of the population live in rural areas. Two-thirds of all Indians depend directly on agriculture for their livelihoods, and many others indirectly due to strong rural-urban linkages and agriculture-based industries. Workers employed in agriculture are the poorest (80% of India's poor live in rural areas) and a high proportion are female (one-third of the agricultural labor force are women, and agricultue remains by far the major employer of rural women). Agriculture has an important impact on present macroeconomic problems: subsidies for agricultural inputs amount to 2.5% of GDP, while declines in already modest agricultural exports (agriculture now accounts for less than 18% of exports) in the last decade have exacerbated the tade deficit  3 1.6 There are growing concerns that agriculture will not be able to increase its contribudon to overall GDP growth, nor sustain its centWal mle in resolving several problems high on the agenda of social development. Although the share of agriculture in employment and output has been falling, consistent with expected patterns of development, India has yet to experience the surge of agriculturl growth rates that accompany and support faster industrial development. There has been no statistically significant change in the all-India trend rate of agricultural growth even after the Green Revolution two decades ago (para. 1.31). Yet agriculture's impact on overall growth and development patterns can be powerful, either posiively, as evidenced in India's Punjab (Box 1.1) and/or negatively, as in the impact of lagging agriculture in the very poor states of eastem India (Box 1.2). In India, agriculture's role is also prominent in addressing several key development issues: poverty and employment, women in development, rural non-agricultural growth and the environment. Povery and Em2loymentI 1 1.7 India has made progress iti reducing the level of poverty. In the 1970s and early 1980s both real wages and real expenditures by the poorest groups increased and the percentage of the population below the poverty line fell. Yet poverty remains a very real and pressing social problem in India: even though the percentage of poor has fallen, the absolute numbers of those living in poverty have risen, and poverty has increasingly become concentrated in regions (eastem and central India) with serious development and financial constraints, and in rural areas (80% of the poor are agricultural workers). Two-thirds of India's poor (and nearly 70% of the ultra poor) in 1983 lived in states with low or moderate rates of agricultural growth combined with medium or high levels of production instability. Government strategies for labor absorption and poverty alleviation have thus concentrated in large part on expanding rural - and especially agricultural - employment opportunities, and in contending with lagging economic and agricultural growth in the eastern and central states. 1.8 That agriculture can play an important role in poverty reduction is clear, yet agriculture by itself cannot carry the burden of coping with India's poverty problem. Modelling of the agriculture sector shows that no agricultural development measures can affect rural labor demand or wages as much as changes in population growth or growth in demand for non- agricultural labor (Quizon and Binswanger, 1986). Measures to create jobs in other sectors of the economy are vital. The need is all the more pressing because despite slowing population growth rates, the sheer numbers of new labor force entrants will remain high for the next 20 years. These people already exist in India, and plans must be made to increase employment generation radically to cope with this problem. 1.9 At the same time, agriculture will remain the major source of new jobs for at least the next decade, even at high industrial growth rates, due to the relatively small industrial base and currently high capital:labor ratios. What is the prognosis for increased labor absorption by agriculture? In the short-term, the number of jobs in agriculture is likely to grow, although agriculture as a proportion of total employment will fall. The most recent rounds of National Sample Survey (NSS) data show that between 1977,78 and 1983, about 15 million agricultural jobs were created, more than in any other sector. In the same period, however, agricultural employment fell from 68% to 64% of total employment, and agriculture accounted for less than half the total growth of 31 million new jobs. This indicates that labor is moving out of agriculture, consistent with its falling share in GDP and the process of structurl transformation. llssues of poverty and employment were discussed in detail in the 1989 India Countr Economic MemdWL (World Bank, 1989), from which this review draws information.  4 BOX 1.1I The_ E re ou A particularly interesting model of the potential of agriculture in the process of structural transformation is the Punjab (G.S. Bhalla et.al. 1990). Punjab is one of the fastest growing and richest states in India. In the period 1960/61 to 1983/84, the primary sector in Punjab grew at an annual rate of 4.1%, while the secondary sector grew at 5.6% and the tertiary sector, 6.8%, compared to all-India annual growth rates of 2.2%, 4.2% and 5% respectively. Although agriculture in Punjab had been growing rapidly since Independence (largely due to the spread of surface and tubewell irrigation and bringing new land into production), production took off after the Green Revolution in the mid-1960s, with unparalleled increases in both output and productivity (which both grew at nearly twice the all-India average rate). Although Punjab (together with Haryana) quickly became the breadbasket of India, due to large yield increases and expansion of area under wheat and rice, it also diversified the agricultural economy as incomes rose and demand patterns shifted. One prominent success was the rapid take-off in milk production, encouraged both by higher incomes and increasing mechanization (which allowed fodder lands p eviously used to support draught animals to be diverted to feeding milk cattle); the livestock sector grew at a rate of 6.9%, and doubled its contrbution to agricultural income, from i4% in 1960 to nearly 30% by 1983. Factors contributing to rapid agricultural growth included attractive new technologies (HYVs and fertilizer), consolidated land holdings, public and private investment in infrastructure (especially irrigation) and rural services (including education and health), and attractive agricultural price policies. The introduction of new rice and wheat technologies quickly led to a predominance of these foodgrains in state production: their shares increased from 30 to over 60% between 1950 