Report No. 1 3092-PAK Pakistan A Strategy for Sustainable Agricultural Growth - World Bank ·...

Report No. 1 3092-PAK

PakistanA Strategy for Sustainable AgriculturalGrowthNovember 3, 1994

Agricultural Operations DivisionSouth Asia Region

Document of the World Bank

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

CURRENCY EQUIVALENTSRupee 1.00 = USS0.033

USSI = Rs. 30.0

WFIGHTS AND MEASURFSMetric System

I Acre = 0.405 haI Acre foot = 1.233.S cubic meters

GOVERNMFNT FTSCAT. YFARJuly I - June 30

ABRRFVTATTONS AND ACRONYMS

ADA Agrnoualral Derdopnemnt Authority MLG&RD Ministry of Local Government and RuralADBP AgriculosnJ Development Bank of Pakstan DerelopnAMRJ Agricultural Machinery Rescarch Institute NCBs Nationl Commtmercialized BalnksAPCOM Agricuulal Prices Conumission NFC National Fertilizer CorporationARI Agricultural Research lnstintze NRP Nominal Rate of ProtectionATI Agricultural Training Inticute NWFP Northwest Frontier ProvinecAUP Agricultnrl Uriversiry of Pcshavar NFML National Fertilizer Marketing LimitedAUF Agriculzural university of Faisalabad O&M Operations and MaiucnanceAZRI Arid Zone Rescarrh Institute OFWMD On-Farnn Water Management DirectorateBalo Balochisg n PACO Paistan Agicultunal Census OrganizabonBAC Balochistan Agneulural College PAD Provincial Agriculrtre DeparutentBARB BalocJistan Agricultural Research Board PAED Provincial Agricultural ExtensionBoR Board of Revenue DeparmrentCEC Cotton Expon Corporation PARB Punpab Agricultural Research BoardCollector Distria Collector PARC Pakistan Agricultural Research CouncilCornm Banks Commercial Banks PADSC Punjab Agricultural Development andCRD Crop Reporting Directorate Supplies CorporationCRMR Crop and Resource Management Research PASSCO Pakistan Agricultral Services and StorageDC Disrirt Council CorporationDRC Domestic Resrouce Cost PCCC Pakisan Cental Cotton CommrniteeDRIP Drainage and Reclamation Institute of PCD Provincial Cooperatives Departnent

Pkistn PDF Provincial Deparomtent of FoodEngD Engineering Directorate PEPA Pakistan Environrmenal Protection AgencyEPA Environmental Protection Agency PFD Provincial Forest DepartmentERP Effective Rate of Protection PFI Pakistan Forest InstituteEW Etonomic Wing PFsD Provincial Fisheries DepartmentFBC Federal Bank of Cooperatives PID Provincial Irrigation DepartmentFID Fenilizcr Irnport Dcparener PIRI Punjab Irrigation Research InstitutcFS Forest School - PLD Provincial Livestock DepartmnentFSCD Federal Seed Cerification Depart'nent PPPD Pakistan PlaLns Protecaion DepartmentFTM Farm-to-Market PSG Punjab Seed CorporationFWMC Federal Water Management Cell PSE Producer Subsidy EquivalentFY Fiscl Year RECP Rice Export Corporation of PakistanGATT Gneral Agreerent on Tariffs and Trade SARO Sindh Agricultural Research OrganizationGCP Ghee Corporation of Pakistan SASO Sindh Agricultural Services OrganuzationGDP Gross Domestic Product SAU Sinddh Agricuiranl UniverairyHa Hectare SBP State Bank of PakistanIBIS Indus Basin Irrigation System SCARP Salirity Control and Reclanution ProjectIWARSI International Waterlogging and Research SCD Soil Conservation Directorate

Institute SD Statistics DivisionKg Kilogranrnme SDIR Sutistics DirectorateKARINA Karrakoram Agricultral Research Institute SRPO Sindh Regional Planning Organization

for Northem Areas SSC Sindh Seed CorporationLG&RDD Local Govermc ent and Rural Development TFP Total Factor Productivity

Department T&V Training and Visit Extension SystemNIFD Marine Fisheries Deparunrun UNESCO United Nations Educationala Scientific, andMINFAC Minisuty of Food. Agrculsturc and Culntral Organization

Cooperatives WAPDA Water and Power Development AuthorityMIRI Mona Irrigation and Reclamation Institute WUA Water User AssociationMOW&P Ministry of Water and Power

Glossaz

kharif - wet season (mid-April to mid-October)mogha - outlet from parent supply channel to teniary irrigation commandpatwari - Board of Revenue functionary at local levelrabi - dry season (mid-October to mid-April)warabandi - weekly rotational schedule of irrigation delivery to farmerswatercourse - irrigation distribution system within a tcrtiay irrigation command

TABLE OF CONTENTS

EXECUTIVE S UMMA R Y ................................................... i

CHAPTER I. The Agriculture Sector: Role, Prospects, and Objectives. 1

I. Role of the Agriculture Sector ............................... 1

Agriculture and Growth ....................................... 1

Agriculture and Adjustment .................................. 2

Agriculture, Poverty, and Inequality ......................... 4

Agriculture and Environment ................................. 6

II. Past Performance. 6

Overall Trend. 6

Assessment of Performance .10

Productivity Growth .11

III. Prospects and Objectives .12

Supply Potential .12Demand Prospects .14

Sector Objectives in the Future .14

CHAPTER II. Resource Constraints and Related Issues .17

I. Land Resources and Constraints ............................... 17

Soil Erosion ................................................. 18

Irrigation-Related Problems .................................. 19

Inappropriate Use of Soil Resources .......................... 20

Forests and Rangeland ......................................... 21

II. Size, Tenure, and Productivity ............................... 21

III. Water Resources and Constraints .............................. 27

The Rainfed Areas ............................................ 30

IV. Human Resources .............................................. 30

V. Rural Infrastructure ......................................... 32

This report is based on a field mission to Pakistan in January/February 1994. The missionconsisted of Rashid Faruqee (Leader), Kevin Carey (Consultant), Derek Byerlee (AGR), YusufChoudhry (Consultant), and Omar Noman (Consultant). Rashed Qayyum and Qaiyum Sheikh from RMPworked closely with the mission. The report, prepared by Rashid Faruqee with help from KevinCarey, is based on extensive consultation within and outside the Bank, as well as with expertsand policy makers in Pakistan. Several seminars were held in the Bank on selected issues ofagricultural growth, and similarly, two workshops were held in Pakistan to design the study.Background papers were contributed by Sarfraz Quereshi, Moazam Mahmood, Nadeem Ilahi, A.R.Saleemi, Tayyeb Shabbir, Shahid Zia, and Waqar Malik. Rashed Oayyum and Qaiyum Sheikh helpedwith institutional issues and Yusuf Choudhry prepared a paper on public expenditure inagriculture. Derek Byerlee and Omar Noman also contributed background papers. The green-coverdraft of the report was extensively discussed with federal and provincial governments in Pakistanin October, 1994. The Country Director is Mr. Paul Isenman (SA3DR) and the Division Chief is Mr.Ridwan Ali (SA3AG). Professor T.N. Srinivasan and Mr. Hans Binswanger are peer reviewers.

1

CHAPTER III. Policy Distortions .................................... 35

I. The Incentive Effects of Price and Trade Policy ... ........... 35

Taking Account of Input Prices ............................... 39

Sectoral Policy Bias: Comparison with Other Countries ... ...... 43

II. Resource Transfer and Tax Reform ............................. 44

Tax Reform Proposals ......................................... 46

III. Policies Affecting Input Markets ............................. 48

Agricultural Credit ........................................... 48

IV. Effect of Distortions on Growth and Poverty .................. 51

CHAP"TR IV. Public-Private Sector Role, Institutions, and Public

Expenditure ........................................... 53

I. What Should be the Public and Private Sector Roles in

Agriculture? .................................................. 53

II. Institutions and Services .................................... 55

III. Public Expenditure ........................................... 62

Subsidies .................................................... 62Inadequate Expenditures ...................................... 69

Inefficient Expenditure on Public Enterprises in Agriculture .. 70

CNAPTER V. The Strategy .. 73

I. Policies Affecting Incentives ................................ 73

II. Policies Affecting Input Markets ............................. 74

III. Government Expenditure ....................................... 76

IV. Poverty ...................................................... 78

V. Environment .................................................. 78

VI. Core Elements of the Strategy ................................ 80

Moving Towards Market ........................................ 80

Addressing the Irrigation Crisis .............................. 80

Reforming Extension .......................................... 81

Correcting Distortions in the Land Market ..................... 81

Bibliography ................................................... 82

2

LIST OF TABLES

Table 1.1 Survey of Estimates of Headcount Poverty in Pakistan.... 4

Table 1.2: Percentage of Population Below the Poverty Line ........ 5

Table 1.3 Social Indicators in Pakistan and India ................. 6

Table 1.4: Growth in Agriculture, 1960-1993 ........................ 7

Table 1.5 Growth of Agricultural Production and Population ....... 10

Table 1.6 Domestic Resource Costs for Major Crops, Trend

Values 1991-92 ......................................... 13

Table 2.1 Distribution of Land Classes by Province ... ........... 19

Table 2.2 Distribution of Land ................................ 22

Table 2.3 Operations and Maintenance Expenditure and Recoveries.. 29

Table 2.4 Cross-country Comparison of Education Levels .......... 31

Table 3.1 Annual Average Overvaluation Due to Trade Policy

Distortions ........................................... 37

Table 3.2 Effect of Interventions on Agricultural Prices,

Nominal Rates of Protection ......................... 39

Table 3.3 Nominal and Effective Rates of Protection, 1991-92 .... 40

Table 3.4 Effective Rates of Protection andProducer Subsidy Equivalents, 1991-92 ................. 42

