Pakistan : public expenditure in agriculture

Report No. IDP-153 SOUTH ASIA REGION INTERNAL DISCUSSION PAPER PAKISTAN: PUBLIC EXPENDITURE IN AGRICULTURE Yusuf Choudhry & Rashid Faruqee The World Bank June 1995 The views presented here are those of the authors, and should not be interpreted as reflecting those of the World Bank. SOUTH ASIA REGIONAL SERIES Title Author Date Originator IDP1 11 How Composition of Public Expenditure Affects Competitiveness: The Case of Bangladesh K. Matin March 1992 P. Mitra (80419) IDP1 21 Labor Retrenchment and Redundancy Compensation in State Owned Enterprises: The Case of Sri Lanka A. Fiszbein December 1992 G. Nankani (84641) IDP126 Some Guidelines for the Appraisal of Large Projects W. Jack February 1993 A. Estache (81442) IDP109 Reforming Higher Secondary Education in in South Asia: The Case of Nepal H. Abadzi May 1993 H. Abadzi (80375) IDP1 27 Some Lessons for South Asia from Developing Country Experience with Foreign Direct Investment M. Fry June 1993 A. Estache (81442) IDP1 29 Quasi-Fiscal Deficits: Latin American Lessons for South Asia C. A. Rodriguez August 1993 A. Estache 181442) IDP1 31 The Impact of Rural Infrastructure on Rural Poverty: Lessons for South Asia E. Goldstein June 1993 G. Nankani (846411 IDP1 34 Infrastructure and Industrial Policy in South Asia: Achieving the Transition to a New Regulatory Environment P. Seabright December 1993 A. Estache (81442) IDP1 35 Taxation of Foreign Investment in South Asia J. Mintz T. Tsiopoulos December 1993 A. Estache (81442) IDP142 Regional Trading Arrangements and Beyond Exploring Some Options for South Asia Theory, Empirics and Policy T. N. Srinivasan July 1994 G. Nankani (84641) IDP146 Participation in the South Asia Region's Project Portfolio: Towards Defining Conditions for Success B. Parker August 1994 G. Nankani (84641) IDP1 53 Pakistan: Public Expenditure in Agriculture Yusuf Choudhry Rashid Faruqee June 1995 J. Wall (85045) PAKISTAN: PUBLIC EXPENDITURE IN AGRICULTURE Yusuf Choudhry & Rashid Faruqee June 1995 TABLE OF CONTENTS ABSTRACT SUM MARY ....................................................................................................................................... i 1. INTRODUCTION ................................................................................................................ 1 Policy Objectives and Role of Government in Agriculture ....................................................... 1 Performance of Agriculture ..................................................................................................... 2 2. NATIONAL AGRICULTURE EXPENDITURE: TREND AND STRUCTURE .............. 3 3. RELATIONSHIP OF AGRICULTURAL EXPENDITURE TO OVERALL PUBLIC EXPENDITURE ................................................................................................. 4 4. COMPOSITION OF PUBLIC EXPENDITURE IN AGRICULTURE .............................. 6 Current Expenditures ............................................................................................................... 6 Development Expenditure ........................................................................................................ 7 Current vs. Development Expenditure ...................................................................................... 9 Federal vs. Provincial Expenditure ........................................................................................... 9 Subsidies ................................................................................................................................. 9 The Impact of Subsidy on Sectoral Objectives and Performance ............................................... 11 5. SUBSECTORAL EXPENDITURE ..................................................................................... 15 Crop Agriculture .................................................................................................................... 16 Fisheries, Livestock and Forestry ............................................................................................ 20 Irrigation ................................................................................................................................ 21 O&M of Irrigation and Drainage Systems ............................................................................... 21 6. EXPENDITURE ON PUBLIC INSTITUTIONS ............................................................... 23 7. THE FINAL ASSESSMENT ............................................................................................... 27 8. CONCLUSION .................................................................................................................... 28 LIST OF TABLES Table 1 Growth in Agriculture in Pakistan .............................................................................. 3 Table 2 Comparative Agriculture Expenditure Ratios in Three Developing Countries .............. 4 Table 3 Development and Current Agricultural Expenditures Excluding Subsidies .................. 8 Table 4 Provincial/Federal Expenditure and Subsidy .............................................................. 10 Table 5 Total Subsidy as Part of Total Agriculture Expenditure ............................................. 10 Table 6 Food Subsidy and Total Current Expenditure (Including Food Subsidy) ..................... 12 Table 7 Transfers from-and-to Agriculture Due to Current & Development Subsidies ............. 16 Table 8 Crop Agriculture: Establishment & Non-Establishment Expenditure ......................... 17 Table 9 Estimates of Marginal Products and Marginal Internal Rates of Return for Agricultural Research in Pakistan .................................................................. 18 Table 10 Establishment to Total Expenditure Ratio for Agriculture Research ......................... 19 Table 11 IrrigationO& Expenditure ................................................................................... 22 Table 12 Difference Between O&M Allocation and Need for Canals under ISRP Targets (in %/) ............................................................................................ 22 Table 13 Consolidated Profit & Loss of Major Public Corporations in the Agriculture Sector ............................................................................................... 24 Table 14 Change in Selected Indicators of Expenditure Efficiency .......................................... 28 LIST OF CHARTS Chart 1 Pattern of Agriculture Expenditure in Pakistan ........................................................ 3 Chart 2 Pakistan: Agriculture's share of Government Investment to Share of Total O utput .................................................................................................. 5 Chart 3 India: Agriculture's Share of Government Investment to Share of Total Output .................................................................................................. 6 Chart 4 Establishment cost Ratios for Agriculture Public Expenditure in Pakistan ............... 7 Chart 5 Current & Development Expenditures Excluding Subsidies .................................... 8 Chart 6 Fertilizer Subsidy and Consumption in Pakistan ..................................................... 13 Chart 7 Fertilizer Subsidy Trend ........................................................................................ 14 Chart 8 Crop Agriculture - Current and Development Expenditures .................................... 18 Chart 9 Crop Agriculture - The Trend in Establishment Expenditure ................................... 20 BOX Box 1 Pakistan: Actual vs. Planned Expenditure in Agriculture ....................................... 4 ABSTRACT Public expenditure policies in Pakistan have neither served the need of the agriculture sector nor have they been used efficiently. A considerable portion of the budget outlay has gone for consumer food subsidy which has affected agricultural producers and the private markets adversely. Subsidized wheat prices, procurement by the government, etc. have depressed prices and deterred private investment in marketing and distribution. Likewise, input subsidies like fertilizer subsidy have also worked against private market growth without yielding a lot of benefit for smallholders. Although the National Agricultural Policy in Pakistan has projected the need for government actions in many relevant areas, such as agriculture research, infrastructure, development of high value products, and checking natural resource degradation, the direction of public expenditure does not seem to support these policy goals. Even where expenditures have been made, ostensibly in direct support of the agricultural sector, the efficiency of such outlays has been extremely poor. Support of public enterprises, such as the fertilizer, seed and storage corporations, have been a high cost, low yield venture for the government. Most of the public corporations have run at a loss and have hindered free markets. There is, this, considerable scope for improving both the composition and efficiency of public expenditure in Pakistan's agriculture. SUMMARY The public sector's role in agriculture has been overextended and hence a substantial part of expenditures is in inappropriate areas. The state's role of supplying public goods in the economy has been interpreted too loosely in Pakistan. Instead of concentrating on creation of true public goods through productivity and growth enhancing activities, such as support of agricultural infrastructure, research, and proper regulatory environment for boosting private sector confidence in agricultural investment, the government has spent an inordinate amount of resources for affecting income transfer which provided little relief to the sector. Support of public enterprises, which do not produce public goods, have also absorbed large quantities of scarce public resources, besides creating harmful externalities such as holding back the growth of competitive and efficient private markets which could have been more beneficial to the poorer farmers. The composition of public expenditure program does not conform to stated national agricultural policy (NAP) objectives. Development expenditure on agriculture has been falling in relationship to overall expenditure (down from 62 percent in 1984 to 40 percent in 1994). This negates the investment and growth objectives of the NAP. A large chunk (21 percent) of development expenditure is going for input subsidies whose effectiveness and benefits are questionable. This does not auger well for national development activities like agricultural research, rural infrastructure development, etc. which are important elements of the NAP. The division of government expenditure between subsectors has not been entirely consistent with the NAP objectives either. Irrigation absorbed the largest share (51 percent), followed by crop agriculture and food (40 percent). The remaining subsectors, including forestry, fisheries and livestock received but a fraction of the total budget. This is not totally consistent with the objectives of the NAP, which allocates high priority to forestry, fisheries and livestock. Even though irrigation received such a high proportion of public expenditure in agriculture, the allocation towards O&M have been inadequate causing widespread deterioration of irrigation facilities. This also seems to be inconsistent with the NAP which recognizes the deficiency in O&M expenditure and calls for steps to redress it. In terms of both size and trend, agriculture in Pakistan is squeezed as a public expenditure priority. Between 1982 and 1994, total provincial and federal spending on agriculture decreased in real terms by an average of 0.57 percent per year. Over 55 percent of total government spending on agriculture was current expenditure. While this has risen significantly in real terms between 1982 and 1994 (3.26 percent a year), development expenditure has declined by 