and 1984. Fertilizer use increased by a factor of 17, from 9 kg/ha in 1965 to 150 kg/ha in 1984. This led to rapid yield increases of 2.6% a year for wheat and 5.7% a year for rice. Favorable land legislation resulted in the predominance of self-cultivation (74% of land was owner- operated by 1970). The state devoted significant public resources to investrnent in irrigation, power, agriculture and community development, which accounted for nearly 70% of public expenditures on average; public investment was also high in per capita terms. This was complemented by private investment in groundwater development. By the early 198( , over 80% of net area was irrigated, compared to only 27% in India as a whole. Punjab also undertook massive programs of infrastructural development in roads, extension services, agricultural research and education. By 1985 nearly every village was linked by a paved road. The state, as the major surplus producing area of India, benefited significantly from Government price supports. Booming agricultural growth in tum promoted growth in other sectors of the state economy. The increasing use of intermediate inputs in agriculture, coupled with agricultural and income growth, resulted in significant linkages to manufacturing and services. The pattern of these linkages also changed over time: while in 1969170 only agriculture and agroprocessing were generating high forward and backward linkages, a decade later many new sectors joined this group, including some machine-based and metal- based industries, trade, transport, banking, real estate and other services. Due to high growth in all sectors, Punjab today is one of the richest states in India, with the lowest rural poverty ratio (10.87 in 1983/84) and highest per capita income. The structural transfornation of the Punjab economy largely resulted from technological breakthroughs in agriculure and demonstrates the powerful potential of agricultural growth in an agriculture- dominated, labor-surplus economy.  5 BOX 1.2 Prbems of Ictra Growh in ASt India In direct contrast to the successful experience of agriculture in the Punjab is eastem India (the states of Bihar, West Bengal and Orissa). In 1988, eastem India had the largest number of rural people in poverty and the largest proportion of the rural population below the poverty line of any region in India. As in the case of the Punjab, agriculture is the predominant activity in the region - agriculture accounted for 42% of state domestic product in eastem India, compared to an all-India average of 35% in 1985-87. Yet while agriculture has flourished in Punjab and has made major contributions to rural and urban development, the Green Revolution has been slow to take off in eastern India, and lagging agriculture growth has directly contributed to high poverty rates. Growth rates of output and input use have been significantly below those of the rest of India (Table 1.1), although there is evidence that growth in several eastern states is now accelerating (Ahluwalia, 1990). Eastem India has become the focus of numerous studies seeking both explanations and solutions for faster growth and development, in part because these states are the poorest in India and in part because of the significant untapped productive potential. Table 1.1 Growth Rates of Ou!ut, Inputs & Total Factor Productivity, 1956-1983 Output 3.1 2.1 2.6 2.2 Inputs 1.8 1.1 1.1 1.0 Total Factor Productivity (TFP) 1.4 0.8 1.5 1.2 a] Haiyana, Punjab, Rajasthan, Gujaat Mauarashr Madhya Prdesh, Uttar Pradesh, Andhra Radesh, Kamataka and Tamit Nadu. Source: Pray, 1991. The roots of slow growth in eastern India lie in the relative neglect of the factors which promoted higher growth in the Punjab and elsewhere: technologies, infrastructure, institutional arrangements, investments, and price policies. First, suitable technologies for the region's major crop, rice, have not been developed. Existing rice varieties are not suited to conditions of flooding and deep water which characterize nearly 50% of ricelands in the area. Nor have significant resources been devoted to resolving this problem: research expenditures in eastem India have been the lowest of all regions (Table 2.28). Thus, yield per hectare of rice has grown much more slowly than in the rest of India (0.8% vs. 1.7%). Second, eastern India is relatively deficient in basic infrastrcture - roads, irrigation, markets, communications and power infrastructure are the least developed in India. Eastern India's agiculture is predominately mainfed and as such, subject to greater variation and risk. Fewer investments have been made in water management (irigation, drainage and flood control) that would help temper these problems. Development of irrigation has been proportionately less: only 60% of eastern India's irrigation potential has been developed, compared to 82% of the Punjab's, and only 29% of the land is irigated, compared to 91% in the Punjab. Even where irrigation exists, there are fundamental problems in water management and control due to the inherently more complex problems of managing surface iurigation in high rainfall areas (Berkoff, 1990). The combination of lack of suitable HYVs and limited and poorly performing irrigation explain much of the stagnation of productivity growth in the region in the first post-Green Revolution decade. Lack of good infrastructure has hindered the development of extension, input supply and maketing networks.  6 BO|. (continued Third, levels of public investment per capita of the rural population in eastern India have been less than half those in wealthier states: public expenditure in 1980/81 prices in the 1980s averaged Rs. 50 to 75 per capita in eastern India, compared to Rs. 144 per capita in Punjab and Rs. 195 per capita in Maharashtra. Partly this is due to the fact that these are poorer states. But the eastern Indian states also tend to spend relatively less on irrigationi as a proportion of public expenditures (Ravishankar, 1990) Private investment has also lagged: the eastern states have throughout the 1980s accounted for the smallest percentage of term credit (a proxy for investment). Private investment is discouraged by lack of complementary public infrastructure (surface irrigation, roads, power), the low investment capacity of the region's small farmners, and higher risks due to lack of water control. Fourth, institutional factors figure predominantly in lagging investment rates. Eastern India has a much higher percentage of small farmns, tenant farmers and fragmented landholdings which discourage investment. Sociocultural factors, primarily the conception of agriculture as the residual sector, have worked against improving the basis for growth in the