Table 3.3 Producer Subsidy Equivalents 1992-93 .................. 42

Table 3.6 Trade-Weighted All Inclusive Import Tax Rates .......... 43

Table 3.7 Sectoral, Economy-wide, and Total Nominal Rates of

Protection for Exportable Products, 1980-84 ............ 44

Table 3.8 Transfers from Agriculture due to Price Interventions,

1984-87 ................................................. 45

Table 4.1 Agricultural Institutions ............................ 57

Table 4.2 Public Expenditure in Agriculture (Current) ............ 63

Table 4.3 Public Expenditure in Agriculture (Development) ........ 64

Table 4.4 Provincial/Federal Expenditure and Subsidy ............. 65

3

Table 4.5 Food Subsidy and Expenditure ........................... 66

Table 4.6 O&M Expenditure and Requirements ....................... 70

Table 4.7 Consolidated Profit & Loss of Major Public Corporations

Agriculture Sector ..................................... 71

LIST OF BOXES

Box 1 Cotton and the Economy 1982-1993. 8

Box 2 Conceptual Framework for Productivity Growth

in Agriculture. 11

Box 3 The Emerging Horticulture Sector. 13

Box 4 Small Versus large Farms: Labor Supervision, Access toInputs, and Risk Aversion .24

Box 5 History of Land Reforms in Pakistan .26

Box 6 Structure of the Irrigation System .28

Box 7 The Importance of Transportation .33

Box 8 Marketing Arrangements for Major Crops .36

Box 9 The Exchange Rate and Agricultural Incentives .38

Box 10 Are the Estimates for Wheat Realistic .39

Box 11 Tractor Prices in Pakistan .41

Box 12 The Case for an Optimal Export Tax .47

Box 13 The Case for Government Intervention .54

Box 14 Land Management and Records System .56

Box 15 Comparing Extension in Pakistan and India .61

Box 16 Has the Fertilizer Subsidy Helped Fertilizer Consumption? 68

List of Charts and Figures

Chart 1.1 Production of Major Crops ................................ 9

Chart 2.1 Cultivated Area .......................................... 17

4

Chart 2.2 Irrigated Area by Method ................................. 17

Chart 3.1 Credit Disbursement ...................................... 49

Chart 4.1 Fertilizer Subsidy and Consumption in Pakistan ........... 68

5

Executive Summary

1. The Government of Pakistan has launched a major economic reform program.Although the agriculture sector has not been at the forefront of this reformprogram, it is an important vehicle towards the government's twin objectivesof adjustment and growth. Moreover, agriculture, which still accounts formore than 20 percent of GDP, and more than 50 percent of employment, has aspecial role in alleviating poverty, given the concentration of the poor inrural areas. As the primary user of the natural resource base, agriculturehas a central role in protecting the environment. The purpose of this report--prepared in close collaboration with Pakistani experts-- is to address themajor issues facing Pakistan's agriculture and to suggest a strategy toimprove the performance of the sector in the future. The report draws onprevious studies done in the Bank and in the country, particularly the Reportof the Task Force on Agriculture (1993) and the Report of the NationalCommission on Agriculture (1988).

2. Agriculture has generally performed well. It grew by more than 3percent a year for the past 30 years. Sources of growth, however, havechanged over the years -- from the seed, fertilizer, and irrigation package ofthe 1960s, to intensification of water and fertilizer use in the 1970s, toimprovements in crop management and incentives in the 1980s. In the lastdecade, cotton has been a key source of growth and livestock also did well.

3. Although the past growth has been impressive, a careful look at thesources of growth seems to suggest that the momentum may be running out. Thereis no chance of a significant increase in total cultivable land or inirrigation. At best, a 10 percent expansion in water resources can beexpected, and only at huge cost. Moreover, cotton has recently been adverselyaffected by viruses. With past sources weakening, future growth will depend onincreasing productivity. At present, there are clear signs of stagnation inproductivity growth. In fact, total factor productivity, a good measure ofoverall productivity, may even have declined since the mid-1970s, due tofactors such as resource degradation, failure to adapt technical change, andpoor incentives. Future growth through productivity increase, however, willrequire major changes in systems, policies and institutions for agriculture.And these changes are crucial because agriculture is still going to be key inachieving growth with poverty alleviation and environmental protection.

4. Overall productivity growth will come from such sources as allocation ofresources to crops in which the country has comparative advantage, improvementin technical efficiency of inputs of each major crop, and increases incropping intensity. On all counts, Pakistan potentially can do much betterthan it does now. Supply potential (as measured by domestic resource cost) insome crops is quite good, notably cotton, basmati rice and wheat. Irri riceand maize are also efficient users of resources, at least in certainenvironments. Sugarcane, however, is a high-cost crop for Pakistan andoverall productivity will increase with movement away from sugarcane. Thelivestock sector also has good potential. Pakistan's climate is ideallysuited to certain types of horticulture, as rapid growth in this sector shows.Key inputs such as water and fertilizer are not currently used in the mostefficient manner, and constraints in water management are holding down

i

cropping intensity. If the demand side is taken into account, diets willdiversify away from wheat as per-capita incomes rise. The domestic textilesector will remain the most important source of demand for cotton, and thissector is facing big competitive challenges from other countries.

5. Thus, growth in the future will have to be achieved through increasedproductivity, through more efficient use of limited resources. There is aparticular need to ensure efficiency and sustainability in the management ofthe most important natural resources i.e. water and arable land.

6. There are good prospects for increasing technical efficiency for allmajor crops. For most, particularly wheat, rice and maize, there aresignificant yield gaps of 30-50 percent (that is, the gap between average andbest farmer yields) . At present, realization of the potential is impeded byconstraints inherent in the way resources are managed and developed, or arisefrom policy distortions. Easing constraints and correcting distortions wouldbe the way to sustain (and improve) the future growth, as well as generallyimprove agriculture's contribution to poverty alleviation and environmentalprotection.

Resource Constraints

7. Pakistan has almost 20 million hectares of arable land, of which about60 percent is rated as good or very good. Variations in land quality can betraced to soil erosion and irrigation related problems -- e.g. salinity andwaterlogging. Soil erosion has so far affected about 12 percent of the totalland (but a much smaller percentage of high quality arable land) . About 10percent of the top quality soils are affected by salinity, and about 2 millionhectares by waterlogging. Because of imperfections in land titling and weakextension services, soil resources are also being damaged by farmers. Insteadof being concerned about long-term sustainability, they try to maximizeprivate, short-run benefits by cultivating crops for which soil is not ideallysuited. Cultivation of rice on loamy soils is an example of how some farmersare only looking at short-run profitability and ignoring long-termsustainability.

8. Distribution of land resources and systems of land tenure are alsoconstraints on productivity growth. Evidence suggests that small farms aremore productive than large farms. But land distribution is concentrated, with25 percent of farmland in farms of 50 acres or more. At the other end ofscale are extremely small and ill-equipped peasant farms. Lack of access totubewells is particularly severe constraint for these farmers. So there isscope for improved productivity if land distribution was more even, but theexisting economic forces are allowing the present land concentration topersist, which is another way of saying that land markets are not workingfreely to allow land to those who would use it most efficiently.

9. Security (or rather insecurity) of tenure is another major problem. Manysmall-farm tenants can be evicted at any time, except during the croppingseason. This undoubtedly affects their attitude towards long-term investmentsand toward planting crops with long gestation periods. An equitable landdistribution and security of tenure have played an important role inagriculture's success in East Asia. However, the speedy land reforms in some

ii

of these countries that were carried out without full compensation to ownersmay not be an option for Pakistan at this time.

10. Pakistan's agriculture is almost wholly dependent on irrigation. Twobroad types of problems which have plagued irrigation are now at a crisislevel. The present system design means that water delivery is supply-, notdemand-driven. The system does not have enough reservoir capacity to smoothwater consumption. Drainage (which would be costly, given the lack ofelevation in the plains) is inadequate. Other serious problems have emergedfrom the way the system has been managed. The government has(inappropriately) retained a big role in irrigation. Delivery efficiency isextremely low, because operations and maintenance expenditure has been wellbelow required levels, and deferred maintenance has led to significantdeterioration in the system. One reason for the short-fall in operations andmaintenance expenditure is insufficient recovery of those costs, let alonecapital costs. Water charges are not related to actual use, which encouragesoveruse of water, a scarce resource.