3.89 percent per year. In fact, current expenditure has risen from about half of development expenditure in 1982, to about one-and- one-half in 1994. This growth cannot be regarded as favoring agriculture, because food subsidies (wheat, sugar, and edible oil) have accounted for almost half of it. Some of these subsidies (e.g. wheat) are actually meant to lower prices for consumers, and that affect procedures adversely. Effective development spending was constrained by the inclusion of fertilizer subsidies. Between 1982 and 1994, constant mpee development expenditure without subsidy actually declined by 3.74 percent a year . Although the share of agricultural expenditure in the budget by itself is not a good indicator of either the government's commitment to agricultural development, or the quality of the expenditure program, a comparison with other countries shows that agriculture in Pakistan has been under- represented in the budget. Agricultural public expenditure, as a percentage of total public expenditure, has varied between 3 percent and 8.5 percent between 1982-94, with an average of 5.6 percent. Compare this to the average figure for 1984-92 for other countries such as India (7.2 percent), Indonesia (7.2 percent) and Malaysia (6.8 percent). The largest fiscal outlay made by the government in agriculture and food has been on subsides. The food subsidy (mostly on wheat) has actually helped consumers at the cost of producers. The other agricultural subsidies did not fulfill their stated purpose. Subsidies fall under two categories--current subsidies which comprises mostly of food subsidy, and development subsidy which includes fertilizer, seed and tubewell subsidies. The provincial government's have borne a large portion of the current subsidy, but development subsidies have almost exclusively been federal. The ratio of current subsidy to current expenditure for the provincial governments, over the past 12 years, has varied from 19 percent to 44 percent. The same ratio for the federal government has varied from 36 percent to 97 percent. Similarly, development subsidies borne almost exclusively by the federal government have also been high, from 19 to 45 percent of development expenditure. Aggregate subsidies have fallen to some extent in recent years, both absolutely and relative to total expenditure. Current subsidies, that have gone to support food procurement and price stabilization have been the cornerstones of government's food policy. The government has tried to insulate consumers from major swings in wheat price through massive procurement and subsidized imports (in Pakistan, the percentage of wheat production handled by public marketing agency consistently exceeded the twenty percent level -- more than those in India and Bangladesh). This policy of market intervention has come with a price. The level of wheat production has remained low because both farmers and traders were discouraged by low prices. The government buying agency has been purchasing post harvest stocks and releasing them later at predetermined fixed margins over procurement prices, uniformly all over the country. Such practice has inhibited the growth of private storage, transportation, and active private trade in food grain. The government has also controlled international trade of wheat by prohibiting export and importing irregularly to replenish domestic stock. Thus, domestic prices of wheat have been lower than import parity prices. Meeting this price differential, along with the cost of storage and maintaining buffer stocks, has been a heavy fiscal drain for the government The cost of maintaining food subsidies has been 29 to 66 percent of current expenditure during the period 1984-85 to 1993-94. 11 Unfortunately, this system i. not achieving its stated goal because the price of flour is already market-determined. Consequently, millers are absorbing most of the subsidy on wheat as rent, and the milling industry has excess capacity. In 1993-94, the retail price of flour was 95 percent of the import priority price of flour, so the effect of the subsidy on retail price is negligible. Subsidies on inputs as fertilizer and seeds distort input markets for these products. Moreover, government activities in procurement, distribution and marketing of inputs are almost always inefficient and unresponsive to the needs of buyers. For example, the cost of fertilizer distributed by government agencies in Pakistan has been much higher than the private sector, and this inefficidncy loss is borne by the public exchequer. The farmers have been deprived of the benefits of competitively driven prices and an active commercial market. Cheap inputs (in the short run) also encourage waste through overuse, particularly by poor farmers not well versed in agronomy, and add to environmental degradation. Besides, fertilizer and seed subsidies are essentially regressive since they benefit mostly the larger farmers that use large amount of these subsidized inputs. Also, the administration of the subsidy program encourages rent- seeking. Other subsidies have outlived their usefulness. A subsidy on seedlings was designed to spur the growth of private forestry. Rapid growth in forestry has been achieved, but the principal beneficiary of the subsidy have been large farmers, some of whom have evicted tenants in order to develop block plantations. There are, also, a host of indirect subsidies that have serious impact on fiscal deficits. Qureshi (1993a) has shown that most of these indirect subsidies (electricity, water and credit) have added anywhere from 3.64 percent to 7.67 percent to the overall fiscal deficit. Furthermore, the average consumers, who are the intended beneficiaries of government intervention, have not received significant benefits because of leakage and inefficiency in state organizations. Subsidies from agricultural loans (subsidized interest and often not recovered) and irrigation water (charges collected are well-below cost) are also substantial. Electricity has a sizable subsidy too. In Qureshi's (ibid.) estimate, the subsidies on credit, water and electricity amounted to almost Rs. 5 billion in 1990-91, which is 2.6 percent of total expenditure of the government (revenue plus capital). Public expenditure in agriculture has been inadequate in certain critical areas such as alleviation of natural resource degradation and operations and maintenance of irrigation. Natural resource degradation, arising from waterlogging and salinity, has not been sufficiently addressed. To the extent that these problems are due to inadequate drainage (a public good), the government could play an increasing role. At the same time, the government has provided funds for the development and maintenance of public tubewells, even though most of the benefits of tubewells are privately appropriated (although there may be some public benefit of tubewells resulting from lowering of water tables). Particularly serious is the neglect of operations and maintenance expenditure of the irrigation system, which has consequently deteriorated. The shortfall varies by province (as high as 54 percent in Punjab) and essentially stems from low water rates and inadequate assessment and collection of iii charges. In addition, the revenue collected does not go directly to the Provincial Irrigation Departments, which creates poor collection incentives. Rural infrastructure deficiencies have also been widely documented. Witness the road density in Pakistani Punjab, which is just one half of the road density in Indian Punjab. Road maintenance expenditure has been seriously neglected; it was Rs 1.3 billion in 1990-91, far short of the Rs 8 billion required for proper maintenance of the road network. It is also clear that education spending has been seriously neglected. Expenditures on public enterprise in agriculture have been highly inefficient. Although the expenditure of over a dozen public enterprises are not included in the budget, the losses incurred by them are met mostly by borrowing from banks guaranteed by the government, and sometimes directly from the public budget under provisions of lumped grants. There are many public institutions in Pakistan for implementing government interventions in agricultural markets, including output market price control, control of input supplies and external trade of agricultural output. A host of public sector institutions were set up for such controversial reasons as providing cheap food to urban consumers and guarding the farmers against "exploitation" of traders. Most public corporations in agriculture are highly inefficient with high cost of operation compared to the private sector. Audited accounts of many of the corporations show persistent losses; and even those corporations which show financial profits would have economic losses if all the subsidies received by them were taken into account. The role of the govern in providing marketing services in an economy not characterized by market failure is extremely tenuous. A review of marketing institution by the World Bank in 1990 has indicated that many problems could arise with parastatal marketing in developing countries. Some are evident in Pakistan. First, under a system of controlled prices, inadequate marketing margins are the primary reasons for the inadequacy of marketing services provided by both the public and private marketing channels. In Pakistan, fertilizer and seed are heavily subsidized and the state organizations (PSC, PADSC, SASO, and so on) have difficulty in realizing their full operational costs form the consumers. The prices set by the government are also imposed on the private sector, either through administered prices or by the presence of large government marketing agencies. This has resulted in inefficiency of supply, such as untimely availability and poor distribution in low demand areas. Second, for large parastatals in Pakistan, the cost of inefficient operation is reflected both in the price and the level of service. There is evidence that private sector prices have been less than state enterprises in many instances of distribution of fertilizer and seed. Even with the presence of parastatals, private sector marketing channels have continued to service farmers, and are often preferred by them. This seriously calls into question the argument that expansion of public-sector marketing service is needed to check alleged exploitation by merchants. So, there is not justification for public sector presence in marketing and the records show that parastatals have been generally wasteful. The resources devoted to them could have been directed towards expenditure that would benefit agriculture, such as promoting the adoption of productivity-enhancing technology by farmers, building infrastructure to link markets, and supporting private entrepreneurship. iv There is a need to overhaul the public expenditure program in agriculture in Pakistan. The reform of the expenditure program should start with a clear identification of the areas of public involvement in agriculture. Obviously the focus should be on productivity improvement of the sector, best achieved through support of a well functioning, efficient private agricultural sector. This support should include building of strong infrastructural and regulatory structures that foster a competitive and low risk environment. The government should take part in building rural roads to link markets and to provide access to both growers and suppliers of output. It should support a strong research and development program to bring new technology to all farmers. It should assist in the development of rural financial markets and provide regulatory cover for protecting investors and borrowers. More resources should also be diverted towards creation and operation and maintenance of primary public irrigation and drainage networks through which farmers could develop their own operated and maintained water supply systems. Monitoring