sector. Finally, institutions in the public sector in eastern India are among the weakest and most politicized in India. Eastern India needs more investment in research, human capital and infrastructure to increase agricultural productivity and reduce rural poverty. There is at least implicit concern that such investments would not have the high payoffs that they did elsewhere in India because of the daunting physical, institutional and socioeconomic conditions. However, analysis of productivity indicates that government investments in research, extension and infrastructure have been as productive or more productive in eastern India than in the rest of India despite the floods, droughts and institutional constraints. Slow growth rates in eastern India have been caused by slow growth of these inputs, rather than slow growth of productivity (although this varies by state) (Pray 1991). This is in contrast to the common perception of eastern India as a technically backward and inherently low productivity region. In particular, irrigation has a larger impact on total factor productivity in eastern India than in the rest of India; together with the higher untapped potential in the region, this argues for more investment in irrigation in eastern India particularly in groundwater development. The large impact of low growth in cultivated land (and the high amount of apparently fallow but productive land in Bihar) suggests that investments in land reform, despite their political difficulty, would have high payoffs. Agricultural productivity has begun to grow in West Bengal and Orissa, due in part to more Govemment investment in recent years, and supports the analysis that eastern India can be a region of growth in the 1990s. There is a strong case to be made to reorient investment resources from the rest of India to increase investment in eastern India, both to improve efficiency (due to higher levels of productivity) and to increase national agricultural growth as a whole. This will, however, require attention to problems of weak institutions and administration.  7 1.10 In the longer temi, agriculture's ability to continue absorbing significant amounts of labor will depend on a number of factors. First, growth rates of agricultural employment are slowin- compared to the last two decades, as the underlying factors responsible for employment growth tiave run their course. In particular, growth in irrigation, rapid yield increases and subdivision of land - all of which contributed to employment growth in the 1970s and early 1980s - have slowed, and are unlikely to be as rapid or as far-reaching in the futre. Second. employment elasicities are falling. The overall employment elasticity for Indian agricultue was estimated to be 0.71 in 1962-65, 0.59 in 1970-73, and 0.51 in 1980-83 (Gulati, 1991). These are averages, and crop and regional estimates of employment elasticities vary considerably (Table 8.5, Statistical Appendix). In general, cereal production (except rice) has lower employment elasticities, especially in the northwest, than crops such as oilseeds, sugarcane and cotton. Despite the decline in agricultual employment elasticities, these are still greater than those in industry. 1.11 Third, given wide variations in labor use by crop and region, employment growth will doubtless vary across the country (Table 1.2). In the rapidly growing states in the northwest, there has been a drop in employment creation and the increasing use of mechanization. Despite considerable debate on this issue in India, use of mechanization may have had a net employment creating effect, through expanding gross cropped area.2 Although labor absorption has dropped in these states, worker productivity has increased. Patterns of labor absorption in other areas of India show high agricultural growth, high labor absorption and increasing worker productivity in three states (Andhra Pradesh, Gujarat and Maharashtra), and low employment absorption and growth in per hectare employment in seven states (Karnataka, West Bengal, Rajasthan, Orissa, Bihar, Madhya Pradesh and Tamil Nadu). These trends are especially worrisome as the central and eastem states will account for a growing share of future labor force age increments (65% in 1980- 2000, compared to 51% in 1960-80). Although these areas tend to have higher employment input per hectare, stagnating productivity suggests low quality of employment (e.g. underemployment) and the inability of labor to transfer to other sectors. Table 1.2: Laebr Use fr Maior Cr-i (mandays/ha) Weighted M Averame Range Paddy 101.84 72.26 - 15936 Wheat 66.03 43.36 - 92.03 Maize 83.45 60.34 - 115.71 Jowar 61.63 50.09- 72.03 Bara 45.46 29S55 - 85.62 Gramn 41.71 24.9: - 64.87 Mustard 55.85 34.47 - 71.88 Groundnut 80.16 63.84 - 136.01 Soybeam 'O.6 46.10 - 51.92 Sunflower 41.73 38.62 - 45.16 Coton 99.76 66.89 - 129.17 Jute 179.25 102.00 - 197.77 Sugarcane 191.76 122.41 - 376.49 Sow= Gulad ad Shame. 1991. 21ntduction of mechanizadon has resulted in more land under producdon, in part because tactors can prepame land more quickly, enabling double cropping in irrigated areas (a wator, for example, can prepare land for wheat a few days after sugan:ane is harvested, compared to nearly two months for bullocks, thereby enabling another cop), and thus may have a net labo absorbing effect. In addidon, switches from bullocks to tmctors has freed up land used for fodder prduction, enabling cropping or switches to milk production (S. Bhalla, 1987).  8 BOX 1.3 W Employment opportunities in agriculture are especially important for women. Agricultural labor is the major source of paid employment for women, and there has been rapid growth in women's labor force participation. There is some debate as to whether rising female participation in paid outside labor is a result of increasing opportunities for women to transcend the invisible bamers that restrict women to "inside" (domestic and family) activities as opposed to "outside" (in the fields and marketplaces) work in India, or whether it is an indication of growing pauperization. Micro studies and national data suggest that both factors are at work, especially as the economic and cultural factors affecting female labor force participation differ radically among regions in India. In northem and eastern states, where women's role is strictly limited by cultural norms, female labor force participation is inversely related to farm size, land ownership, and income. The poorest groups - landless families, scheduled castes and scheduled tribes - have the highest percentage of females working (although scheduled tribes and castes are also less affected by the cultural restrictions on women accepting outside work). However in the southern