11. The human resource base in Pakistan is comparatively weak. Educationlevels are poor, a significant constraint on productivity and efficiency.Status of women, who play a key role in agriculture in Pakistan, iscomparatively low. The transport network is inadequate, which inhibits timelydelivery of inputs and outputs.

Policy Distortion

12. Although direct intervention by the government in agricultural marketshas diminished in recent years, major interventions still remain -- subsidizedimports of wheat; duty and period restrictions on cotton exports; andprotection of sugarcane. Major indirect intervention arises from theprotection of domestic industry, which penalizes agriculture through theimpact on sectoral relative prices.

13. Nominal protection coefficients reveal a continuing policy bias againstcotton and wheat (-18 percent and -35 percent, respectively), while sugarcaneis highly protected (56 percent) . Subsidies on nontraded inputs (water,electricity, and credit) offset the discrimination to some extent, but notcompletely. Producer subsidy estimates thus show falling, but stillsignificant, transfers from cotton (21 percent of revenue) and wheat (30percent of revenue) producers. Moreover, inefficiencies and rent-seekingundermine the subsidy system. Subsidies which appear as a transfer toagriculture may have little beneficial impact on the sector.

14. In the past, the resource transfer from agriculture (including theeffect of policy distortions) was significant, averaging 15 percent ofagricultural gross product in the early 1970s, but now has come down(according to one estimate, about 6 percent of agricultural product in 1992-93). If we take account of the capital costs of the irrigation system andnon-payment of agricultural loans, the transfer would be close to zero. Atthe same time, the sector pays almost no direct taxes. Of course, sectors arenot the only way of analyzing the incidence of the tax burden. Nevertheless,taxation of income and wealth in agriculture should be on the same basis asincome from other sectors, particularly to avoid distortions in the taxsystem. The net transfer between sectors should be irrelevant to anyindividual tax liability. Both the recently introduced agriculture income tax

iii

and a strengthened wealth tax are expected to generate little revenue from thesector. The new taxes represent a modest beginning, and they will have to bestrengthened in the future.

15. Government policy is also affecting vital inputs. Imports of phosphatefertilizers are handled by the government inefficiently, which results insubstantial delays and high costs to farmers. Moreover, the development ofthe seed market is restricted by parastatals and deficiencies in the law.

16. Problems in the credit market are also serious, reflecting both policymistakes and inherent difficulties in rural finance -- correlated risks andliquidity needs, and high transactions costs. Some farmers cannot obtaincredit, while others can get as much credit as they want, despite lowrepayment rates. The commercial sector has almost completely withdrawn fromagricultural lending, leaving state institutions as the primaryintermediaries. Deposit mobilization is poor and most lending is funded bythe tax payer. Most farmers use the informal credit market, where interestrates (reflecting transactions costs) are sky-high.

17. All these policy distortions have significantly retarded agriculturalgrowth and poverty reduction. Estimated output losses from past pricingdistortions are huge. Estimated output losses for the mid 1980s range from 10percent a year for cotton (the main export earner) to 6 percent a year forwheat (an increasingly costly import item). These losses were tolerable in thepast because other favorable factors were maintaining growth at a reasonablelevel. But these favorable factors are now exhausted, raising the burden ofpolicy distortions. The impact of policy is particularly evident when setagainst other countries that pursued different policies. The East Asianpolicy mix of fewer biases against agriculture, plus rural investment,promoted agricultural growth and rural poverty reduction.

Public-Private Sector Institutions, and Public Expenditure

18. Roles of the public and private sectors are (in principle) easy todefine. The public sector should confine itself to public goods and marketfailures; all other areas should be left to the private sector. Thegovernment of Pakistan, however, has its role expanded way beyond thatboundary. Public agricultural institutions have proliferated. They areinefficient, and problems of overlapping roles and coordination are severe.Public presence in the fertilizer and seed industries crowds out privateinvestment in these important areas.

19. Meanwhile, research and extension are particularly weak; none of themany research institutions were able to cope with the recent problems thatafflicted the cotton crop. Farming systems research does not receivesufficient priority. Extension is, perhaps, the weakest link in the chain andseems to be a key factor in explaining stagnating productivity in most cropsother than cotton. Diffusion of new varieties is particularly poor inPakistan. Extension is too centralized and incapable of giving site-specificinformation to farmers. Field workers are poorly motivated and lack thenecessary skills and operational funding.

20. The level and composition of public expenditure on agriculture leavemuch to be desired. Spending is dominated by subsidies, which do not helpfarmers either because of rent-seeking and inefficiency (in credit and

iv

procurement), or because the subsidy is designed to help consumers at theexpense of producers (wheat subsidies). Subsidies and parastatals arecrowding out private sector investment in key markets such as fertilizer andseeds. Furthermore, subsidies (on electricity and water) are of limitedbenefit to farmers too -- witness the shortages and unreliability of deliveryin these key areas.

21. Most agricultural public enterprises in Pakistan are supposed to beself-financing through their operating revenue, but in reality they have addedto the budgetary pressure by shifting some of the financing burden to thegovernment. For instance, the Sind Agricultural Services Organization has hadlosses averaging Rs 48 million per year between 1990 and 1993--losses thatwere equivalent of 13 percent average annual sales during that period.Support to corporations like the Pakistan Agricultural Storage and ServicesCorporation and the Punjab Seed Corporation has also run into millions ofrupees.

Strategy for the Future

22. In formulating agricultural strategy for the 1990s, the underlying focusmust be to build on the momentum created by reforms initiated by thegovernment. In a time of worldwide change, measures are being adopted innumerous countries which emphasize the need for revisiting views on the roleof government in economic decision-making, without compromising itsresponsibility for ensuring the pre-conditions for growth, povertyalleviation, and protection of the environment. The program of necessaryreforms is summarized in a matrix at the end of the executive summary. Thismatrix is based on a first round of discussions, and will be modified andfurther specified on the basis of future dialogue with the government. It isintended that the measures listed in the matrix will form the basis for anaction plan, to be jointly prepared with the government, for implementation ofstructural reforms in agriculture.

23. It is vital that distortions in agricultural incentives be corrected. Atpresent, government involvement in output markets distorts market signals andhas a huge fiscal cost. Prices of outputs should therefore be market-determined, implying that all forms of price supports, controls, and subsidiesfor wheat imports will have to be phased out. This will reallocate resourcestowards areas where Pakistan's comparative advantage is strongest. Forexample, if the prices are market-determined, production of cotton and wheatwill go up, and production of sugarcane will go down, but overall productivitywill go up. Although it is better to stabilize producer income rather thanproducer prices, if the goal of intervention is to stabilize prices, there aremore cost-effective means of doing so, such as promoting private storage andfutures trading. There should be an operationally relevant study to determinewhat these alternative means should be in Pakistan. It is crucial to ensurethat there is an orderly transition to market determination of prices, whichmay require that the reform program be phased in over 4-5 years.

24. Trade policy should reflect Pakistan's comparative advantage, andprotection for industry (which hurts agriculture) should be removed. Asrecommended by the recent Bank report on trade and tariff policy, it would bedesirable for the government to aim at replacing the present distortionarysystem of high custom duties, exemptions, and non-tariff barriers, with a

v

uniform duty scheme. In agriculture, the government needs to consider removingtaxes on cotton exports, duties on sugar imports, and quantitativerestrictions on both. Any loss of revenue would be offset by removing thewheat subsidy. Along with removal of price and trade policy distortions,revenue from direct taxation of agricultural income and wealth shouldincrease. All income should be taxed in the same manner, regardless of source.

25. The functioning of input markets has to improve. The best option for thegovernment is to liberalize input markets, again with an orderly transition tofull private-sector orientation. The government should seriously considerprivatizing as soon as possible parastatals involved in seed, fertilizer, andother inputs. Directed credit in any form will have to end, and community-based banks could be considered as a solution to rural credit problems.

26. Reform of irrigation should be based on market pricing of water andinstitutional reform. The latter involves public utilities on the supply-sideand water user associations on the demand side, with the market policed by apowerful regulatory authority. Market pricing of water will helpsustainability, as well as the funding of operations and maintenanceexpenditures. User participation has worked in other countries, includingIndonesia, the Dominican Republic, and the Philippines (where half ofirrigated land is operated and maintained by self-financing communalirrigation associations).

27. Pilot schemes and experimentation will be useful in guiding theimplementation of irrigation sector reform. Water charges can be raisedimmediately. Preparatory work can begin on the institutions that will operatethe decentralized irrigation system. The current operations of irrigationinstitutions can be made transparent and restructured along commercial lines,to facilitate the transition to a market-based system. For more efficientwater use, successful on-farm water management programs (which involvesformation of user groups for improved water management), could be extendedthroughout the country.