and regulation of scarce water resources and providing guidelines for environmental conservation are also important aspects of public sector activity that need to be strengthened in Pakistan. Resources for the above and other legitimate public sector activities could be generated from the cuts in such wasteful public expenditure as output and input subsidies and propping of loss making public enterprises. In the financial year 1993-94, almost 1.76 billion rupees out of the agricultural budget of 7.85 billion rupees, or 22.45 percent, were spent on subsidies. Although difficult to quantify in the absence of accurate data, an even larger amount could have gone for supporting loss making public enterprises. The appropriate solution to the problems of both fiscal hemorrhage and market destruction would require the government's withdrawal from agricultural marketing and diverting some of the saved resources towards the activities listed above. V PAKISTAN: PUBLIC EXPENDITURE IN AGRICULTURE 1. Introduction The purpose of this paper is to analyze the trend and composition of the public expenditure (PE) program in Pakistan, with a view to ascertain: a) how well the program conforms/with the government's policy objectives in agriculture; (b) whether the public expenditure program is consistent with the appropriate role of the government (providing enabling environment for private sector agriculture); and (c) how has the public expenditure program affected the performance of the sector. Policy Objectives and Role of Government in Agriculture The public expenditure program in agriculture is a key instrument for implementing government policy objectives in agriculture. Government papers, including various plan documents, outline the broad policy objectives. In 1991, the government developed a National Agriculture Policy (NAP) which incarnated the free market thinking of the later eighties and essential recommendations of the National Commission on Agriculture (1986) incorporated in the Seventh and Eighth (draft) Five Year Plans. The key policy objectives of the government, as expressed in these documents, are to:1 * obtain agricultural growth rate higher than population growth to ensure food security, self sufficiency and export surpluses; * increase productivity of crop sector, livestock, fisheries and forestry sectors; * evolve an export-oriented strategy to exploit export potential; * conserve and develop natural resources; * promote institutional development and bring social and economic equity to the agrarian structure; * focus on small farmers and barani area development; * achieve full employment in rural areas through rural agro-based industrialization. These policy objectives of the government, though broadly defined, provide a clue as to how public expenditure in agriculture would be organized and should also provide a rough bench- mark to judge the efficacy of the size and composition of public expenditures. More importantly, the size and composition of public expenditure program depends on the prevailing belief or norm about what the government role should be in the sector. A review of the public expenditure program will help clarify the framework in which the government performs its role. Ideally, the government role should be to provide an enabling environment for private-sector agriculture, while assisting in reducing rural poverty, and ensuring sustainable resource use. This means that government will have to confine itself to certain functions, the nature and extent of which will be dictated by the nature and extent of market failures in the agriculture sector, including public goods, externalities, moral hazard problems, infant industry situations and monopolies. There are clear areas 'Ministry of Food and Agriculture and Cooperatives, National Agriculture. Policy, 1991 where the need for government intervention is pressing. Specifically, the public sector should have a key role in growth-enhancing public goods mainly agricultural research, extension, public health and rural infrastructure, while correcting environmental externalities associated with forestry, soil conservation and integrated pest management. Even where government has a legitimate role, its actual activity may not be serving public good. While it may appear to be intervening to correct market failure for the public good, in fact, intervention may be serving private interest, including those of the public officials themselves. This paper will show that the government of Pakistan seems to have defined "public goods" or "market failure" too loosely and current public-sector intervention seems over-extended. The accepted principle of state intervention -- that it should be used when there are significant externalities and where individuals do not absorb the full costs or benefits of their action -- mainly apply to creation of public goods as improved agriculture technology, or control of environment degrading activities such as excessive use of chemical fertilizer, or the regulation of plant and animal movements that could be the source of epidemic among crops and livestock. In Pakistan, either too little importance have been given to such activities or there have been actions by the government that run contrary to the avoid goal of intervention. For example, public expenditure on agriculture research have been grossly inadequate and the use of fertilizer and tubewell subsidy have encouraged over-use of both these inputs resulting in run-off poisoning of groundwater and lowering of water tables, thus raising costs for other users of aquifers. Intervention justified on the grounds of important market or political sensitivity have also been misapplied in Pakistan. A case of limited intervention could be seen, due to the importance of food and nutrition in the budget of the poor, by means of well-targeted food subsidies and other measures to enhance the income of the poor farmers. But, in Pakistan the use of a general food subsidy and controlled food prices has defeated the purpose since this type of subsidy has benefited mostly the non- deserving population at huge costs to the state and the price controls have cut down the income of the poor farmers. If price stabilization is desired for basic food commodities, experience from many countries show this to be better achieved through a system of variable border taxes and subsidies instead of direct government procurement, control and ownership of the crops.2 In the final analysis, the sector's need for investment in true public goods have been compromised in favor of food subsidies benefiting mostly the non-rural consumers and input subsidies utilized mostly by large farmers; and on the expansion of agricultural bureaucracies. Budgetary transfers to cover the losses of state marketing organizations, set up to implement the subsidy and interventionist policies of the government, have also taken their toll on the health of the sector. Performance of Agriculture Growth of the agriculture sector, so far, has been good -- more than 3 percent average over a thirty-four year period (1959-60 to 1993-94). Table I shows this growth in different periods. The 2 See Odin Knudsen and John Nash (1990), Redefining the Role ofGovernment in Agriculture for the 1990s, World Bank Discussion Papers, No.105. 2 strong growth in the 1960s was driven by greater confidence in the use of irrigation water, productivity enhancing fertilizer-seed packages, growth of tubewells and some policy changes that increased the profitability of agriculture. The paper will argue that the performance of the agriculture sector could be even better with improvement in composition and efficiency of the public expenditure program. Table 1: Growth in Agriculture in Pakistan 1960-1994 Period Average Annual Growth 1959-60 to 1969-70 4.9 1969-70 to 1979-80 2.3 1979-80 to 1993-94 3.6 Source: Estimated from data in Government of Pakistan, Economic Survey. 2. National Agriculture Expenditure: Trend and Structure Annex I and Chart I (below) show the evolution of real public expenditure for agriculture in Pakistan between 1982-1994. Taking the period as a whole, total public expenditure on agriculture decreased in real terms by an average of 0.57 percent per annum. Development expenditure, fell by approximately 3.69 percent a year while current expenditure increased 3.26 percent. In some countries budget allocation (planned expenditures) differ significantly from actual expenditures. In Pakistan, they seem to be generally close (Box 1) and since 1992, the gap has narrowed. Chart i PATTERN OF AGRICULTURE EXPENDITURE IN PAKISTAN 1400000 1200000 0 Current 10000.00 -o-Developmert -a-Total Expenditure -X--Subsidies 8000 00 01 6000 00 4000. 00 2000 00 0.00 I - - - 1 2 3 4 5 6 7 8 9 10 11 12 Period 1982-1994 3 Box 1 Pakistan: Actual vs. Planned Expenditure in Agriculture A comparison of actual with planned expenditure (chart 2 below) shows that the actual expenditure exceeded the planned figures between 1982-87 and 1989-1992. But from 1992, the gap is narrowing. 14000 00 1 2000 00 10000 00 8000.00 -+-Pliannie d 6000 00 --Act ual 4000 00 2000.00 0 00 - - * 1 2 3 4 5 6 7 8 9 10 11 12 PsrIo.dl1S92.1 994 3. Relationship of Agricultural Expenditure to Overall Public Expenditure. The volume of agricultural expenditure by itself is not a good indicator of the government's commitment to agricultural development. Comparisons have to be made with the rest of the economic sectors to obtain a better understanding. In Annex II, a number of aggregate indices have been presented to show the relative size of agriculture public expenditure in Pakistan. Agriculture PE, as a percentage of the total PE on all sectors, has varied between 3% and 8.5% in the twelve years of review, with an average of 5.8%. Comparing agriculture PE as a percentage of total public expenditure across three South Asian countries, one finds that Pakistan has generally trailed behind India, Indonesia and Malaysia (Table 2). Table 2 Comparative Agriculture Expenditure Ratios in Three Developing Countries (As % of Total Public Expenditure) Country 1984 1985 1986 1987 1988 1989 1990 1991 1992 Avg, India 9.65 7.91 7.87 7.47 8.14 7.47 8.33 8.07 7.48 7.21 Indonesia - - 4.55 8.51 6.73 8.05 7.64 7.92 - 723 Malaysia - - - - 7.63 7.08 6.90 6.47 5.80 6,78 Source: IMF, Government Finance Statistics Yearbook, 1993. 4 A recent study by the World Bank estimated the ratio of the agriculture spending share to the agriculture production share at 0.4 for 40 developing countries and 0.7 for 15 industrial countries (time period 1972-88).3 The same ratio for Pakistan has varied between 0.17 and 0.32 in the past twelve years, averaging about 0.25 (Annex II). If the output (or food) subsidy, which has actually worked against agriculture, is removed from the expenditure the ratio varies between 0.13 and 0.23 with an average of 0.19. By contrast, in India the average ratio between 1985 and 1990 was 0.28. Pakistan's ratio is thus weil below the 40 developing countries stated above, and also below neighboring India, illustrating to some extent that agriculture has been less favorably treated in government expenditure. Chart 2 Pakistan: Agriculture's share of Government Investment to Share of Total Output 0.86' 0.70 0.60 . 0.40l3e 0.30 0.20 0.1 0.00 """""" . . . . . . . . . m 82-33 84 - 85- 86- 187- 88- 89 90 - 91 92 - <with no subsidy) < with subsidy) Polynomial Trend <with no subuidy) Another indicator of government's commitment to agricultural development is the agriculture investment ratio - estimated by dividing the share of agriculture in total government investment (i.e. development expenditure) by its share of the total output. This ratio, also, is relatively low for Pakistan, averaging 0.52 in the past twelve years. Comparable statistics for Mexico (between 1977 and 1984) was 0.91. In India, it was 0.41 between 1985 and 1988, which on the face of it may seem worse than Pakistan. However, the same measure in Pakistan, after taking out the input subsidies from development expenditure, becomes 0.41 (Annex II). What is striking, however, is that the trend has been downwards in Pakistan 3 Blarcom, Knudsen and Nash (1993), The Reform of Public Expenditure for Agriculture, World Bank Discussion Paper No. 216. 