and central states, cultural norms are less restrictive, there is a high participation of women in the paid labor force, and many workers are drawn from small farm cultivators. Studies of wage trends and labor use patterns suggest that there are increasing opportunities for female workers, especially in rice-growing areas, as women's real wages are rising (and the gap is narrowing with men's) and as HYV varieties of rice and cotton in particular have increased demand for female agricultural tasks (Bennett, 1990). Women's participation in the paid labor force has benefits for family health and nutrition as well, as nearly all women's wages are spent on the family (although for the poorest families, which make up the bulk of farm laborers, this does not substantially alter the fact that they are on the margin of survival). Overall, however, the labor picture for women in other sectors of the economy remains sobering - women account for only 6.6% of the jobs in manufacturing, and clearly have fewer opportunities to obtain the minimum education and technical training needed for such jobs. Choices in the pattern of agricultural development provide an excellent opportunity to improve women's lives and their gains from development. Efforts to improve the position of Indian women need to focus on women as critical economic actors rather than as target groups for poverty alleviation (World Bank,1989) - and agriculture is the sector where women are overwhelmingly concentrated. While there is no single strategy which will work in all regions or with all types of women, a consensus is emerging that it is only through better access to key inputs (such as education, land and credit) that real changes will occur in cultural ideology, in gender-based distribution patterns and in women's opportunities. Support for _cultu intensification and infastrt develoment can increase the demand for female labor, especially in certain crops (cotton, rice, vegetables, fruits) and animal husbandry. Efforts at redirecting txLsn to the needs of female cultivators (who provide much of the labor on small and marginal farms) are especially important in encouraging the adoption of new agricultural practices. Ample evidence from Africa, Asia and elsewhere in the developing world indicates that women can block agricultural modernization if they are only expected to provide additional labor without seeing clear advantages. Creative efforts can be made to increase women's access to crdit and other serices through demand groups, which assist women in dealing with public systems and the marketplace. Finally, there is little to rival female educafion as a means of weakening the inside/outside dichotomy and expanding women's stock of human capital - especially important to ensure that women can move out of agriculture into higher paying industrial and service jobs as structural transfonnation progresses. Incentives to educate girls include providing day care facilities (to relieve female children of childcare responsibilities), improvements in the quality and coverage of basic education, and strengthening nonformal education to reach older girls.  9 1.12 Future scope for employment increases in agriculture will depend on the choices made in investments, cropping patterns and intensification of irrigation, as well as on demand for agricultural products and the relative price of labor. Intensifying production through irrigation investments has a significant impact on labor absorption. However, indications are that employment creation stagnates once irrigadon potential is reached, so that a strategy based on increasing irrigation alone will be insufficient (Singh, 1988). High employment absorbing commodities, such as rice and cotton, have been relatively discriminated against in the structure of incentives (para 2.23); shifts into these commodities (and out of low-labor absorbing oilseeds) would increase employment. Crop diversification and animal husbandry, both of which are labor- intensive, also have the potential to increase labor absorpdon, and consumer demand for these products is inceasing rapidly (Table 1.9). Increased adoption of HYVs in rice growing areas will encourage greater labor absorption as well. Green Revolution technologies have generally maximized retums to land (through increasing yields and allowing double cropping), but havc not always maximized employment. This does not mean they are inappropriate technologies, for they have enabled higher returns to scarce land resources. The more practical and longer term solution to the problem of falling labor use in agriculture, in any event, is expanding employment opportunities in other sectors. Rural-Urban 0rowth Lnkaes 1.13 Development of the nual non-farm economy will be especially important to the rural poor, and consequently, merits close attention in India's rural development strategy. The rural nonfarm economy, dominated by commerce, service and small-scale manufacturing activities that cater mainly to agricultural and ural consumer demands, now accounts for 20%o of employment in rural areas and 30% of rural income. Nonfarm activities are especially important to the poor. Landless laborers and small farmers typically obtain half or more of their income from nonagricultural activities, and women are active especially in food processing and household manufacturing. Estimates of the multipliers for the impact of agricultural growth on the nonfarm sector are Rs. 0.64 based on econometic analysis (this is a national multiplier, state multipliers range as from Rs. 93 in Punjab and Haryana to Rs. 0.46 in low productivity states such as Madhya Pradesh and Bihar) and Rs. 1.35 based on input-output analysis (this multiplier is higher because it pertains to the national economy, not just the rural section).3 The income multiplier is Rs 1.56 fcr irrigated agriculture and Rs 1.23 for rainfed agriculture (Table 1.3). 1.14 Projections through the year 2040 of the volume of nonfarm income, employment, K and rural demand for manufactured goods that will result from agriculture indicate several important conclusions for growth and employment policy. First, because of strong agricultural growth linkages, rural nonfarm income and employment will both grow faster than their agricultural counterparts. A sustained agricultural growth rate of 2.4% (equal to past trend) will lead to 3% growth in nonfann income in rural areas and towns, and 2.8% growth in nonfarm employment. These growth rates increase to 5.8% and 4% respectively if agriculture grows at an accelerated rate of 4%. Second, continued trend growth in agricultural output by itself is unlikely to provide the growth in employment required to absorb the ruml labor force increases, although secondary rounds of growdt induced in the rural nonfarn economy could bridge this employment gap given even moderate rates of agricultural growth. Third, despite the strength of the rural-urban linkages, agricultural growth alone cannot provide the necessary market to sustain rapid growth in India's manufacturing sector (notwithstanding recent public rhetoric on emphasizing basic goods in industrial production and against "VCR - led growth"). An agricultural growth rate of 2.4% p.a. will generate 1.8 to 1.9% growth in national manufacturing output (based on rainfed or irrigated 3The first approach is an econometric analysis of cross-secdonal ste- and district-level &d.