28. A careful study of the costs and benefits of land reform could beinitiated. However, some immediate measures would be highly desirable tocorrect land market distortions, which include low machinery prices (ofteneffectively zero through the use of delinquent loans to buy machinery), andunequal access to credit (which makes it difficult for small farmers to expandtheir holdings) . The land title process can be modernized and streamlined,including the establishment of a system of permanent title deeds to land.Security of tenure shall be assured, without creating further disincentives torent-out land. Groups finding it hard to gain access to credit (small farmers,rural entrepreneurs) could be helped not by directed credit, but by community-based credit provision.

29. Some redirection of public expenditure is necessary. Public expenditureshould attach greater importance to funding crop and resource managementresearch. Spending on research and extension needs to be restructured so thatoperational funding can increase. Joint public-private research ventures canbe encouraged. Pressing research issues that are currently neglected requireattention. There has been a serious buildup of pest and disease problems dueto unchanging cropping patterns, such as rice-wheat or cotton-wheat. Researchneeds to focus on solutions to such problems. Publicly funded research needs

vi

to stress growth-enhancing public goods, natural resource management issues,and poverty reduction.

30. The extension service needs to be reduced in size and improved inquality. Additional public funding should be contingent on an improvement inorganizational capability. Decentralization of the extension service isessential. Generally recommended packages need to be replaced by site-specific information. Farmer participation in the functioning of extensionservices will be helpful, and group-based approaches to extension (includingmass media) would be highly desirable. The private sector and NGOs should beencouraged to increase their provision of extension services.

31. The quality of human capital and infrastructure needs to be improvedthrough investments in education, health, and roads. Electricity supply willimprove with energy sector liberalization.

32. To tackle poverty, it is important to ensure that agricultural growth isbroad-based and has appropriate spillover effects on non-farm growth.Subsidies on capital, which reduce the employment impact of growth, need to beeliminated. However, measures to improve land records and strengthen securityof tenure, as noted earlier, can also help reduce poverty. Participatorycommunity-based organizations offer great promise in poverty reduction.

33. Policy towards the environment will call for more efficient use ofnatural resources, while arresting the degradation of the resource base.Correct incentives in input and output markets and reforms in land titlingwill ensure better natural resource management, particularly with respect towater usage. Additional steps to promote natural resource management -- suchas encouragement of integrated pest management, improvements in regulatoryagencies, institutional strengthening, and redirection of public expenditureto emphasize natural resource management -- will be necessary, particularlywhere market failure is an issue.

Prioritizing Reforms

34. While this entire strategy will have to be implemented to ensuresustainability, some elements have higher priority. Topping the list shouldbe price and trade policy reform; irrigation sector reform; restructuring ofextension services; and removal of distortions in land markets.

Moving Towards Market

35. Those policy reforms that support the ongoing structural adjustmentshould be top of the list. This includes government withdrawal from activitiesbetter suited to the private sector. These actions will include withdrawingfrom agricultural crop markets (such as wheat, rice, cotton, and sugarcane)and leaving prices to be determined solely by market forces. Reforms alsoinclude the withdrawal of the government from input markets, such as importsof fertilizer. With liberalization and rationalization of trade and tariffs(to low, uniform rates), the incentive structure for agriculture compared toother sectors will improve. Agriculture must contribute a fair share togovernment revenue, which will also advance the goals of internal balance andstabilization.

vii

Addressing the Irrigation Crisis

36. To address the present crisis in irrigation management and ensureoptimal use of a scarce resource, market pricing of water must be allowed todirect the irrigation system. Pilot programs and experimentation will beuseful in phasing in market-based reforms. The government needs to givepriority to drawing up enforceable property rights to water. Administeringinstitutions will be restructured along commercial lines. The best option isto develop autonomous, commercially-oriented public utilities on a canalcommand basis, ensuring cost recovery of all current operations andmaintenance and future capital expenditures. Farmers themselves will becomeresponsible for operations and maintenance. Water user associations will bevital to the new market-oriented irrigation system.

Reforming Extension

37. Given the potential of increasing productivity of most crops by 30 - 50percent, effective diffusion of existing technology and of best farmingpractices is paramount. Reform of extension services will have to take placein four key areas (a)closer links with research institutions, (b) reducing thenumber of front-line extension workers and replacing them with fewer, bettertrained workers more responsive to the needs of farming systems and not justgiven crops, (c) opening consulting services by adaptive research institutesto better off farmers on a cost-sharing basis, and (d) extensive use of massmedia and other group approaches for basic messages about available technologyand better farming systems.

Correcting Distortions in Land Markets

38. A full-fledged land reform is difficult and can be considered after acomplete study of costs and benefits -- a study that needs to be initiatednow. Some important measures, however, need to be implemented now. Foremostis rectifying the insecurity of tenure of many farmers, especially tenants-at-will. Protection from arbitrary eviction, enshrined in the 1972 Land Reform,should be backed up by the legal system. Property rights can be reinforced byimproving and streamlining land registration and land records. Finally, itwould be highly desirable to eliminate artificial incentives to self-cultivate, such as low machinery prices and unequal access to credit.

viii

Pakistan: Suwmary and Timeframe for Structural and Policy Reforma inAgriculture

Policy Area Objectives Measures Proposed Timing ofInitiating Measures andExpected ImplementationPeriod

Incentives Market-determined *End subsidy on wheat imports 1994/95, short-runoutput prices

*Remove wheat and flour 1994/95, short-run

import restrictions

*End support prices forsugarcane ~~~1994/95, short-runsugarcane

End support prices for all 1995/96, medium- to long-other crops run

Initiate study into 1994/95, short- toalternative means of reducing medium-runvolatility of agriculturalprices

*Build up enabling 1995/96, medium-run

environment for private sectorentry into storage anddistribution

Regulation of processing 1995/96, medium-runindustry where necessary

Trade Policy *Complete trade reform 1995/96,short- to medium-reflecting runcomparative

advantage Remove export taxes, import 1994/95 short-run

duties, and quantitativerestrictions

Efficient and Extend income and wealth tax 1994/95 short-runEquitable tax system to agriculture ,

Eliminate all agricultural tax 1994/95, short-runexemptions

Update tax base in agriculture 1995/96, short- toby revaluation of Producer medium-runIndex Units

Installation of mechanism for 1995/96,short- to medium-periodic review of Producers runIndex Units

Note: For implementation periods, short-run implies 1 to 2 years, medium-run, 3 to 5years, and long-run, 5 to 7 years.

* Element of high priority

ix

Pakistan: Suary and Timeframe for Structural and Policy Reforms inAgriculture, continued

Policy Area Objective Measures Proposed Timing ofInitiating Measures andExpected ImplementationPeriod

Input Markets Private sector *Privatize urea production 1994/95, short-runproduction and and distributiondistribution of

commercial inputs Divest National Fertilizer 1994/9S, short- to

Corporation, including 19/5 hr-tplants operated by its medium-runsubsidiaries

*Expedite privatization of 1994/95, short-run

phosphate imports

*Level playing-field between

public and private sector in 1995/96, short-runinput markets

*Commercialize seed 1994/95, short-run

corporations

Privatize seed corporations 1995/96, medium-run

Strengthen seed 1994/95, medium-runcertification process

Market-determined Remove subsidy on 1995/96, medium-runinput prices electricity

Reform the irrigation *Decentralize irrigation 1995/96, medium-runsystem system based on water user

associations, publicutilities, and market inwater rights

*Raise irrigation charges 1994/95, short-run

Remove distortions inland markets and Initiate a study of land 1994/95, short-runreform credit reformprovision

*Improvement and 1995/96, medium-run

computerization of landrecords

Clamp down on delinquent 1994/95, short-run

loans and end cheap loansfor machinery purchase

End directed credit 19/5 hr-u



x

Pakistan: Sumary and Timeframe for Structural and Policy Reform. inAgriculture, continu-d

Policy Area Objective Measures Proposed Timing ofInitiating Measures andExpected ImplementationPeriod

Government Reorient public Reduce burden of administrative 1995/96, short-runExpenditure expenditure towards expenses on research and extension

changing needs ef budget and increase operationalagriculture and funding.enhance efficiency ofexpenditure program Introduce new research priorities 1994/95, short-run

including research on farmingsystems, growth-enhancing publicgoods, and the environment.

Induce private sector to undertake 1994/95, short- toprivately profitable research. medium-run

Introduce patent protection for 1994/95, short-runseed varieties

Reduce duplication and increase 1995/96, short- tocoordination of research medium-runinstitutions to minimize wastage ofresources

*Reduce the number of front-line

extension workers and replace them 1995/96, medium-runwith fewer, better trained workersmore responsive to the needs offarming systems and not just given

crops

*Open consulting services by 1994/95, medium-run

adaptive research institutes tobetter off farmers on a cost-sharing basis

*Extensive use of mass media and 1994/95, medium-run

other group approaches for basicmessages about available technologyand better farming systems.