4Estimated from Government of India, Ministry of Finance, Budget Volume for 1989-90, Tables S-6 and S-44. 5 from 1988 and is at its lowest level today - i.e. 0.32 (Chart 2), in contrast to India's rising trend (Chart 3). One reason for this particular situation is that the fertilizer subsidy, which comprised a large part of the development expenditure in Pakistan, is being gradually withdrawn without concurrent increase of other investment in the sector. Chart 3 India: Agriculture's share of Government Investment to Share of Total Output nart of the Seventh Plan Period) 0.46 0.44 042 0.41 0.42 0. 039 0.36 : 0.34 .. 1985-86 1986-87 1987-88 Source: Goverrunent of India, Ministry of Finance Budget Volume for 1989-90 4. Composition of Public Expenditure in Agriculture The size and trend of total expenditures in agriculture are only indicative of Pakistan relative position, and they are not good indictors of either the government's commitment to agricultural development or the quality of the expenditure program. The most relevant question is the composition and efficiency of public expenditure in agriculture. Current Expenditure Over 55 per cent of the total government spending in agriculture has gone for meeting current demand. But food subsidy (wheat, sugar and edible oil) has accounted for almost half this current expenditure.5 This ratio could be even higher if proper accounting is done for the recurring expenditure in the Public Sector Development Program (PSDP), a somewhat difficult task under the present reporting system. Current expenditure without the subsidy has averaged 26 percent of the total expenditure in agriculture (Annex 11). A substantial part of the current expenditure goes towards establishment costs. The ratio of establishment cost to the total cost, in real terms, for each subsector in the four provinces is shown in Chart 4. A generally accepted norm for this ratio is 40:60. In Punjab, with the exception of forestry, payroll expenditure has been more than 50 percent, going as high as 90 percent in the cooperation division. In Sindh, 5 Although expenditure on food may be considered not a part of agriculture, it is treated as such in Pakistan's agricultural budget under the Ministry of Food and Agriculture and Cooperatives. 6 payroll expenditure in crop agriculture is close to 80 percent. In NVFP, the situation is only slightly better, with crop agriculture, forestry and irrigation maintaining a ratio of around 50 percent. In Balochistan, the situation is almost as bad as the Punjab, and in some cases worse, as for example in irrigation, where payroll expenditure takes up almost all of the allocation. It is also significant that the ratio has been steadily getting worse over the past 12 years. Chart 4 Establishment Cost Ratios6 for Agriculture Public Expenditure in Pakistan Punjab Sindh 1.00 1.00 080- .80 0.80 ~ -0.80 ... 0.60 .A 0.60 060 V 0.6 0.40 -0.40 0.40- 0.40 0.20 0.20 0.20 0.20 0.00 . .... . . 0.00 0.00 I 82-83 84-85 86-87 88-89 90-91 92-93 0.00 0.00 - - - - _82-83 84-85 86-87 88-89 90-91 92-93 -- Crops -E- Fisheries -+- Livestock ---- Cooperation 0 Forestry ---- Irrigation a Cops - - rrigation Balochistan 1.00 0.80 0.80 1.00 0.80 0.60 0.60 - 0.80 0.60 0.4 a , a * 0.60 040 0. 0.40 0.40 0.20 0.20 0.20 - 0.20 0 .00 . . . 0.00 0.00 1 ::::!:: 0.00 82-83 84-85 86-87 88-89 90-91 92-93 82-83 84-85 86-87 88-89 90-91 92-93 rops Forestry Fisher --Crops - L---Livestock -4-- Forestry ---Uvestock g Irrigation L Fisheries ---Irrigation Development Expenditure The development expenditure is also inflated by the large quantum of federal government subsidy on fertilizer. This subsidy has averaged 31 percent of the total development expenditure between 1982-94 (Table 5). But even with the large subsidy, development expenditure has been less than current expenditure in all but three of the twelve years taken in this analysis, viz. 1982-84 and 1985-86 (Chart 1). Particularly after 1988, there has been an almost consistent fall in development expenditure, in contrast with current expenditure, right up to 1991. Rising briefly in 1991-92, it has 60 6Establishment costs as a percentage of total cost in each subsector. 7 again fallen right through 1994. If the subsidy element is taken out of the development expenditure, the pattern of investment appears as shown in Chart 5. Chart 5 Current & Development Expenditures Excluding Subsidies 450001 - 4000 3500 30 0.... ..... . -.-- Current ..*--e- Deve lopmeant 2500 I ----- Linear Trend 2000 150 0 100 0 500 1 2 3 4 5 6 7 B 9 1 1 1 0 1 2 Table 3 Development and Current Agricultural Expenditures Excluding Subsidies (In millions of constant 1980-81 Rupees) Growth Development Year Current Growth Development Rate Expenditure as Expenditure Rate Expenditure % % of Total % Spending 1982-83 2301 .. 4437 .. 0.66 1983-84 2572 0.12 4156 -0.06 0.62 1984-85 2816 0.10 3801 -0.09 0.57 1985-86 3091 0.10 4827 0.27 0.61 1986-87 3389 0.10 4497 -0.07 0.57 1987-88 3028 -0.11 3864 -0.14 0.56 1988-89 3132 0.03 2904 -0.25 0.48 1989-90 3062 -0.02 3711 0.28 0.55 1990-91 3140 0.03 3580 -0.04 0.53 1991-92 3586 0.14 3464 -0.03 0.49 1992-93 3455 -0.36 3201 -0.07 0.48 1993-94 3644 0.05 2445 -0.23 0.40 Source: Estimated from data in Government of Pakistan, Demand for Grants. 8 It is very clear from the trend shown in this chart that government real investment in agriculture has steadily declined in the past 12 years. The real values of this investment (with subsidy removed) are shown in Table 3. A comparison of actual development spending with the Plan and the Public Sector Development Program (PSDP) allocations shows that during the Sixth Plan period (1983-1988) the actual expenditure was only 66% of the Plan allocation and 85% of the annual PSDP allocations. But the actual spending on fertilizer subsidy (Rs. 6362 million) exceeded the Plan allocation. Current vs. Development Expenditure Trends in current and development expenditures, excluding the corresponding subsidies, show that current expenditure has risen from about half of development expenditure in 1982 to about one-and-half times in 1994 (Table 3 and Chart 5). It is also significant that the growth rate of current expenditure without subsidy has declined slowly in the period 1982-1992, whereas the growth rate of development expenditure without subsidy has fallen more rapidly and has been negative in most of these years. The subsidy free development expenditure fell from 53 percent of total agricultural expenditure in 1982-83 to 31 percent in 1993-94. Declining investments may have affected the pace of agricultural development. In comparison, a large and increasing part of the government spending during the 1980s was in defense, interest payments and subsidies. The share of these three items rose from 39.5 percent in 1979-80 to 52 percent in 1989-90.7 Federal vs. Provincial Expenditure There are differences between the federal and the provincial budgets in intrasectoral allocations because of their respective areas of responsibility. The provinces bear a major part of the operational costs (current expenditure). The share of provincial governments in total agricultural expenditure, without counting the subsidies, have gone up from 59 percent in 1984 to 81 percent in 1994 (Table 4). Their share in total current expenditure also remained high, between 49 percent to 94 percent in the period 1984-1994. Also, their share of the total development expenditure went up from 26 percent to 53 percent in the same period. This shows that relatively less development activities are being undertaken in recent years by the federal government in comparison to the provincial governments. Also, most of the federal development expenditure was consumed by the food and input subsidies. Subsidies The largest fiscal outlay made by the government in agriculture have been subsidies. Subsidies fall under two categories - current subsidy which comprises mostly of food subsidy, and development subsidy which includes fertilizer, seed and tubewell subsidies. The provincial governments have borne a large portion of the current subsidy, but development subsidies have almost exclusively been federal government's responsibility. Combined government subsidies, including food (in current budget) and fertilizer (in development budget) reached a high of Rs.7393 million in 1988-89 which was about 55% of total public expenditure in agriculture (Table 5). In relative terms, subsidies have been a significant part of the total 'John Mellor & Associates (1993), Agriculture Price Study. 9 Table 4 Provincial/Federal Expenditure and Subsidy (in constant Rupees - Million) Year Provincial Provincial Subsidy/ Federal Federal Subsidy/ Expend Subsidy Expend Expend Subsidy Expend (Total) (Total) CURRENT 82-83 2155 na na 146 na na 83-84 2462 na na 109 na na 84-85 3792 1149 0.30 2917 2744 0.94 85-86 4409 1498 0.40 1498 1318 0.88 86-87 5025 1836 0.37 313 112 0.36 87-88 5129 2278 0.44 413 236 0.57 88-89 4173 1209 0.30 4919 4751 0.97 89-90 2894 na na 2670 2501 0.94 90-91 3950 975 0.25 2393 2229 0.93 91-92 4236 821 0.19 2085 1914 0.92 92-93 4114 877 0.21 1196 978 0.82 93-94 3474 963 0.28 540 492 0.91 DEVELOPMENT 82-83 1458 2 .00 4604 1623 0.35 83-84 1259 13 .01 4130 1220 0.30 84-85 1286 13 .01 3665 1137 0.31 85-86 1365 12 .01 5240 1767 0.34 86-87 1633 13 .01 3665 789 0.22 87-88 1638 6 .00 3621 1389 0.38 88-89 1187 5 .00 3150 1429 0.45 89-90 1422 4 .00 2941 648 0.22 90-91 1837 0 0 2422 678 0.28 91-92 1897 0 0 2791 1224 0.44 92-93 1802 0 0 1732 333 0.19 93-94 1306 0 0 1139 309 0.27 Source: Estimated from Government of Pakistan, Demand for Grants. Table 5 Total Subsidy as Part of Total Agriculture Expenditure (Constn t Rupees - Million) 1984-85 1985-86 1986-87 1987-88 1988-89 1989-90 1990-91 1991-92 1992-93 1993-94 5043 4594 2749 3909 7393 3153 3882 3959 2188 1764.00 43.18% 36.66% 25.81% 36.16% 55.09% 31.77% 36.58% 35.97% 24.74% 27.31% Source: Government of Pakistan, Economic Survey 1992-93. 10 expenditure varying between 25 percent to 55 percent between 1984-1993. From 1992 they declined somewhat, falling to 24.74 percent in 1992-93 but rose again to 27.31 in 1993-94. It should be borne in mind that this figure reflects only the budget provision for 1993-94 and may differ from the actual in the final analysis. The slightly downward trend may reflect government's effort to reduce the levels of subsidy. The ratio of current subsidy to current expenditure for the provincial governments, over the past 12 years, have varied from 19 percent to 44 percent. The same ratio for the federal government has varied from 36 percent to 97 percent (Table 4). Similarly, development subsidies borne almost exclusively by the federal government have also been very high, from 19 percent to 45 percent of the development expenditure during this period. The Impact of subsidy on sectoral objectives and performance While the total subsidies dominate the expenditures (as noted), the two broad types of subsidies -- output subsidy and input subsidy -- have different types of impact on sector objectives and performance. Effects of Output Subsidy: Currentbsidsidies, which have gone for supporting food procurement and price stabilization, have been the government's corner stone for food autarky. This is evidenced by the standard deviation of wheat production which has been less than 8% of the mean during the period 1978- 1987.8 Moreover, government had insulated consumers from major swings in wheat prices through massive procurement and marketing operations supported by the food subsidy. State procurement of wheat between 1970-1980 have averaged 24 percent of production and beyond 1980 about 31 percent.9 In comparison, food procurement in Bangladesh and India have been below 10 percent per year. Although the country has achieved some sort of food price stability at a level favorable to consumers, the intervention has come with a price. The level of wheat production has remained low because both farmers and traders were discouraged by low prices. Until recently, Pakistan had used the rationing system to ensure low wheat flour prices to urban as well as some targeted rural consumers. The rationing system is now gone, but it has been replaced by a program of open market operation - buying post harvest stock and releasing it later at a pre-determined fixed margin over procurement price, uniformly all over the country. Such practices have inhibited the growth of private storage, transportation, and active private trade in food grain. The government has also controlled international trade of wheat by prohibiting export and importing irregularly to replenish domestic stock. Domestic prices of wheat, because of these intervention, have been lower than import parity prices. Meeting this price differential, along with the cost of storage and maintaining buffer stocks, have been a heavy fiscal drain for the government. The cost of maintaining food subsidies over the years as a percentage of current expenditure and agriculture GDP are shown in Table 6. The subsidies have been very high ranging from 33 percent to 66 percent of current expenditure and from 1.45 percent to 4.42 percent of the agricultural GDP . 