% Thc second is based on a semi-input - output model fitted to the nadonal input-output table for 1979/80. Both approches prov.-e esfimates of the agricultural income multiplier, defined as the increase in value added in the nonfarn sector atbutable to a one mpee incre in agricultual value added. See Hazel and Haggbladde (1990).  10 agricultue, respectively). Even a 6% agricultural growth rate will sustain only about 5.5% growth in manufacturing output. Table 1.3 National Aricutur Ouu adL1oe MuliliryarIj a me er RaLufe Agriculture Jirigated RaLnfed Tolal ONE RUPEE INCREASE IN CROP OUTPUT (Resulting lncrase in Sector Gross Output) Manufacting 1.05 (77) 0.94 (92) 0.98 Tertiary 1.14 (87) 1.11 (95) 1.12 Nontadable Agriculture 0.45 (94) 050 (90) 0.48 Toml. Nonfarm 2.19 (82) 2.05 (94) 2.11 Total 2.64 (84) 2.55 (93) 259 ONE RUPEE INCREASE IN CROP INCOME (Resulting Increase in Sector Income) Manufactuning 0.47 0.35 0.39 Terary 1.09 0.88 0.96 Nontradable Agriculture 0.51 0.47 0.48 Total Nonfarm 1.56 1.23 1.35 Total 2.07 1.70 1.83 Notes: al Weighted average using base-year gross output (income) weights of 0.4 (0.36) for irrigated crops and 0.6 (0.64) for rainfed crops. Figures in parentheses are the percentages of the increases in total sectoral outputs that are attributable to consumption linkages. Source: Hazell and Haggbladde, 1990. 1.15 Finally, although the size of the agricultural income multpliers depends primarily on the level of per capita agricultural income, public policy can influence their magnitude. The multipliers are positively rm.ated to the development of rural infrastructure (roads, electrification, banking services, etc). They are stronger under irrigated rather than rainfed agricultural growth, and larger for small to medium-sized fanns than for very large farms. Hence appropriate regional and farm targeting of agricultural technology, supported by adequate infrastructural investrnents, will be especially important for improving the nonfarm response to growing demand from agriculture. Moreover, well-developed rural towns foster stronger rural-urban growth linkages, and government policies towards small nonfarn businesses are important It is particularly important to avoid tax, regulatory or licensing policies that discriminate against small, labor- intensive businesses in favor of their larger, more capital-intensive cousins. Envdirnmealssues 1.16 Any attempt to increase agricultural productivity cannot overlook the serious environmental problems confronting India. India is already at the margin of its cultivable land (totalling some 140 million ha) and environmental pressures (erosion, overgrazing, expansion into wastelands and salinity problems) are increasing on both marginal and cropped lands (Figure 1.2). Official statisdcs on land degradation and deforestation are weak and often conflicting, so the extent of the problem is not known precisely. Between a third to a half of India's total land area of 304 milion ha is estimated to suffer from problems of degradztion, and the speed with which land degradaton is occurring is alanning: up to 2.5 million ha (or 0.76% of India's land area) turns  11 into wasteland every year by some estimates. Another one million odd ha degrade every year because of mining, waterlogging, soil salinization, creation of artificial reservoirs, expansion of settloments and industries, cultivation of marginal land and overgrazing. According to satellite data, India lost forests at the annual rate of 130,000 ha in the period 1972-75 and 1980-82. These data were acknowledged to have problems, and a more careful assessment done in 1989 estimated actual forest cover at 64.01 million ha (19.5% of land area) against an officially recorded area of 75.1 million ha (22.8% of land area) (GOI, 1989). Most observors believe deforestation rates are understated, and closer to 1.3 to 1.5 million ha per year. (World Resources Institute, 1990). These figures include both closed and open forest area. Nearly three-quarters of forest cover was concentrated in two major tracts - the Himalayan range and the states of the northeast, and the central and eastern Indian states of Madhya Pradesh, Maharashtra, Orissa, and Andhra Pradesh. Most other areas now have a minuscule forest area and almost all ecologically sensitive areas like hill, mountain and plateau regions are devoid of almost any tree cover. Figure 1.2: Land Use in India, 1950/51 - 1986/87 M.ha 2 Area for which no return exists 250 - _l Net area sown 200 - ~ ~~~~~~~~~~~~~~~*Fallow lands O - - ;. lOther uncultivated land 1 5 - -.:.----: 1 .50 ; ;..-.'..'. Area not available for cultivation 1 .0.-.. ..... ........ ...... : ::::.:::: . ... U ; Area under forest 50- 50 - - . - . .~~~~ -' .' - 1950/51 1960/61 1970/71 1980/81 1986/87 1.17 India's water resources can be characterized by first, their enornous size and variety (4,000 billion cubic meters annually in torrential Himalayan rivers, in deep Gangetic alluviums, in seasonal deccan streams and in coastal swamps) and second by their fundamental irregularity and imbalance. In most of India, over 75% of precipitation is within a three month monsoon period and that with a mean variability of about 30%o. This requires major water resource planning, development and management efforts to ensure adequate supplies for agriculture, industry, energy, human and animal consumption. Theoretically, India has substantial water resources to draw upon (it has been estimated that the country today uses only about a tenth of the precipitation that it receives annually). However, poor management of the resource is causing problems. At present, an estimated 80% or more of India's villages are without a safe and clean water supply, and only about 50% of the urban population is said to have access to safe and clean water, causing significant human health problems. Pollution of water resources is growing: most of India's rivers are polluted by sewerage, industrial and agricultural wastes.  