Increase investment in education, 1995/96, medium- toincluding functional education of long-runfarmers



xi

Pakistan: Summary and Timeframe for Structural and Policy Reforms inAgriculture, continued

Policy Area Objectives Measures Proposed Timing ofInitiating Measures andExpected ImplementationPeriod

Poverty Rural Poverty Remove subsidies on 1994/95, short- toAlleviation capital medium-run

Targeting developmental 1995/96, medium-runexpenditure towards poorand marginal farmers

Develop participatory 1995/96, medium-runcommunity-basedorganizations

Ensure enforcement of 1995/96, short- totenancy protection medium-run

Environment Environmental Price water at economic 1995/96, short- toprotection and cost medium-runsustainable

development Phase out sapling subsidy 1994/95, short- toto encourage private medium-runsector participation insapling market

Regulate pesticide use 1994/95, short-run

Provide incentives to use 1995/96, medium-runIntegrated Pest Management

Create community 1995/96, medium- toinstitutions to manage long-runlocal resources and commonproperty

Encourage private and 1995/96, short- topublic sector investment medium-runin soil and waterconservation

Note: For implementation periods, short-run implies 1 to 2 years, medium-run, 3 to 5

years, and long-run, 5 to 7 years.

xii

Chapter 1. The Agriculture Sector: Role. Prospects, andObjectives

1.01 Pakistan is a relatively poor country of 117 million people with per-capita income of $440 in 1992-931. Economic growth, however, has been strong,at 6 percent per year over the past 30 years. Although the share ofagriculture in the economy has been slowly decreasing (and that of industryslowly increasing), it is still the backbone of the economy, employing morethan 50 percent of the labor force and earning (directly or indirectly) 70percent of export revenues.

1.02 Crops remain the most important subsector, but livestock now accountsfor almost 40 percent of agricultural GDP. Crop production is diversified,with more than half of cropped area devoted to cereals, one-fifth to cashcrops, and the remainder to fodder, pulses, vegetables, and fruits. Pakistanhas three broad agro-ecological zones for crops - the irrigated lowlands, therainfed lowlands, and the mountain areas. The irrigated plains are dominatedby cotton-wheat, rice-wheat, and sugarcane. Wheat is the basic crop in therainfed areas. Crop-livestock farming dominates most mountain regions.

1.03 So what precisely is the role and performance of agriculture in theeconomy? What are its prospects? What are the key objectives that will haveto be pursued?2

I: Role of the Aqriculture Sector

1.04 Agriculture has an important role in both overall economic growth and inthe ongoing adjustment program. It also is central to poverty reduction andnatural resource management.

Agriculture and Growth

1.05 The agriculture sector has a direct role in determining economic growthgiven its one-fifth share in the economy. It has an important indirect role,

1 Economic and agricultural statistics quoted in this reports were taken fromthe following sources:Economic Survey, Government of Pakistan, Finance Division. 1992-93 issue.Agricultural Crops: Long-Term Trends. Government of Pakistan, AgriculturalPrices Commission.

2 The report of a high-powered Agricultural Commission, set up by thegovernment of Pakistan in 1986, addressed these questions and outlined astrategy and targets for the future. Though a fews years old, this CommissionReport (1988) is a highly informative reference which this report has used,along with other recent studies on agriculture in Pakistan, such as theexcellent report of the Task Force on Agriculture (1993).

1

too. A cross-country study of Asian countries (where agriculture's share inthe economy remains significant) shows that for each one-percent increase inper-capita agricultural growth, there is a 1.5 percent increase in per-capitanonagricultural growth. -This is partly because agriculture requires inputsand output processing from the nonagricultural sector. Moreover, increasedincomes in agriculture are, in large part, spent on locally produced goods andservices, which have elastic demand and high employment content. Thus,agricultural expenditure patterns have a strong growth multiplier, and,through increased rural employment, can help reduce poverty. Linkages betweenthe agriculture and industry sectors are particularly important.

1.06 Agriculture has provided a significant boost to Pakistan's overallgrowth in recent years. In 1986-91, agricultural growth was 4.3 percentannually, giving per capita growth of just 1.3 percent annually. Given thefinding that per capita nonagricultural growth contributes 1.5 times percapita agricultural growth, this leads to 2 percent per capita growth in non-agriculture, or 5 percent overall growth in non-agriculture( the actual growthrate was 5.3 percent). In Pakistan, agriculture's contribution tomanufacturing is especially strong. Manufacturing value-added is highest inindustries that have linkages to agriculture: food, beverage, and tobacco (30percent in 1986), and textiles and clothing (21 percent). Value-added isgenerally smaller in other industrial subsectors. The indirect effect ofagricultural growth thus is an important influence on Pakistan's overallgrowth performance.

1.07 Agriculture's role is equally evident in the bad harvest years. In thesummer of 1992, rain and floods ruined an estimated 15 percent of the cottonand rice crops, along with 10 percent of the sugarcane crop, and even more ofsome smaller-scale food crops. This severely affected the economy. Growthfell to 3.2 percent, and exports tumbled. In the current year, cotton yieldshave been badly hit by a virus, and wheat yields have been depressed by anunusually warm and dry winter, which will likely depress 1993-94 growth.

Agriculture and Adjustment

1.08 Pakistan is in the midst of an adjustment program, and agriculture willbe crucial in correcting the internal and external imbalances that theadjustment program seeks to address. Growing macroeconomic imbalances andunderlying structural weaknesses (such as poor human capital and continuedpublic sector orientation) warranted the program. A narrow and inelastic taxbase, coupled with growing expenditure requirements, led to a deterioration inthe fiscal accounts. As explained in Chapters III and IV, the government'sactive role in the agriculture sector has been overly extended and ofteninappropriate -- particularly evident from marketing restrictions, subsidies,and parastatals. Reforms within agriculture are, therefore, crucial for thegrowth of the sector, and form an important part of the adjustment agenda. Inaddition, agriculture has an important contribution to make to the success ofthe overall adjustment program.

3 Mellor (1976).

2

1.09 Internal Balance: Initial progress in reducing the fiscal deficit in theearly 1990s was not sustained. In FY 92, fiscal performance slipped, andagriculture contributed to the slide. Compensation for crop damage andcollapsed credit co-operatives put additional upward pressure on expenditure.Even in other years, agriculture has contributed to the deficit problem.Subsidies remain a significant burden on the budget. In particular, wheatflour consumption is subsidized in Pakistan to try to improve the livingstandards of the poor(see Chapter IV for details). Hidden subsidies are alsoprovided through non-economic pricing for services such as energy andirrigation. With public expenditure on agriculture dominated by subsidies,spending can be difficult to control. Fiscal difficulties are compounded bylack of cost recovery of development expenditures.

1.10 On the revenue side, agriculture also has an important role to play. Taxrevenues in Pakistan have been stagnant for many years at about 13-14 percentof GDP, which compares unfavorably with 23 percent of GDP for the Middle Eastand North Africa, and 18 percent for Latin America and the Caribbean. Directtaxes as a share of revenue are just 18 percent in Pakistan, compared with 39percent in Sub-Saharan Africa, 56 percent in the Middle East and North Africa,37 percent in Asia, and 46 percent in Latin America. Pakistan's heavy relianceon indirect taxes is a source of major distortions, distortions which

4agricultural reform can help correct. Until recently, agricultural income oragricultural holdings were not taxed. Although the situation is now changing,tax reform is proceeding slowly. Potentially, agriculture can contribute moreto total revenue, and the need for new sources of revenue will increase withthe loss of tariff revenue as a result of ongoing trade reform.

1.11 External Balance: Agriculture also plays a key role in external balance.The current-account deficit has remained near 5 percent of GDP over the lastfive years and external indebtedness (as a ratio to GDP) has not improved.5

Strong export performance, including agricultural items, has, howeverstabilized the current-account deficit and improved the ability to service thedebt.

1.12 Export performance had been impressive before 1992-93, due to cottonmanufactures and leather products. Leather goods have emerged as an importantexport category, now accounting for 10 percent of exports. Agricultureincluding livestock is thus a key element of exports.

1.13 On the import side also, agriculture has a key role. Major importcategories in 1992-93 included wheat ($464 million), fertilizer ($250million), and edible oils ($580 million). Together, these imports have nearlytrebled since FY 88, and the rise in wheat imports is particularly pronounced.

4For example, in 1992-1993, 35 percent of tax revenues originated from tradetaxes. Such high trade taxes distort relative prices of tradables andnontradables and creates a bias against agriculture (see chapter 3).