8See Thomas Pinckney (1989), The Demand for Public Storage of Wheat in Pakistan, Research Report No.77, International Food Policy Research Institute, Washington D.C. 9Harold Alderman (1993), "Issues for Food Security in Pakistan," in Agha Saijad Haider et.al. ed. Agricultural Strategies in 1990's, Pakistan Association of Agricultural Social Scientists, Islamabad. 11 Subsidies and direct intervention into the food markets arise from policy makers' belief that without control the price of food is likely to go out of reach of the ordinary consumers. It must be borne in mind, however, that food security could only be ensured by raising the real income of households and increasing the supply of food through freely operating market prices. Without adequate prices in the market, the incentive to increase production does not exist. Price control thus can work against the achievement of food security. Again, if raising real income for the entire population is a policy target, control of food prices runs counter to the interest of the farmers. In Pakistan, the government on one hand is squeezing the farmer's income by forcibly lowering the price of output through administered prices and distribution controls, and on toe other hand trying to alleviate the damage by giving him some help through fertilizer and seed subsidies. On the suiface, this dole may appear to relieve the pain of the producer, but in reality the effect is disastrous for reasons shown in the next section. Overall, the damage done through price controls and output subsidies is far in excess of the explicit relief provided through input subsidies. The picture is even more dismal if we count in the long term harm created by the subsidy regime. Table 6 Food Subsidy and Total Current Expenditure (Including Food Subsidy) (in Millions of Constant 1980-81 Rupees) Year Total Total Current Subsidy as % Subsidy as Subsidy Expenditure Current % Ag.GDP Expenditure 82-83 na 2301 na na 83-84 na 2572 na na 84-85 3893 6709 58% 4.42% 85-86 2816 5907 48% 3.00% 86-87 1948 5337 37% 2.01% 87-88 2514 5542 45% 2.53% 88-89 5960 9092 66% 5.63% 89-90 2501 5563 45% 2.29% 90-91 3204 6344 51% 2.80% 91-92 2735 6321 43% 2.18% 92-93 1855 5310 35% 1.54% 93-94 1764 5408 33% 1.45% Source: Estimated from data in Government of Pakistan, Demand for Grants. Effects of Input Subsidy. There are a number of insidious effects of input subsidy on markets. First, subsidy on such essential items as fertilizer and seeds distorts the input markets, destroys private trade and development of efficient markets. Second, government activities in procurement, distribution and marketing of inputs is almost always inefficient and non-responsive to the needs of the buyers. For example the cost of fertilizer distributed by government agencies in Pakistan have been much higher than the private sector, and this inefficiency loss is borne by the public exchequer. Government agencies are capable of delivering at most average services at average prices unlike the private sector which, due to pressure of competition, could provide efficiency prices and quality services. This is quite apparent in the seed market. Giant government corporations like the Punjab Seed Corporation and the Sindh Seed Corporation have provided only average quality of services at a high cost to the public. Most of these corporations have been running at a loss, as 12 shown later in this discussion. The farmers have been deprived of the benefits of better service, competitively driven prices and the public has paid for the inefficiencies. Third, cheap inputs (in the short run) encourage wastes through overuse, particularly by poor farmers who are not well versed in agronomy, and add to environmental degradation.. Fourth, fertilizer and seed subsidies are essentially regressive since they benefit mostly the larger farmers that use large amount of the subsidized inputs. Fifth, the administration of the subsidy program encourages corruption and rent seeking interests in the functionaries who carry it out. In Pakistan, fertilizer subsidy has represented the largest fiscal outlay of the government after the food subsidy. However, its benefit over a long period of time is questionable. In Chart 6 the relationship between fertilizer subsidy and use of fertilizer have been shown. There is a remarkable lack of congruity between them - the correlation being -0.51 (negative).'0 Between 1983-84 and 1984-85, the use of fertilizer increased when subsidy was falling. From 1985-86 to 1986-87 the subsidy plunged, but consumption of fertilizer shot up. The following year, the pattern reversed for both. Again between 1988-89 and 1989-90, subsidy took a steep dive but fertilizer consumption rose steadily. Since then the consumption remained steady while subsidy went through significant change. These negative correspondence suggest strongly that fertilizer usage rates are not affected by the subsidy. Under the circumstance, the huge fiscal outlay on fertilizer subsidy has been a dead loss to the government. Chart 6 Fertilizer Subsidy and Consumption in Pakistan" 2000.00 -.... . ... 1800.0 1600.0 1400.0 1200.0 10000 800.00 600.00 400.00 200.00 0.00 82-83 83-84 84-85 85-86 86-87 87-88 88-89 89-90 90- 91- 1OPearson correlation cocfficient between subsidy and fertilizer consumption. "Fertilizer consumption is measured in thousands of metric tons. 13 Under the covenants of the Agriculture Sector Adjustment Loan (World Bank), fertilizer subsidies are being cut and are expected to be phased out completely by 1995. Chart 7 depicts the falling trend in fertilizer subsidy in the recent years. Chart 7 Fertilizer Subsidy Trend (Period 1982-1993) 1600.00 1400.00 1200.00 1000.00 800.00 600.00 400 00 200.00 0.00 1 2 3 4 5 6 7 8 9 10 11 The government budgets list the open subsidies fairly clearly. However, a host of indirect subsidies also exist that have serious impact on fiscal deficits. Qureshi (1993) has shown that most of these indirect subsidies, comprising of electricity, water and credit subsidies, have added anywhere from 3.64 percent to 7.67 percent to the overall fiscal deficit.2 The government of Pakistan, like so many other developing countries, has used subsidies to compensate for the explicit and implicit taxes on agriculture imposed by administered prices, trade distortions and exchange rate misalignment. It is estimated that about Rs. 272 billion was transferred out of agriculture due to these taxes between 1981-1990." Although part of this transfer, amounting to Rs. 80 billion, was returned through various public development expenditures, a difference of Rs. 167 billion was leftover as the net cost to agriculture. Moreover, the subsidies did little to reduce the burden of the small farmers. The benefit they provide is, by definition, proportionate to the use of '2Sarfraz Khan Qureshi (1993). "Fiscal Impact of Government Interventions in Agriculture," in John Mellor Associates, Agricultural Prices StuIv Vol.11. The study omits another important subsidy -- seedlings in the forestry sub-sector. 13See Mellor Associates (1993), ibid. 14 the subsidized product, and the very poor farmers did not use much of the subsidized inputs. Most of the benefit of input subsidies went to the large farmers. So, instead of helping the poor, resources were misallocated, rural income distribution was distorted, and environmental degradation perpetuated due to overuse - all at a heavy fiscal cost. In providing subsidized fertilizer to the farmers, the implicit assumptions seems to be that they are too ignorant to understand the cost-benefit relationship of its use and would never use the product if it was priced correctly. Such beliefs have been proved wrong by numerous studies and the real motive of the government behind such moves now seems to be a desire to maintain political support of the farmers by partially compensating them for policies that depress output prices.14 Therefore, on the balance, between taxation of agriculture and the subsidization of agriculture, the sector has been a net looser with an added adverse long term affect on market efficiency. Even the average consumers have not received significant benefits because of leakage through inefficiency of state organizations like PASSCO. Overall Impact of Government Market Interventions and Subsidies. Government price dictation imposes an implicit tax on agriculture. This tax, net of subsidies, is shown in Table 7. The indirect subsidies were the single largest transfer to agricultural producers - totaling about Rs. 23 billion between 1982-90. On balance, the transfer out of agriculture to the rest of the economy averaged over Rs. 26 billion between 1982-90. The negative effects of this transfer, summarized by Chaudhry and Maan (1993)"6 were diminishing private savings and investment potential in agriculture, growth of inefficincies, underpricing of agricultural commodities, falling profitability and reduced private incentives. The policies have thus run counter to the objectives of the National Agriculture Policy (1991) which emphasized the provision of appropriate incentives in output prices for accelerating transition of subsistence farming to commercial farming with high levels of productivity. The fertilizer subsidy policy has also negated the objectives of supporting private sector investment in the production and distribution of all fertilizers. 5. Subsectoral Expenditure The division of public expenditures between subsectors of agriculture are shown in Annex 1. 17 On the average, the largest overall (current and development) allocations were made to irrigation (47%) and crop agriculture including food (40%). Other areas, such as fisheries, forestry and animal husbandry (veterinary) were a small fraction of the two majors items. Looking at development expenditures alone, the average expenditure on crop agriculture and food came to about 1.9 billion Rupees (51%) between 1982- 1994. This was slightly higher than the average expenditure on irrigation - about 1.5 billion Rupees (41%). Over the past two decades, the government laid very heavy emphasis on development of cereal crops, like rice and wheat, for attaining food self sufficiency. This objective has now been attained, by and large, and future strategies may have to be directed towards producing non-cereal high value agriculture products for 14 See Odin Knudsen and John Nash "Redefining the Role of Government in Agriculture for the 1990s," World Bank Discussion paper No.105, 1990. '5John Mellor & Associates (1993), ibid. '6M.Ghaffar Chaudhry and A.H. Maan (1993), in Agha Sajad, et.al ed. Agricultural Strategies in the 1990's, Pakistan Association of Agricultural Social Scientists, Islamabad. " Excluding subsidies and other service expenditures. 