12 1.18 The major environmental issues now facing Indian agriculture are (a) the impact of irigation and groundwater development, including the impact from resettlement of displaced populations; (b) land management issues, including forestry; and (c) incentives to overuse fertilizers and pesticides, and overexploit groundwater. 1.19 Env.nmental Imoact o fio. Irrigation has probably had, on balance, a positive environmental impact, largely because it has reduced pressures on land where rainfed agricultural practices predominate by permitting increased production of biomass for food, fodder and fuel. The problems of erosion from cultivation of marginal soils, overgrazing and deforestation would have been far more acute without rapid development of irrigation. 1.20 Yet there are negative environmental repercussions of irrigation development. Loss of land for reservoirs, canals, and other infrastructure is inevitable, but relatively small. River diversion schemes usually take 2-5% of the potential command. Storage schemes take another 3- 8%, depending in the dam site. Larger reservoirs tend to involve relatively smaller land loss. Sardar Sarovar, for example, results in a 2% land loss. Some of the land lost may be designated forest area.4 In view of the rapid rate of deforestation in India, this is a serious matter. Although losses due to irrigation probably account for less than 3% of this and may be compensated by production of other biomass, the extent of land loss and the possible effects on biodiversity should be systematically taken into account in the evaluation of iigation investment. 1.21 The loss of land brings with it displacement of families: on average 6 families per 100 ha provided with irrigation. Displaced families, or "oustees", must be provided with the means of at least sustaining and preferably improving their standards of living, ideally through sharing in project benefits. Indian policies for dealing with oustees are not yet satisfactory. State policies are deficient, funds are inadequate and implementation capacity is low. Govemment land on which oustees were to be resettled has either been unavailable or of inadequate quality, and cash compensation has been insufficient to purchase equivalent land. 1.22 Some progress is, however, being made. In 1989 a draft National Policy on Development Resettlement was prepared jointly by the Govemment's Commission on Scheduled Castes and Tribes and nongovernmental organizations. The states of Maharashtra, Madhya Pradesh and Kamataka have passed legisladon defining resettlement policies; Gujarat, Bihar and Orissa have issued Government orders which effectively provide a legal and policy framework. However, in the rest of the country the policy, legal and institutional framework remains inadequate. The revision and issuance of a national resetdement policy paper is an essential first step. This should be followed by enactment of similar legislation by the states. At the project level, land acquisition, resettlement and rehabilitation plans have to be worked out in detail with local authorities, community leaders and nongovemmental organizations. Displaced people should be incorporated into the irrigation commands of projects (thereby sharing in benefits) through land purchase - this approach has been successful in Gujarat. Agreed actions, organizational procedures and expenditures should be included in the project 1.23 While a number of other environmental problems are associated with irrigation, waterlogging and salinization, siltation of reservoirs, reduced river flows in dry season and dangers of increases in waterborne diseases, these are not inevitable. Waterlogging and salinization, which may affect as much as 13-20 million ha of land, is not entirely or even mainly the result of irrigation. Nevertheless, it can be exacerbated by poor maintenance, inadequate investment in drainage and weaknesses in water management systems. Low river flow can reflect inadequate river basin water resource use planning. These problems indicate the need for a more comprehensive and better planned irrigation strategy giving higher priority to imnproved efficiency of existing systems, river basin plannig, and a shifting of expenditure priorities away from new systems to reconstruction, maintenance, drainage and modemization. 4Meas designated as fawest ieives in India are not infiequently severely degrded.  13 1.24 Resoutce Management. India's rapid depletion of forests and vegetative cover is of particulr concrn in raind areas, which stil account for nearly 700o of total croplands. Depletion IS due to rapid population growth and density, lack of alternative sources of fuel for domestic cookng needs, extensive cate gzing in forests, and commercial demand. The impact of human activity on the forests has been characterized as an ecological disaster, resulting not only in loss of wildlife habitat, but also in soil erosion, siltation, floods and droughts. Catle production is a particular pmblem. The low productivity of the current cattle population is reflected in the dispani't between available feed fodder and the level needed for maintenance and production. In 1985/86, for example, a total of some 460 million mt was produced against a national requirement of about 750 million mt, leaving a deficit of some 290 million mt. The problem is compounded by the role of cattle in Indian society, notably the status that cattle ownership conveys, despite their lack of productivity and their preservation, at almost any cost, until natal death. 1.25 Incentives. The current structure of incentive policies has environmental effects as well. Low prices for water (through both surface irrigation and subsidized electricity for groundwater pumping) results in mismanagement and overuse of scarce water resources. Water- intensive crops such as rice and sugarcane are grown in areas with limited water resources. Overexploitation of groundwater, due to very low power rates, has reduced water tables and caused concems over depletion in some areas. Low power rates discourage energy conservation as well. Although fertilizer use remains at relatively low levels in India, field reports from the Punjab and Haryana region (where use is highest) suggest that indiscriminate fertilizer use is adversely affecting soil conditions as well as posing ecological hazards such as groundwater pollution (Chopra, 1990). These problems could all be checked by pricing inputs correctly. At the same time, susiainability of production systems could be ensured. 