5Pakistan: Country Economic Memorandum FY93. World Bank, South Asia Region.3

Agriculture. Poverty, and Inequality

1.14 Poverty is a multi-faceted problem which agriculture alone cannot solve.

Most poverty reduction will have to come from the non-farm sector. Since the

poor are often marginal farmers or landless, wage labor and the rural non-farmsector are their main sources of income. Agricultural income accounts for

just half the total income of poor households, but since the agriculturesector is an important influence on growth in the non-farm sector, the total

dependence of the poor on the agriculture sector is much larger. In addition,

agricultural income contributes to income inequality in Pakistan, via the

skewed distribution of land (see Chapter 2). By contrast, rural non-

agricultural income tends to mitigate inequality.6

1.15 Agriculture has a particularly strong interaction with rural poverty,

which is higher than urban poverty (table 1.1) . Among the rural poor,

agricultural laborers, marginal and small farmers, and rural artisans

constitute the majority. In addition, some urban poverty results from flightfrom rural areas, providing an additional incentive to ensure that agriculture

and poverty interactions are properly managed. Transitions in and out of

poverty are considerable, and anti-poverty policy cannot assume that the poor

are an unchanging group of people.

Table 1.1: Survey of Estimates of Head-count Poverty in Pakistan

Percent of Households

Study Poverty Line Year Rural Urban

Malik (1988) Rs per capita 1987-88 24 19

Ahmad and Allison (1990) Rs per capita 1987-88 20 16

Malik (1991) Calorie intake 1990-91 16 7

Source: Srinivasan (1993)

1.17 Cross country comparisons highlight the direct effect of agriculture on

poverty reduction. Like many developing countries, Pakistan has seen a

decline in absolute poverty - from 40 percent of the rural population in the

early 1960s to 25 percent by the mid 1980s . However, despite growth rates

that rivaled those of East Asian countries, Pakistan did not achievecommensurate poverty reductions (see Table 1.2). Note, for instance,Indonesia's faster rate of poverty rate reduction, achieved largely through

increased incomes in agriculture. Detailed analysis of the 1984-87 period

Alderman and Garcia (1993).

Malik (1988).

4

reveals that 60-70 percent of the decline in poverty in Indonesia between 1984and 1987 came from households primarily engaged in agriculture.'

1.18 Additionally, note that because Pakistan's population growth is so high,the number of poor has fallen by much less than in East Asia. While bothMalaysia and Pakistan more than halved the percentage of the population belowthe poverty line, the number of poor fell by 46 percent in Malaysia, but just19 percent in Pakistan.

1.19 One reason why agricultural growth in Pakistan did not reduce poverty asfast as in some East Asian countries is because after the green revolution,growth tended to be unduly labor-displacing. In fact, Pakistan's record inpoverty reduction could have been worse, if migration to the Middle East hadnot played a vital role in absorbing the vast pool of rural labor. We arguein Chapter III that some labor displacement was rooted in policy distortions,notably policies relating to machinery and credit. This undermined the directcontribution of agricultural growth to poverty reduction.

Table 1.2:Percentage of Population below the Poverty Line

Country Years First Year Last Year

India 1972-83 54 43

Indonesia 1972-82 58 17

Malaysia 1973-87 37 14

Pakistan 1962-84 54 23Source: The East Asian Miracle. The periods are chosen for purposes of comparability.

1.20 In addition to basic poverty-line measures, social indicators are alsoimportant. By various measures, Pakistan is deficient in this regard whetherwith respect to other South Asian countries (Table 1.3) or low-incomecountries in general. A survey (World Bank, 1993) of human capital investmentsin Pakistan over the last 25 years found that investment in education was verylow, controlling for per capita income. Improvements in education and healthindicators achieved through more efficient and larger investment in the socialsectors can alleviate poverty and improve the productivity of the rural laborforce at the same time.

1.21 Finally, there is an important interaction between agriculture andinequality. East Asian economies have benefited from the perception that thebenefits of growth are widely shared. Widespread viable land holdings sharethe benefits of agricultural growth, as does the generation of employmentopportunities in the sector. On the other hand, concentrated land holdings,or a subsidy regime that mainly benefits large farmers, results in a highlyskewed distribution of the benefits of growth. Chapters II and III argue thatthe latter situation (concentrated holdings and capture of subsidies by largefarmers) is present in Pakistan.

a Indonesia: Agricultural Transformation Challenges and Opportunities, WorldBank, 1992

5

Table 1.3: Social Indicators in Pakistan and India

Indicator Pakistan India

Access to Safe Water (% of pop) 50 75

Access to Sanitation (% of pop) 22 13

Male Literacy Rate 49 62

Female Literacy Rate 21 34

Primary Enrollment, Male 48 112

Primary Enrollment, Female 26 83

Source: Srinivasan (1993). Note: Water and Sanitation figures are for 1988-90.Literacy rates and enrollment rates are for 1990, and the latter are expressed as apercent of population in relevant age group that are attending school. This can exceed100 percent if persons from outside the age group are also attending primary school.

Agriculture and the Environment

1.22 Agriculture is at the center of many issues in natural resourcemanagement and environmental protection and, so, sustainable development. Muchof the increased agricultural output has been due to an expansion in land areacropped and liberal availability of water but future growth cannot rely onincreasing supplies of cultivable land and irrigation.

1.23 Part of the past output increases have come at the expense of depletionof the resource base. Management of water resources is poor, and damage tosoil resources considerable. As elaborated later (chapter 2), the landavailable for cropping is being reduced due to waterlogging and salinity,resulting from drainage problems associated with the expansion of irrigation.Soil erosion, aggravated by stripping and excessive grazing of naturalvegetation, is a serious problem. Reliance on a minimal number of croprotations had led to a decline in soil fertility. On' top of that, there havebeen recent problems with the use of chemicals without proper regulation.These problems are not an inevitable consequence of agricultural growth: adifferent pattern of agricultural growth would have done a better job ofpreserving the natural resource base.

II: Past Performance

Overall Trend

1.24 A longer view of agriculture sector performance and past sources ofgrowth can help identify what sources of growth can be tapped in the future.Table 1.4 divides growth in agricultural into three epochs (1960s, 1970s, and1980s-90s), with a different pattern of growth at each stage. The stronggrowth in the 1960s was driven by greater certainty in the use of irrigationwater due to an agreement with India, a productivity enhancing fertilizer-seedpackage, the emergence of tubewells and electrification of rural areas, andpolicy changes which increased the profitability of the agriculture sector.

6

Table 1.4 Growth in Agriculture, 1960-1993.

Period Average Annual Growth

1959-60 to 1969-70 4.9

1969-70 to 1979-80 2.3

1979-80 to 1987-88 3.6

1988-89 to 1993-94 3.6

Source: Economic Survey of Pakistans

1.25 The significant deceleration in the 1970s was due to uncertainty createdby land reforms in 1972 and 1977, and severe climatic shocks, a cotton virusthat depressed production for most of the 1970s, as well as politicaldisruptions. The pickup from the 1970s, and acceleration since 1988 can beattributed to the introduction of new varieties of cotton and improvements incotton management, as well as a gradual improvement in agriculturalincentives. The exceptional overall performance since 1988 (particularly therapid growth in 1989 and 1992) was also driven by cotton, but performance iscorrespondingly weakened when the cotton crop is poor (Box 1).

1.26 The dependence on cotton for good growth and export performance meansthat Pakistan is exposed when the cotton crop suffers any setbacks. Recentdevelopments indicate that the cotton-driven boom has now run its course andnew sources of growth will have to be found in the future.

1.27 Looking over the longer term, there are clear signs of stagnation. Thestagnation in the non-cotton crop sector can be seen in Figure 1.1, whichshows the production of major crops between FY 1983 and FY 1993.

9 Calculations to 1988 are taken from Gross National Product at ConstantFactor Cost of 1959-60, under the old National Accounts Methodology, from theStatistical Supplement to the Economic Survey of Pakistan, 1992-93. Growth isassumed to be continuously compounded. The change in the national incomemethodology after 1988 prevents us from presenting complete comparable figuressince 1979-80. For this reason, figures are presented through to and since1988.

7

Cottan anrd the Ecanomy: 1982-1993-

108*

0 ~ ~ ~ ~ ~ ~ *

I I p I~~~~~~~~~~~~~I1-100 -6 0 0 :5 100

C otton: Productio'n 'Grow th

T1a ~:QC:l c4 rx=ap in OvXe941 ggpwtbi

~Growth in cotton production has been a dominant factor in Pakistan:sincetho eairly: 1 9, 8OS. Cotton : produ:ction :nearly trebl:d bet: wtlen tht: yearl ofl cottondebacle (in 1994) and: 1992. Growth~ ini cotton wan. driven byuse o ultseed,~] increase ~in pesticideuean attractive incentives. Thle steaddepreciation of the rupee: And lower exp'ort duties'cnane h decline] inaverage Cotton export pticeo to ~abut~ 12 prent betwee iaL19ad19.

i0000~~~~~~~~~~~~ep ew 's "6I io4*q irid -. 00:02.,

This .compares. with. a: .29. percent, decline: (in $US.) in international: cottoniprices during this'period.