15 income generation and poverty reduction. Unfortunately, this does not seem to be happening as apparent from the relatively low allocations being made to fisheries, livestock and forestry sub-sectors. Within the limits of available resource, there are major scopes for decreasing expenditure on some items and increasing it in other categories. Public expenditure should be redirected towards activities that build infrastructure and human capital and those that produce public goods for the benefit of all, such as applied research and drainage. The National Agriculture Policy (1991) has emphasized livestock and fisheries development and forestry management for public benefit.'8 It has also supported programs for rehabilitation of irrigation and drainage systems, agriculture research and extension, and the expansion of marketing rural infrastructure. The future expenditure program has to make higher allocations for major investments in agriculture, such as irrigation and drainage infrastructure, research/extension upgrading and efficiency management and support to private sector development of efficient markets. Table 7 Transfers from-and-to Agriculture Due to Current & Development Subsidies (in Million Current Rupees) Transfer Due to Direct Open Concealed Net Price Output Price Taxes'9 Subsidies0 Subsidies2' Intervention Year Intervention (a) (c) (d) Effect (b) (a+b-c-d) 1981-82 -20,133 -288 1782 936 -17,703 1982-83 -28,346 -428 1980 932 -25,861 1983-84 -24,550 -467 1482 1455 -22,080 1984-85 -26,312 -509 1516 2220 -23,085 1985-86 -30,624 -547 2425 2372 -26,374 1986-87 -34,816 -575 1252 3250 -30,890 1987-88 -27,748 -609 2005 4044 -22,307 1988-89 -23,522 -678 2423 4222 -17,555 1989-90 -55,502 -784 1265 4297 -50,724 Source: John Mellor Associates, Agricultural Prices Study, 1993. Crop Agriculture Crop agriculture has been the main stay of agriculture development in Pakistan. Presently, public expenditure on this sub-sector is close to 3.5 billion Rupees a year (Annex III). The current expenditure is taking a 26 percent bite out of this allocation22, mainly due to rising establishment costs which have taken 18National Agricultural Policy (1991), Ministry of Food, Agriculture and Cooperatives, Islamabad. 19Direct taxes include land tax and Usher- a levy on agricultural produce. 20Open subsidies include subsidies on fertilizer, tubewells, plant protection, and seeds. Some of these are off budget items. 21Concealed subsidies are subsidies for water, electricity and credit. Details of these subsidies are given in "Fiscal Impact of Government Interventions in Agriculture," by Sarfraz Khan Qureshi, John Mellor Associates, Agricultural Prices Study, 1993. 220ut of total expenditure that includes subsidies. 16 bigger and bigger chunks of the allocations in recent years (Table 8). For instance, the ratio of establishment to non-establishment expenditure have risen from less than 1 to nearly 1.5 in the past twelve years. The 1994-95 Provincial Budgets have shown large increases in expenditure allocations towards growth of government employee wages.23 Correspondingly, the non-salary component of current expenditure is declining in nominal terms in many instances. This is a disconcerting trend because with increasing resource squeeze, the dwindling operating funds will not be able to support any useful activities. Table 8 Crop Agriculture: Establishment & Non-Establishment Expenditure Establishment Non-Establishment Ratio Expenditure Expenditure 1982-83 261.31 273.34 0.96 1983-84 297.19 316.22 0.94 1984-85 313.93 417.67 0.75 1985-86 355.73 426.45 0.84 1986-87 378.67 464.65 0.82 1987-88 473.50 402.52 1.18 1988-89 347.77 379.98 0.91 1989-90 431.78 378.61 1.14 1990-91 451.23 378.94 1.19 1991-92 517.45 396.55 1.30 1992-93 526.99 414.91 1.27 1993-94 540.63 381.83 1.41 Source: Estimated from date in Government of Pakistan, Demand for Grants. Another disturbing trend is that current expenditure on crop agriculture is increasing and the development expenditure is declining sharply (Chart 8). The ratio of development to current expenditure has deteriorated from 1:5 to 1:0.92 in the last 12 years. This is a sign of increasing inefficiency in the sub-sector as much more is going for meeting personnel costs than for operational purposes. One of the most important expenditure items in crop agriculture is research and extension. Tie generation of new agricultural technologies by research and their delivery to the farmers by the extension system are key to productivity increase. Managed exclusively by the government, none of these have produced very good results. Looking at the pattern of expenditure on research (Annex IV) one can see many reasons for this problem. First, the research program has lacked proper budgetary support. The average annual crop research and extension expenditures have been only half a percent of the agriculture GDP in the last 12 years - which is way too low (Annex II). Again, out of the total agriculture expenditure, the allocations to research and extension have been only about 4.9 percent in this period of time. This is unfortunate, because many past studies have shown that agricultural research in Pakistan is capable of 23 World Bank, Pakistan Resident Mission 1994-95 Report on the Provincial Budgets. 17 Chart 8 Crop Agriculture Current and Development Expenditures 3000.00 -O-Current 2500.00- -20- Developmeni 2000.00 Constant Rupees 15000 0 1000.0 0 500 00 0.00. 1 2 3 4 5 6 7 8 9 1 Period 1982-94 0 1 2 Table 9 Estimates of Marginal Products and Marginal Internal Rates of Return for Agricultural Research in Pakistan Crop Time Period Estimated Marginal Estimated MIRR (%) Product Pakistan Wheat 1955-85 16.52 Cotton 1955-85 43.52 Rice 1955-85 27.30 Sugar 1955-85 Negative Wheat 1956-85 .. 76 Cotton 1956-85 .. 102 Rice 1956-85 84-89 Sugarcane 1956-85 .. Insignificant India Wheat 1959-83 .. 50 13razi Wheat 1974-82 .. 59 Source: Evenson & Bloom (1993) 18 producing high rates of return to investment.24 Estimated marginal products of research and marginal internal rates of return on few essential food crops in Pakistan demonstrate this amply (Table 9). Pakistan has obviously underinvested in agricultural research. Second, the proportion of establishm-nent (wages and salaries) expenditure is too high in the research budget (Table 10 & Chart 9). The 12 year average for this ratio in Punjab is 73:27 and in Sindh 59:41. Table 10 Establishment to Total Expenditure Ratio for Agriculture Research Year Punjab Sindh 1982-83 0.64 0.59 1983-84 0.73 0.56 1984-85 0.69 0.66 1985-86 0.67 0.65 1986-87 0.70 0.59 1987-88 0.73 0.58 1988-89 0.70 0.55 1989-90 0.77 0.56 1990-91 0.79 0.62 1991-92 0.82 0.17 1992-93 0.79 0.70 1993-94 .. 0.87 Average 0.73 0.59 Source: Estimated from data in Government of Pakistan, Demand for Grants. A reasonable ratio of wage to non-wage expenditure in Pakistan has been estimated at 40:60, due to the generally lower cost of labor in relationship to other recurrent costs. By this standard, it is not surprising that effectiveness of research has declined in some provinces as a consequence of low material support. The underinvestment in research is worsened by lack of proper operational funding for scientists and absence of control over appropriate areas of research. For instance, sugarcane research has not yielded much in terms of economic benefits. 24Robert E. Evenson and Erik A. Bloom (1993), "Research and Productivity in Pakistan Agriculture," in Agricultural Strategies in the 1990's, op.cit. 19 Chart 9 Crop Agriculture: The Trend in Establishment Expenditure 500.00 400 00 200.00 . E lat,lrnr,I E *r,-rrjIure 100.0 - 0.00 1982- 1983- 1984- 1985- 1986- 1987- 1988- 189- 1990- 1991- 1992- 1993- Fisheries, Livestock and Forestry Public expenditure in the fisheries, livestock and forestry subsectors have been marginal. Their combined average share in the national budget has been less than 12 percent (Annex III). This is low, particularly in consideration of the special emphasis on livestock and fisheries given in the National Agricultural Policy. Livestock contributes about 7 percent to the GDP and about 32 percent to the agricultural GDP. In relations to this, its share of only 6 percent in the agricultural public expenditure is perhaps too low. The subsector has shown very encouraging growth potentials in recent years with the emergence of large commercial dairy and poultry businesses. It is also holds one of the keys for diversification of agriculture, away from cereal production, and a source of increased nutrition for the rural poor. Presently, the fisheries subsector contributes about 1 percent to the national GDP and 4 percent to the agricultural GDP. Against this, it receives only about 1% of the total agricultural PE. The importance of fisheries in Pakistan's economy arises from the fact that it has considerable potential for export markets in addition to being a high nutritional diet substitute for the domestic population. Present earnings from shrimp and fish exports are about 2.67 percent of total exports.25 25 Fish and fish preparation export in 1992-93 was Rs. 4,733 million out of the total export of Rs. 177,028 million. Source: Economic Surey, 1993-94. 20 Forestry has fared much better than the other minor agricultural subsectors in the public expenditure program. This subsector contributes a little more than quarter percent to the national GDP and about 2 percent to the agricultural GDP. In comparison, it receives over 5 percent of the agricultural expenditure. However, massive resource depletion in recent years, brought about by population pressure, lack of policy initiatives and insufficient inputs including reforestation programs, necessitate considerable attention and budgetary support to this subsector. Irrigation At present, irrigation consumes about 54.6 percent of the total agricultural public expenditure, 54.5 percent of the current agricultural expenditure and 54.7 percent of the agricultural development expenditure (Annex III). Current spending in irrigation is largely a provincial responsibility involving the operational aspects and about 38 percent of the PSDP for irrigation consists of federal programs. There have been some questions about the wisdom of continuing the federal government dominance in water resource development.26 It stands to reason that the federal government should continue supporting inter-provincial irrigation works, and the provinces should take up most of the schemes that are exclusively for their own benefit. A number of important issues in the development and management of irrigation and drainage should be addressed through the public expenditure program. First, the investment priorities in future should be towards stopping the severe natural resource degradation due to waterlogging and salinity. Drainage, since it falls in the domain of public goods, should be given emphasis in the public expenditure program. Second, the government should withdraw from the development and subsidization of tubewell irrigation because it is a private good and use the freed resources for servicing other immediate needs. Third, expenditure on O& M of public irrigation assets in the past have been inadequate and must be increased to the required level on pains of loosing the entire assets. O&M of Irrigation and Drainage Systems O& M expenditures have averaged about 48 percent of the current budget for irrigation for the whole of Pakistan (Table 11). There have been differences in this ratio between provinces, but nowhere it has been adequate to handle the big problem. The country's extensive irrigation and surface drainage system (about 16.2 Million ha) has deteriorated substantially due to deferred maintenance. Through ISRP-I (FY 83- 88) the provinces had caught up with the maintenance backlog and in ISRP-II they had resolved to increase O&M allocations to meet the increasing needs. However, this decision was not implemented, and a fresh maintenance backlog has built up in all the provinces, with the exception of NWFP (Table 12). 