1.26 Actions Neede Addressing environmental issues requires tackling a range of problems (economic, technical and sociocultunil) that contribute to incentives for misuse of physical resources. In some cases, needed adjus-vrents are evident; in others, more work needs to be done on identifying priorities and remedihA a tions. There are already technical solutions to control arable land degradation, dtrough intvILu.:Con of appropriate soil and moisture conservadon measures in conjunction with improved fanning systems. Better management of irrigation schemes and removal of disincentives to overuse water are also necessary. To make best use of water resources, river basin planning, incorporating multiple demands for water, is essential. Adjustments in subsidies for inputs, particularly 1 nidlizer, will reduce incentives for overuse. To control degradation on nonarable lands, efforts si o- ld be made to maintain existing vegetative cover, encourage revegetation of degraded areas, wvitl an emphasis on meeting the needs of local peoples for fuel, fodder, timber and non-forest prodb is; improve better conrrol and managgement of livestock and animal grazing. While technolo-,ic-d solutions for non-arable lands will be important, these are only sustainable if the long-te; :n socioeconomic problems of these areas - which are common property resources - are also tackled. 1.27 While these are broad indications of what can, and in many cases already is being done, it is also evident that the more guidance on practical and long-tenn solutions is needed. Although the broad ramifications of environmental issues for agriculture is klown, serious gaps remain in data, analysis of secondary impacts, and setting priorities. This will partially be addressed in the forestry subsector, where Govemment is preparing to carry out a national forestry action plan exercise, to set priorities for aresting deforestation and encouraging afforestation. This exercise is being complemented by World Bank work on forestry issues, and preparation of assistance to the forestry sector at the state level. A more detailed study on sustainability issues in agriculture with analysis of issues, priorities and actions is needed. India is also poised to develop an environmental action plan to translate these recommendati^ns into practice.  14 B. Prformance and Sources of Growth 1.28 India's agricultural performnance isliproved after Independence, but growth has remained moderate despite impressive successcs in raising foodgrain output. The virtual stagnation in agricultural production since the beginning of the century (output grew at 0.4% p.a. between 1891 and 1946) gave way to steady growth in the post-independence era. Most notable has been the sustained growth in foodgrain production brought about by the introduction of high- yielding varieties of rice and wheat in 1965 - the renowned Green Revolution. The yield and output growth attained in the last two decades for these crops have been impressive, and have allowed India to become food self-sufficient. The Green Revolution marked a turning point in the modernization of Indian agriculture. However, by many standards, sector performance has fallen short: overall growth rates failed to accelerate after the Green Revolution, growth remains highly concentrated in certain crops and regions, rairded agriculture is lagging, and yields and input levels remain low compared to other Asian countries. BOX 1.4 Wat is the Growth Rate of Agrcultum in India? Discussion of agricultural growth rates in India is often confused by different definitions of what should be measured. Growth of GDP in agriculture is the normal indicator from an economic standpoint, as it measures value added; it is the indicator used here. However, in India, growth of output is commonly used (in part because disaggregated statistics on value added at the crop and regional level are lacking). Using this measure, agriculture grew at a trend rate of 2.5% in the period 1950/51 to 1988/89. Another commonly used indicator is growth of foodgrain production in physical terms, although foodgrains account for less than 65% of output; by this indicator, the trend rate of growth was 2.6% p.a. There is also some discussion of what constitutes a "dynamic" versus a "moderate" growth rate for agriculture. Growth rates over 3% for agriculture are usually considered dynamic, although for foodgrains, growth rates 0.5% ahead of population growth are considered rapid (Mellor, 1990). By these standards, India's overall growth has been moderate, but foodgrain production has been dynamic. Aggregate Producdon Growth 1.29 Overall growth rates in Indian agriculture, even after the Green Revolution, have been moderate. In the period 1967/68 to 1988/89, GDP originating in agriculture (proper) increased at a trend rate of 2.6% p.a (Figure 1.3). In comparison, GDP originating in the industrial sector grew 5.2% p.a. and in services 4.6% p.a. in the same period. Agriculture as a share of GDP declined from 56% in 1950/51 to 33% in 1988/89, consistent with growing industrialization (Figure 1.4). Agriculture's share of state GDP, although varying from only 17% to over 55%, also shows a declining trend (Table 1.5). 1.30 India's agriculturl growth has been modest from several perspectives. First, many other developing nations in Asia have performed better. Comparing data from the major countries in the region for the period 1968-88 (the Green Revolution period) (Table 1.4), most of the other countries in the sample have growth rates well ahead of 3%, with only Bangladesh and Nepal growing more slowly than India. Comparing regional growth rates to those of neighboring countries (which might provide a better international comparator because of geographic and in some instances cultural, proximity) between 1976/77 and 1986/87 shows that none of the regions in India except the east did as well. The 4.3% p.a. trend growth in north India is less than the 4.6% p.a. achieved in Pakistan (as a whole); growth in the south (1.1%) is much less than in Sri Lanka (2.9%); the east (2.7%) did somewhat better than Bangladesh (2.0%); and the central states, which have no intemational borders, lagged behind all neighboring countries, with only 1.3% growth.  15 Figure 1.3: GDP frm Agiiculture and Rainfall RA. billion (1880/81 pricea) Rain indlootor 700 200 600 500 - -- - - ---- 5 400 - 2.0% per annum 1 00 300 200 o0 100 0 0 68 69 70 71 72 73 74 7676 77 78 79 80 81 82 83 84 86 86 87 88 89 Indian Fiscal Years GDP agricultuare = Rain indicator Mean Rain -+ GDP trend Figure 1.4: Sectoral Shares in GDP - All India 100 .,......,..,- ..., ..- :,-,.:-- -'-' 80 60 40- __ 40- 20 1951 1955 1959 1963 1967 1971 1975 1979 1983 1987 Indian Fiscal Years | Agnculture & allied 0l dstry U Services Note: Data are from the GDP series (at factor cost) at current prices. Source: Ahluwalia (1990).  