Area under cotton has risen frn around~ 2 2 mil4iox hectAres . hearly 1580S. to..close ,to 2Z.9 million ~hectares. in, recent years. Th:e builk Iofthe. production 4n-rease is: accounted for : by- 'the increase in yields - .cottOn(lint yields have rincrea :ed f -om just ovdr :3 KIsa0: in fic 18 to is: over75Q0 Kg/l in 1992. The vrieary oConsunw X GC cowtton is$ ;the) 6 estic textileindustry: in 1993, domestic mill consumption ~will be about 1.5 Million metric

on.

The importance of cotton in deteermining the ~overall growth rate can beseen in the figure, which shows GOP g qrowth. aaihOn rowth in cotton produc-ti6nsince 1982. There it a clearpocsitive relationship, highlighted in years whenthe cotton harvest is particularly bad orgood. Performance in 1994 will :beAdversely affected by the pook cotton harvest this year.

Part of the relationship between cotton growth and overall growth isdirect, as is apparent from growth-accounting c$alculations: between 1989 and1592, cotton production grrew at over 10) percent ~per year, contributing one-halt of the overall growthin the crop soctor: diring this 0-period, or one-4thirdof growth in agriculture.. However, there, ~are also impwortant~ downstreameffects. When cotton: isaabundant, thetextile industry bans the advantage ofcheap local sourcingr(avin: on transport costs). Conversel ywen cotton is

(lintY~~~~ yils aveinreaedtrom 03u t - bo;3t0id Y. Co 4nlfsat1Y.3t oVhen

scar e (as in 1994), the textilel inidustr aces aS At availabiltproblems, all the more so because icotton tends to be scarce in other countriesAt the same time.

8

Chart 1.-1 Production of Major Crops

1 1

10.5 xX-x-x x-x-X-XX x-X-x-x x x

10 -X

9.5

-~8.5

_ 87.5

6.57 + ,l6

75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93

-I--Rice - Maize -*-Cotton A Wheat -x- Sugarcane

Source: Statistical Survey of Pakistan.

1.28 Export Crops: The reasons for cotton's strong performance havealready been noted. Rice is both an export and a food crop. As the graph

shows, production of rice has barely increased since the early 1980s. Areaunder rice has fluctuated moderately at around 2 million hectares over the

last 10 years. Overall yield growth has been rather disappointing, decreasingfrom nearly 1750 Kg/Ha in 1982-83, to just under 1,550 Kg/Ha in 1991-92. The

sluggishness in yields has been fairly uniform across varieties. Basmatiyields declined 15 percent, coarse yields by 4 percent, and Irri-Pak yields by

4 percent.

1.29 Import/Food Crops: Wheat is by far the major food grain in the

country. It is estimated (Alderman and Garcia, 1993) that wheat contributesover 50 percent of the per capita daily availability of calories, and 85percent of the total protein intake. Some wheat is also used as feed grain.Total consumption in 1992-93 was about 17.5 million metric tons. Area under

wheat has slowly increased from 7.4 million hectares in 1982/83, to about 7.9

million hectares over the past three years. Wheat yields have shown some

increase, from just under 1700 Kg/Ha in 1982-83 to nearly 2,000 Kg/Ha in 1991-

92.

1.30 Like many other countries, Pakistan aims at self-sufficiency in sugarproduction which has increased slowly since the mid 1980s. Area under sugar

has averaged around 0.85 million hectares over the last 10 years. Sugarcaneyields have shown moderate increase, from nearly 35,700 Kg/Ha in 1982-83, toover 43,300 Kg/Ha in 1991-92.

9

1.31 Livestock and Other Products: The livestock sub-sector contributes

about 37 percent of agricultural GDP. This sector has recently grown quite

rapidly. Milk is the most important animal product, and 70 percent of this is

accounted for by buffalo milk. About 5.5 million households own livestock,

generally in small mixed farming systems. Between 70 and 80 percent ofhouseholds own less than four head of cattle or buffalo, 60 percent own fewerthan five sheep, and 80 percent own fewer than five goats.

1.32 Since 1981-82, red meat and milk production are reported to have grown

by 5-6 percent a year in volume. The poultry sector has been even more

dynamic, with production growing at 12-13 percent a year in recent years.

Poultry meat now accounts for 14 percent of total meat production, compared

with 9 percent in 1984/85, and 6 percent in 1980/81. At present, the

livestock sector is secondary to the crop sector and is not very

commercialized. There are many questions about the performance, potential,

and issues facing the subsector (such as a possible trade-off between growth

of crops and this subsector), and future analysis is needed before policies

can be recommended for the commercialization of the sector.

Assessment of Performance

1.33 Growth in the past has been quite good -- over a thirty-three yearperiod (1959-60 to 1992-93) average growth has been more than 3 percent. Theaggregate agricultural growth rate in Pakistan stands up well with some other

comparable countries, especially for the 1980s (see table 1.5). However, the

nature and sources of growth raises some concerns. First, Pakistan'sperformance is less impressive in terms of per-capita growth, or taking its

resource base into account. Pakistan's rapid population growth provides an

additional hurdle against which its agricultural growth must be assessed. East

Asian countries have maintained reasonably strong growth in agriculture with

much lower population growth, allowing stronger per capita growth.

Table 1.5: Growth of Agricultural Production and Population, 1970-91.

Country 1970-91

Agriculture Population

Pakistan 3.2 3.1

Indonesia 3.6 2.1

India 2.5 2.2

Egypt 2.6 2. 3

Source: World Development Report, 1993.

1.34 Pakistan is comparatively well endowed with natural resources (arable

land, water, and sunshine). With such an excellent resource base, and other

favorable factors (a sizable domestic market, and favorable location) one

could expect a faster rate of growth than was actually achieved. Second,

10

recent growth performance has become unduly dependent on cotton, which leavesPakistan vulnerable to setbacks in this crop.

1.35 Third, growth has generally come from extensive, not intensive,agriculture. Witness the low cropping intensity in Pakistan - 130 percent.Egypt, a country with strikingly similar resource endowments, has a croppingintensity of 180 percent. This is surely one reason why Pakistan with anagricultural land base six times as large as that of Egypt, its agriculturalproduct was less than twice that of Egypt. Egypt does have the benefit ofgreater water resources than Pakistan - which highlights the need for Pakistanto use its water resources as efficiently as possible.

Productivity Growth

1.36 Production estimates by themselves do not provide sufficient informationto evaluate the growth process. One needs to consider output growth relativeto inputs -- that is, productivity and productivity growth. Following theframework in Box 2, Pakistan is now in the second post-Green revolution phaseof productivity growth, where improvements in input efficiency are the mainsource of growth. High yielding varieties have already diffused widely, andinput use is at high levels.

Productivity- Growth in Agricflture.- Byrl.e (19944 divides -tecnicl ge inland intensive sy0tems into four stages, distinguished by-the soiirces: of growthin each phase: l)- theO pre-Green Revolutiobn phase, iwhen growth is driven by(irrigated) area .ea i6n:, and: produet.ivity growth is modest,. (2) the GreenRevolutio pha . whe: groth is dret y bilh yelding rt:ie withincreaed respnsineas t :inuts, (3) the- first post-Green on phase,when growth is driven by inten-sifi cation of -'put use, especially chemicalfertilizer, and jrrigatipn wa4ter twhich 1i2ilitates multiple cropping}, and(4>the second post-Gre e revoluti0n: ph se t . wh input use begins -to plateau , andthe source of growth bi ¢Xe6 dincreases in inut ef ieny, p d.with theongoing release:of new varieties-

The Greeni t evutioa--*i-fts tbi nt4 tion u$ er i -d raisesthe marginal responsiveness to inputs. Farmers do not operate;initially on theproduction frontier. In jthe first poet-Geen- Revolution phase, use ofcomplementary inputs rises,- and farmers improve allocative efficiency(equalizing mar4i.al products idftpK icesY). It the sacaonodpoetG4Xee Revolution

phase, farmers encounter :diminiahing - returns toinputs, and. move towards theproduction frontier -by- eaiuing their. ffitincy. :Reio rcdegradation is aform of technical regress:which will shift the.productio.n function downward.

1.37 However, neither absolute levels of productivity, nor productivitygrowth over time, have matched the progress on the input side. Various methods

11

can be used to measure productivity, but most measures yield similarconclusions. Partial productivity measures (output per unit of a singlefactor) suggest that productivity is much lower than comparable countries suchas India and Egypt. Growth in yields has been very disappointing. Forinstance, in the past decade, wheat yields have grown at 1.6 percent perannum, compared to 2.9 percent in the Indian Punjab and 2.7 percent for alldeveloping countries. In both levels and rates of growth, Pakistan's cottonperformance remains strong, however.

1.38 Yield gaps between Pakistan's farmers are also significant. Whileinterpretation of yield gaps is difficult, careful controlling for otherfactors still suggests a yield gap of 30 percent in the case of wheat, forinstance. For rice, measures show that up to half the potential yield remainedto be exploited for average farmersl0. In general, farmers are constrained bylack of inputs when needed, insufficient water, and seed impurities.