26See World Bank, Pakistan Public Expenditure Review, 1993. 21 Table 11 Irrigation O&M Expenditure (in Million constant Rupees) Year Current Current O&M- Punjab Sindh NWFP Baloch Irrigation Irrigation Budget Budget O&M Ratio 1982-83 1310 721 55.03% 548 97 58 17 1983-84 1430 798 55.80% 602 107 71 18 1984-85 1570 879 55.96% 656 117 87 18 1985-86 1708 936 54.79% 677 124 103 32 1986-87 1875 969 51.69% 685 129 125 30 1987-88 1452 699 48.11% 426 122 124 26 1988-89 1724 754 43.77% 487 108 149 11 1989-90 1603 693 43.23% 462 82 144 4 1990-91 1597 727 45.51% 443 143 139 2 1991-92 1932 906 46.92% 650 105 149 2 1992-93 1768 621 35.15% 416 72 130 3 1993-94 1967 882 44.83% 667 84 129 2 Source: Estimated from data in Government of Pakistan, Demand for Grants. Table 12 Difference Between O&M Allocation and Need for Canals under ISRP Targets (in %) 1988-89 1989-90 1990-91 1991-92 Operation Punjab -42.08 -53.19 -52.78 -53.80 Balochistan -31.41 -35.50 -52.78 -52.52 Sindh -3.83 -8.04 -10.85 -26.53 NWFP 13.51 3.34 11.81 10.63 Establishment Punjab 15.93 28.03 22.52 40.95 Balochistan 3.13 8.13 8.70 51.08 Sindh -5.32 -3.27 0.51 9.15 NWFP 15.49 7.64 10.27 58.01 Source: World Bank, Pakistan Public Expenditure Review, 1992. Analysis done by the Bank shows that in Punjab the budgetary allocation is 12 percent lower than the O&M requirements for FY93-94, as defined under IDA assisted Projects. For Sindh, the allocation is 42 22 percent below the targets for FY93/94. For Balochistan, the allocation is 25 percent less than the target. Only in NWFP, the FY93/94 allocation exceeds the target (by 16%). Part of the funding problem for O&M stems from the fact that that the government has not been able to achieve adequate cost recovery through appropriate adjustments of the water rates and/or drainage cess. If the target level of O&M funding is to be provided in Punjab, cost recovery has to be increased by 56 percent to meet the Fordwah covenant of 45 percent recovery by July 1, 1994. Unfortunately, the collection rate in Punjab has been less than 60 percent of the assessed amounts during the last three years. In Sindh, the gap between recovery and requirement was more than 40 percent on July 1, 1993. According to the covenants of ISRP-II, Sindh should achieve full cost recovery by July 1997. This means that the water charges would have to be further increased. Incidentally, in Sindh, full cost recovery does take place, but there are leakage at the assessment phase. It would be better for Sindh Government to transfer the assessment work from the Revenue Department to the PID which has shown 20 percent higher assessment in pilot areas. NWFP and Balochistan are not required under ISRP-1I to achieve full cost recovery by 1997, but are required to increase water charges along with Punjab and Sindh. NWFP has increased rates by 25 percent recently, following Punjab. Balochistan had increased water rates by 25 percent in 1992, but did not follow the increase by Sindh in August 1993. Although better collection would improve cost recovery, by itself it would not be sufficient to bridge the cost-revenue gap. A host of other measures will be necessary to stop the hemorrhage in the public budget including rationalization of water charges and drainage cess, increasing system efficiency, and transferring responsibilities involving the generation of private goods to the beneficiaries. 6. Expenditure on Public Institutions Growth of public enterprises reflect a number of policy pursuit of the government such as filling in market deficiencies, balancing regional development, egalitarian distribution of income, and prevention of concentration of economic power. Government ownership, however, by itself may not ensure efficient allocation of resource and a certain degree of market behavior, particularly the acceptance of opportunity cost pricing is generally needed.28 In many developing countries, the belief that distribution is the mandate of the public sector have led governments to make massive investments in public enterprises. In general, most of the public enterprises have pursued non-commercial objectives, although it has been strongly argued that they should be commercially oriented like any other private sector counterpart. In other words, public enterprises should work on the same principles of efficiency as a private enterprise.29 Therefore, it is expected that even though public enterprises 27 Source: Pakistan Resident Mission Report. 28 See A. Premchand, Government Budgeting and Expenditure Controls, IMF Publication, 1993. 29 See United nations, Improving Profit performance ofPublic Enterprises in Developing Countries (1969). 23 should respond to general public demand in a positive way, they should at the same time protect the interest of their owners (the public) and their consumers.30 There are a number of public institutions in Pakistan for implementing government policy of market interventions, including output market price control, input supplies control, external trade of agriculture output, etc. Government's involvement in the markets for agricultural products is sometimes necessary for the explicit purpose of addressing market failures, but in Pakistan the reasons have been quite different. A host of public sector corporations were set up, for such controversial reasons as providing cheap food to urban consumers and guarding the farmer's against private 'exploitation'. Notable among these agencies are the Punjab Seed Corporation, the Sindh Agricultural Services Organization (SASO), Punjab Agricultural Development and Supplies Corporation (PADSC) Pakistan Agricultural Services and Storage Corporation (PASSCO), Agriculture Development Authority (ADA), Cotton Export Corporation (CEC), and Rice Export Corporation of Pakistan (RECP). All of these corporations are treated outside the budget as presented in the Federal and Provincial Demand for Grants, and their accounts are difficult to audit. Nevertheless, they have been consistently criticized for inefficiency and high cost of operation compared to the existing private sector. For example, a study in 1989, showed that the private rice mills operated at 40 percent less cost than the RECP. Table 13 summarizes the operating profit and losses of four major agriculture sector corporations in Pakistan. Their corresponding income statements are shown in Annexes V. Table 13 Consolidated Profit & Loss of Major Public Corporations in the Agriculture Sector (in Million Current Rupees) 1988 1989 1990 1991 1992 1993 PSC 1.74 4.38 18.99 14.65 21.66 31.00 SASO NA -10.01 -42.37 -50.84 -40.82 58.77 PADSC -93.14 -118.70 -101.29 -39.62 -9.50 -17.62 PASSCO -281.60 -170.60 78.40 86.00 56.00 -89.00 Source: Corporations' Annual Accounts Data. Audited accounts of the Punjab Sindh Corporation (PSC) show that the company recovered from early losses and had profits in the past six years. Although, on the surface this corporation does not seem to have done too badly, certain expenditure items were not taken into these accounts whose inclusion would have in all certainty depressed the performance picture. First, the seed farms acquired from the Punjab Agriculture Development and Supplies Corporation (PADSC) were not valued. The implicit land rent of these farms must be added to the operating costs to get an unbiased estimate of costs. Second, the corporation's loans from the nationalized banks were underwritten by the government and given at around * Sec Premchand, ibid. 24 12.5 percent. This is an indirect subsidy whose cost must be added to the operating cost for correct evaluation. Third, The corporation had borrowed working capital of Rs.50 million from the Government of Punjab, but has not admitted the interest on this loan fully yet. Fourth, the corporation had received occasional grants from the Government of Punjab for expansion of facilities, whose effect were not fully reflected in the operating statements. The Sindh Agricultural Supplies Corporation (SASO) distributes fertilizer and seed to farmers in Sindh. The corporation receives reimbursement of the distribution costs on a predetermined formula, but it is not sufficient to cover the costs. As a result, the organization has had operating losses in all the years for which data was available (1989-93). Available data for the Punjab Agricultural Development and Supplies Corporation (PADSC) also show net operating losses in all years between 1988-93. The Pakistan Agricultural Storage and Services Corporation (PASSCO), which is engaged in procuring, preserving, storing and distributing food grains (especially wheat) has been operating with a budget, close to 3 billion rupees. The operating finance of this organization shows major losses prior to 1990, which were partly offset in 1990, 1991 and 1992. But the corporation again went in the red in 1993. The role of the government in providing marketing services in an economy not characterized by market failure cannot be justified. A review of marketing institution by the World Bank in 199031 has indicated that a number of problems could arise with parastatal marketing in developing countries. Some of these are examined in context of Pakistan below. a) Under a system of controlled prices, inadequate marketing margins are the primary reasons for inadequacy of marketing services provided by both the public and private marketing channels. In Pakistan, fertilizer and seed are heavily subsidized and state organizations like PSC, PADSC, SASO, etc. have difficulty in realizing their full operational costs from the consumers. The prices set by the government are also forced on the private sector, either through administered prices or by the very large presence of the government marketing agencies. This has resulted in inefficiency of supply, such as untimely availability, poor distribution in low demand areas, etc. b) For large parastatals, the diseconomies resulting from poor management tend to outweigh the potential gains. The cost of inefficient operation is reflected both in the price and the level of service in Pakistan. There are evidence that private sector prices have been less than state enterprises in many instances of distribution of fertilizer and seed in Pakistan. c) The private sector also has better performance records where the regulating agencies are alert -- a factor lacking in Pakistan. d) Even with the presence of parastatals, private marketing channels have provided valuable services to farmers, and are often preferred by them. This is very true in Pakistan and it thus seriously calls into question the argument that expansion of public sector marketing services is necessary to check alleged exploitation by merchants. 