16 Table 1.4: t R f (% per annum) Staiscally Significant Break between 1978-88 and 1968-88 1268-22 192us 198- 77 atA5% Lvl Bangladesh 2.0 0.5 2.0 Yes at 6% Burma 4.8 2.8 5.6 Yes China 4.3 3.1 6.2 Yes Nepal 1.9 1.4 3.8 Yes Pakistan 3.6 2.4 4.6 Yes SriLanka 3.0 1.6 2.9 Yes India 2.3 2.3 2.5 No Indonesia 3.9 4.0 3.5 No Malaysia 4.0 5.7 3.5 Yes Philppines 3.7 4.4 2.3 Yes Mhailand 4.0 4.1 3.4 Yes at 6% Notes: For Burma data until 1986: for Malaysia data from 1970. Soure: Ahuwalia (1990). 1.31 Second - and this is important for understanding the dynamics of growth - in six of the eleven comparators, there is evidence of a significant upward break. with trend in the decade 1978-88 compared with the previous decade. In India, there is no significant upward shift in the growth trend in agriculture in recent years, unlike industry. Overall, growth in the post-Green Revolution period is not significantly different than in the pre-Green Revolution period. Third, the rapid increase in population in India (about 2.1% p.a. over the last four decades) has meant that the modest aggregate growth rates of production have translated into even more modest increases in per capita agricultural production. Given the virtually closed nature of the Indian agricultural economy, this has led to very small increases in per capita availability of food as compared to other countries in the region (Pakistan, China and Sri Lanka, for example). It has also resulted in per capita rural income growth rates of less than half those in the urban sector. 1.32 Growth has been concentrated in a small number of states and regions. India is a vast and hete-ogenous country, and thus aggregate figures obscure highly varying stories at the state and regional levels.5 A handful of dynamic states - mainly in the north - have had robust growth rates (3.5 - 4% p.a), compared with rates of 2% p.a. or less in nearly half the states (Table 1.5). The brighter spots in the picture include evidence of accelerating growth in Uttar Pradesh and Assam from the mid-1970s, and higher overall growth rates in the east in the last decade (although the initial base was much smaller than in other states) as well as in the central states - welcome tends because these states have been the laggards and contain the highest proportion of poor people. The bleaker aspects are evidence of significant declines in growth in Kerala and Tamil Nadu, and stalled growth in the south and west 5Fiftee. main states account for about 85% of all-India domestic product originawg in agriculture. These sttes au Haryana and Punjab in the Northern Region; Uttr Pradesh; Assam, Bihar, Orissa and West Bengal in the East; Madhya Pradesh and Rajasthan in the Center, Gujamat and Maharshra in the West; and Andhra Pradesh, Karuazka, Kemala and Tamil Nadu in the South. Uttar Pradesh is considered separtely (it is normally classified with the North) becs of its dominant size, and the wide variation of production patems from west to east within the Sa.  17 Table 1.5: _nAgigM u.nt. aid the Ralrnv Imete fAvadz iti b Stat (Oa Shaes n Shareof Gwowthof AB-India Domatc Agrkui in Stat Dmatec Pduct Origing Ste Dndic Produ Originag i Apiam Prd=t in Agluixa (average _954d (averag 1984154617 (l960161-1?g817)_ Ngh A8. 43. 4.0 Haryana 3.3 433 3.7 Punjab 5.1 43.8 42 Uw Emdo 12.8 38.0 23 FM 23.0 41.9 2.2 Assasn 2.8 41.9 1.3 Bihar 8.0 453 1.4 Orissa 4.2 55.2 4.2 WestBengal 7.9 34.8 2.5 113 42.9 2.5 Madhya PRndh 6.2 40.1 1.7 Rajasdtm 5.2 46.7 3.6 9ba 11.5 22.8 2.0 Gujarat 4.2 25.0 2.2 Maharaahtra 73 21.7 2.0 ZIwh 17.6 30.6 1.5 Andhra Pradesh 6.6 36.9 1.8 Kamatka 5.0 36.9 2.7 Kerala 3.1 35.4 0.9 Tamil Naed 3.0 16.9 .0.03 lou 84.6 35.0 AD h& 100.0 34.5 2.3 Sa4m GOL CO,. Badrz of Stae D Ne&wm vas Isma AhIuValb(1990) 1.33 These variable growth rates among states reflect -)me extent resource endowment and infiasu ue development, but they also indicate a growinm regional speadion along lines of compaive advantage. Indeed, this specilizion has already begun t occur in the postGreen Revolution era, with the bredbasket of India shifng fom the eastern Gangedfc Plain to the nthwest. Other shifts are be g. n ses such as Kead and Tamil Nadu, agriculte is increasingly constrained by waterresources. I other states with low growth rates, es y in the west, comparative advantage would appe to lie outside agriculture, as the sector's m nce in their economies (20-25%) is much lower than the all-India average (34.5%). In the north, there are indicadons dtat productivity may be plateauing in traditional foodgrains. Future growth here will depend on increased diversification and specialization in high-value crops (horticulture, livestock products) with veical integradon and processing links. Future growth prospects center on Uttar Pradesh and the eastern and cental states, which together account for nearly half of agricultual GDP. These stmes came late to the Green Revolution and many constraints remain (Box 1.2). But these regions are endowed with a number of factors - relatively high irrigation  18 potential, good rainfall and a large rural population - which suggest they can reclaim the prominent position in agriculture which they once held. The potential for faster growth in these regions is more than a promise - it is alrady underway in several crops, productivity is improving and growth is accelerating (para. 1.36). EoWrh *ni*Cxgp 1.34 Although the Green Revolution led to large output increases for some crops, it has not led to any acceleradon in the growth of production at the all-India level. Indeed, there has been slower trend growth in most of the crop groups after the Green Revolution as compared to the pre- Green Revolution period (Table 1.6). The only exceptions are the relatively minor groups of condiments and spices, and plantation crops (which account for less than 4% of production). In fact, there has been a (statistically) significant downward trend for coarse cereals, sugarcane, and fruits and vegetables in this period. Nor has there been much significant change in trend between the first and second decades of the Green Revolution, except for oilseeds, condiments/spices, and plantation crops. Table 1.6: Trd (art Rates of Pmduces of Prncin CMp Ugm Al t AM-In l (% per nnn) Significant Signifin dtn Post-Gre PostGret (ahge Entie Pie-Green Post-Gen between Revoln. Revoln. betwee Weig tn pliod Revohnion Revoluoion pee and psaecsde I Dlede I Decade I Index of 1949/50- 1949150- 1967.68- Green Revhn. (1967/68- (1978-79- & Dec. 11 Qn (ip A&n~d& 12I8912ili64 I988M2 at I 119771781 12AM2X at 5Lh:vIe All Qm 100.00 2.5 3.1 2.5 No 2.3 2.8 No Eg2ftmim 62.92 2.4 2.8 2.5 No 2.2 2.8 No AllCeeals 54.98 2.8 3.2 2.8 No 2.6 2.9 No CoarseCaeals 10.79 1.2 2.2 0.5 0 Yes 0.7 * -0.3 0 No All Pulses 7.94 0.4 1.4 0.7 No 0.3 e 2.1 0 No Non-Foodgmlns 37.08 2.6 3.7 2.6 Yes 2.6 3 No All Oilseeds 12.64 2.1 3.2 2.2 No 1.3