3See Agricultural Marketing: The World Bank Experience, 1974-85 (1990), Operations Evaluation Department, The World Bank, Washington. D. C. 25 Even successful parastatals tend to have difficulty maintaining good management and effective services. In Pakistan, organizations like the Sindh Seed Corporation and PASSCO have tremendous problems of overstaffing and control. This has reflected heavily on the efficiency of their operation. The existence of the large parastatal marketing organizations in Pakistan is not only costly to the government but also costly to the economy in both the short and long run. Apart from the inefficiency costs attached to their operations, these institutions compete with the infant private sector on a non-level playing field. They are often represent monolithic public monopolies, enjoy tremendous government patronage and arrogate financial resources from the market with state backing. The net result is a crowding out effect that leaves the private sector weak, non-competitive and thereby exploitative whenever it gets the chance to be so. The losers are the public and the productive sector. The monopoly powers of state enterprises could be a major impediment to the development of Pakistan's economy because it drives away private initiative and competition. Perhaps the cardinal principle that needs to be acknowledged, first, is that farmers have economic rights to produce whatever they wish, using technology they feel best and sell their products freely at home or abroad. The traders, likewise, have the right to move the goods in their best interest without obstruction from the government or any authority, and the consumers have the right to buy their products at the best price from whatever source, domestic or international.32 Markets work best when individuals are allowed to transact freely, and even in some cases when "private" exploitation becomes an issue, it is good to ask before any intervention whether the market failure is likely to be more serious than government failure. The best cases for public intervention is where externalities exist and individuals do not absorb the full cost or the full benefit of their action. Government's participation should be confined to the creation of such public goods only. This includes basic or applied research for development of inputs, outputs or techniques that could not be patented, prevention of environmental degradation through overuse of chemicals and fertilizers that pollute streams and rivers, and regulation of such activities that create harmful externalities. There may be occasions requiring government intervention in the market to address poverty issues such as nutritional intakes of the poor in the country. Providing targeted food subsidies to the poor, and income support to the very poor farmers are all right as long as they do not distort the price signals and if they go towards addressing the problem at the root rather than the symptom. For example, if it is absolutely necessary to regulate basic food prices for providing access to a very large segment of a poor population, it may be better to do so through a system of variable border taxes and/or subsidy, rather than direct procurement, and distribution.33 This would also minimize the distortion of prices in the distribution chain that arises due to control over distribution margins. Besides it would eliminate the need for maintaining large, inefficient public procurement, storage and distribution organizations that become a drain on the public budget. 32 Much of the arguments made here are derived from Knudsen and Nash, op. cit. 33See Odin Knudsen and John Nash (1990), Redefining the Role of Government in Agriculture for the 1990s, World Bank Discussion Papers, No. 105. 26 7. The Final Assessment An overall assessment of the composition and efficiency of agricultural public expenditure in Pakistan could be made in perspective of the major objectives of the National Agricultural Policy. Looking at the composition, first, we see a significant deviation between the actual expenditure and the avowed objectives. Over 55 percent of the total agricultural budget is going for non-development expenditure, of which a sizable chunk (28 percent) is taken up by food subsidy. An important goal of the NAP is to reduce the role of government and bureaucratic constraints on economic interactions between productive enterprises. Obviously, this cannot be accomplished without cutting down the size of current expenditure. The government has too large a payroll burden (between 50 to 90 percent of current expenditure). Although, to some extent this may be justified in institutions like agriculture research, in Pakistan the payroll weight has been excessive in other operational and non-operational areas. Development expenditure has also been falling in relationship to overall expenditure (down form 62 percent in 1984 to 40 percent in 1994.) This negates the investment and growth objective of the NAP. Also, relatively less of the development expenditure is being made by the federal government in comparison to the provincial governments, and most of it is going for input subsidies. This does not auger well for development and growth of the agricultural sector which are important elements of the NAP. The division of government expenditure between subsectors has not been erairely consistent with NAP objectives. Irrigation absorbs the largest share (54 percent) followed by crop agriculture and food (29 percent). The remaining subsectors, including forestry, fisheries and animal husbandry receives but a fraction of the total budget. This is a serious limitation on the growth of these special high value product subsectors and runs against the objectives of the NAP. Even though irrigation receives such a high proportion of public expenditure in agriculture, the allocation towards O&M have been inadequate causing widespread deterioration of irrigation facilities. This fact has been well recognized in the NAP which calls for steps to redress the deficiency. Thus, on the whole, it can be said that the composition of agricultural expenditure in Pakistan is not consistent with either its National Agricultural Policy or the appropriate role of government in agriculture. Agricultural expenditure in Pakistan has not been efficient either. The government has spent between 25-55 percent for agriculture subsidies in the last ten years, mostly in support of food and fertilizer operations. It is obvious from the arguments presented earlier in this paper that food subsidies have distorted market prices and hurt the process of gaining self sufficiency in essential commodities - a key objectives of the NAP. Likewise, fertilizer subsidies and state marketing activities have hindered the development of private enterprises as perceived in the national plan. Besides, they did not make a significant difference in the uptake of fertilizer by the farmers and thus, have not contributed very much to productivity growth. The large number of public corporations, bome on the budget, in pursuit of marketing and distribution policies, have actually hindered the privatization program and the transfer of agricultural spending from public to private hands. Most of the public enterprises have been a drain on valuable public funds. One redeeming factor is that the pattern of public expenditure has changed for the better in certain categories in recent years. Some key indicators of this change are shown in Table 14 below. Between 1984 27 and 1994, current subsidy as a percentage of current expenditure dropped from 58 percent to 29 percent, which is quite encouraging. Likewise, the share of irrigation O&M in the current budget rose from 13 to 17 percent. Expenditure on agricultural research, however, showed only a marginal increase. Development subsidy as a percentage of development expenditure also went down from 23 percent to 21 percent. There were some regressive tendencies in expenditure also. For example, the ratio of current to development expenditure rose from 1.36 in 1984 to 1.85 in 1994. Personnel costs also increased in a large way, going up from 36 percent to 83 percent in one area of agricultural activities -- agricultural research. Table 14 Change in Selected Indicators of Expenditure Efficiency 1984-1994 Expenditure Ratio 1984-85 1993-94 Current/Development 1.36 1.85 Current Subsidy/Current Expenditure 0.58 0.29 Irrigation O&M/Current Expenditure 0.13 0.17 Agricultural Research/Current Expenditure 0.010 0.013 Personnel/Non-Personnel Exp. in Agriculture Research & Extension 0.36 0.83 Development Subsidy/Development Expenditure 0.23 0.21 In short, government policies have not been generally successful in accomplishing specified or implied goals and the expenditure priorities have not supported agricultural strategies. There is a lot of room for improving efficiency of resource use, overall and within specific activities undertaken by the public sector. 8 . Conclusion The foregoing discussions have shown how the government's expenditure on agriculture in Pakistan have been costly and misdirected. The cost of market intervention, for reasons unjustifiable on economic and social grounds, have been colossal and there is no evidence that they have helped the agricultural sector in any way. There is clearly a pressing need for reforms, starting with a reconsideration of the future role of the government and corresponding institutional changes. The former could start with an acceptance of the principle that free markets provide better opportunity to farmers than government controlled markets and government failures could be more serious and demoralizing than market failures. The only legitimate intervention in the market exists where strictly public goods are involved, which enhance the productivity of all participants and produces a sustainable environment for growth. The present priorities of government spending in agriculture must be reversed and moved towards maximum efficiency. Without increasing the overall expenditure, it is possible to achieve higher efficiency by cutting parts of the budget and shifting resources to productivity enhancing activities. A look at the budgeted expenditure, below, in various categories during FY1993-94 shows that out of 7.8 billion rupees spent on agriculture only 3.49 billion rupees (44.4 percent) went to growth enhancing activities like irrigation and research. On the other hand 1.45 billion rupees (18.5 percent) were spent on untargeted food subsidies 28 (with marginal benefits to the poor) and an additional 0.31 billion (3.95 percent) on fertilizer subsidy most of whose benefit was reaped by the richer farmers. Million Rupees Current expenditure without Irrigation and Agriculture Research 1611.77 Irrigation expenditure without O&M 1085.43 Irrigation O&M 882.00 Agriculture Research operation 65.27 Untargeted food subsidies 1455.00 Fertilizer subsidy 309.00 Public investment without Irrigation and Research 987.13 Public investment in Agricultural Research 119.70 Public Irrigation investment 1338.19 Total 7853.49 If the expenditure on food and fertilizer subsidies are eliminated, the irrigation and research budget could be enhanced almost fifty percent. Also, a much needed support could be given to groundwater monitoring and regulation without which the ongoing privatization of SCARP tubewells is likely to environmentally sound. Outside of political consideration, such actions would be budget neutral since they involve mere category switching and no further demand on the budget. Their benefits would be great, however, since they would move agricultural prices closer to market values and remove the disincentives for private enterprises as discussed earlier, in addition to providing a boost to growth enhancing activities that fall within the legitimate realm of government actions. In addition all other implicit subsidies and transfers to public enterprises should cease. The primary objective of expenditure should be productivity and growth enhancement and not income transfer. Proper policy in support of a vigorous private sector would take care of the income redistribution objectives. In the interim period, any welfare objectives, like nutrition support for the rural poor, should be clearly identified and specifically targeted, as done in many countries like the Food for Work Programs in Bangladesh. Subsidies to promote adoption of improved technology may be used in special circumstances, but only as a temporary short-term measure and removed once the process is well underway. The markets for agricultural goods and services should either be privatized or there should be a concerted effort to involve increasing private sector participation by introducing cost recovery measures or private contracting government agricultural services.14 3 See Blarcom, Knudsen and Nash (1993), The Reform of Public Expenditure for Agriculture. 29 C, 41