1.39 Even when inputs are used, output is less than potential. For wheatbetween 1966 and 1986, Byerlee and Siddiq found that yield growth was lessthan would be expected from the application of green revolution inputs. Theyattributed this to a decline in the quality of the resource base. Indicativeof resource degradation is the fact that the yield of high yielding varietiesin farmers' fields has not grown since 1970, despite the intensification offertilizer use.

1.40 The best measure of productivity is total factor productivity, whichcompares an index of all outputs with an index of all inputs. Two recentstudies" find that TFP has stagnated or even declined in post-greenrevolution Pakistan, i.e. the period since the mid 1970s. While the 4 percentoutput growth in agriculture over the past decade might seem impressive, theTFP statistics suggest that agriculture's true contribution to economic growthis less than this.

III: Pronp@CtS and Obj Ctiven

Supply Potential

1.41 The report of the National Commission on Agriculture (1988) hasextensively outlined the agricultural potential of Pakistan. Comparativeadvantage of different crops in Pakistan can be shown using the domesticresource cost, which is the ratio of domestic (nontraded) inputs to the valueof foreign exchange per unit of the crop, all evaluated at opportunity cost(see Table 1.6).

1.42 The resource cost ratio is less than 1, then there is a comparativeadvantage in resources being devoted to that particular crop. Cotton, wheat,and basmati rice in their respective farming systems are the crops which are

10 Yield gap data are taken from Byerlee (1994) and Saleemi (1994).

Cited in Byerlee (1994).

12

the most efficient for Pakistan. Coarse rice and sunflower are marginally

inefficient. Sugarcane and yellow maize are inefficient crops, at least for

the indicated regions. Livestock generally shows a comparative advantage. Therapidly emerging horticulture sector (without any policy assistance) has

extremely promising potential (Box 3).

Table 1.6: Domestic Resource Costs for Major Crops, Trend Values 1991-92

Crop, Location DRC Crop, Location DRC

Cotton, Multan 0.28 Wheat, Multan 0.68

Cotton, Nawabshah 0.26 Wheat, Gujranwala 0.29

Sunflowerseed, Multan 1.10 Wheat, Nawabshah 0.58

Yellow Maize, Faisalabad 1.29 Basmati, Gujranwala 0.49

Coarse Rice, Larkana 1.09 Sugarcane, Faisalabad 2.09

Broilers, Sind 0.23 Sugarcane, Larkana 1.91

|Broilers, Lahore 0.77 Sugarcane, Mardan

Source: Longmire and Debord, Table 16.

::. - .. . . : -. . . . . . . . .... :---..:

The Emerling Horticult -Sector: . - - .- - . - - - . - - .- . - -. , ., . . ,, -..... -. -

Pakistan climate gives it a. strong inrs±ve .dvatage n

horticulture. witriess the rapid igrowth in this sector -without iy policyinterveAtions. The .oCntys climatic zYoes' give it A UMbl* x2iohes inseveral areas of the horticulture market, notbly off-season (re:lative to thehuropean market) mid-winter harvests, year-round trqpic:a £fUitw 1 . J.-w chillingtemperate fruits (such.h as berries and some :types cf -tne fruit):, .and high*.illing temperate fnridt st ii!fu Xit)...:. ialready has lower production coatsd:.than Greece: Zndiat and.california.

However, the perishabil.ity of horticultural products ..complicates processingand marketing, and Pakistan is only beginning -to address Itsdeficiencies inthese areas. ThO Pakistan -fortiCulture .ip0xt teioenrit Pject .hasestimated that with an appropriate- plan :to correct ' 'these . deficiencies,Pakistan could have Mhrticulture erports of ver $1 .: year lwthin 10 years.

1.43 Pakistan's strong comparative advantage in wheat stands in starkcontrast to its need to import increasingly large quantities of wheat inrecent years. Sugarcane production appears glaringly uncompetitive, and the

gain in productivity from switching water use to a more efficient crop would

13

be considerable. Supply potential depends on moving production towardscompetitive crops, and on closing the large yield gaps described earlier.

Demand Prosnects

1.44 Future production will also depend on the growth of domestic and exportdemand. A relative decline in wheat consumption and an increase inconsumption of milk, meat, and fruits and vegetables can be expected. Cerealsaccounted for 18 percent of average consumer expenditure in 1969-70, and milk4 percent; in 1987-88, the shares were about equal, at 10 percent each12.High population growth will increase the demand for nearly all foodcommodities. A recent projection of livestock food consumption (Akmal 1994)for 1993-94 through to 2004-05 showed rapid growth in consumption of poultrymeat and eggs (over 5 percent a year in both cases), and somewhat slowergrowth in fresh milk (over 4 percent), mutton (nearly 4.5 percent), and beef(over 3 percent).

1.45 On the export side, the World Bank13 projected that Pakistan's effectiveexport market will grow by 3.5 percent per year over the next decade. This isonly slightly above the 1980-90 average, and is not as high as some ofPakistan's competitors. The World Bank projects that world cottonconsumption will grow by 1.9 percent annually between 1991 and 2005.15Consumption within Pakistan is projected to increase from 0.4 percent annualgrowth between 1970 and 1990 to 0.7 percent between 1991 and 2005. lDemandprospects for cotton-based manufactures are also conditional on internationaltrade agreements. The new GATT agreement calls for a ten-year phasing outperiod for the current Multi Fiber Agreement trade constraints on textiles andclothing. Most of the phase out is weighted towards the end of the 10 yearperiod, so no immediate improvement can be expected.

Sector Objectives for the Future

1.46 What, then, are the objectives the sector should be pursuing in thefuture? Developments in agriculture are going to interact with growth,poverty, and the environment, no matter what policy makers do. The aim ofpolicy makers should therefore be to manage these interactions to produce themost beneficial outcome possible for the economy. In the past, little

Government of Pakistan: Federal Bureau of Statistics.

13Pakistan: Country Economic Memorandum FY93

14 Cotton Production Prospects for the Decade to 2005, World Bank TechnicalPaper Number 251

l5 Actual data since 1991 have been consistent with this forecast.

16 This projected growth is markedly slower than the 2.2 percent for India andBrazil, and the 2.3 percent for China. Future domestic consumption may befaster than what is projected.

14

attention was given to the interactions between agriculture, poverty, and theenvironment. Priorities have now changed, as was recognized and extensivelydiscussed in the report of the National Commission on Agriculture (1988). Thegovernment has previously relied on agriculture to provide food security,emphasizing the key grain crops, rice and wheat. Further poverty reduction,and changing demand patterns, will require increased attention to other cropsand livestock. In addition, insufficient consideration has been given todeveloping sustainable systems of resource use.

1.47 Increased pressure on land will require a shift towards higher value,higher yielding crops, and higher cropping intensities. Growth in the pastrelied on expansion of cultivable land, but irrigation water has become acritical constraint on further land expansion (see Chapter 2). At best, anadditional 10 percent expansion in water resources is available, and the costof exploiting these resources is prohibitive. Fertilizer use is now levelingoff, suggesting that returns to further intensification are falling. Futuregrowth must rely almost entirely on efficiency gains, the potential which isconsiderable. Resource allocation should be guided by comparative advantage;in particular, the allocation of scarce water to sugarcane represents a majorefficiency loss.

1.48 There is a growing recognition that the neglected non-farm and non-cropsectors can play an important role in rural poverty alleviation. Althoughmany of the rural poor are landless, they have direct links with theagricultural sector through livestock ownership. Raising the productivity ofthe livestock sub-sector will be crucial to poverty reduction. Other ruralpoor farm small areas of often marginal land and could be helped by raisingthe production of small-scale marginal farmers.

1.49 The entire agriculture-environment-poverty nexus must receive greaterpriority than it has in the past. A depleted and polluted environmentadversely affects the poor through increasing health problems and lowering theproductivity of the natural resources off which they are often trying to live.On the other hand, the short time horizons imposed on the poor by theircircumstances limits their scope for natural resource investments. Lackingother means of diversifying risk (such as storage, migration, non-farmopportunities, or credit), the poor will tend to over-exploit any availablenatural resources. Agricultural strategy in Pakistan will have to harness"win-win" strategies to reverse the downward spiral of worsening of povertyand natural resource degradation.

1.50 Even though agriculture commands a declining share of GDP as an economydevelops, it plays a crucial role in preserving the environment. This isbecause it remains the primary user of the natural resource base - land andwater. Recognition of this fact perhaps makes the sector even more importantthan it has been in the past. In addition to natural resource managementconsiderations, the objective of sustainability will influence growthstrategies more generally. Future growth must come from yield increases, butin a less damaging way than in the past. This will require more emphasis onraising the knowledge of farmers about crop and resource management, and lessemphasis on chemical input use and mono-cropping.

1.51 In sum, agriculture will still be key to future GDP growth, povertyalleviation, and environmental protection. There are still good prospects that

15

the sector will be able to grow as strongly as before or even do better in thefuture. But this will not happen simply as a result of the continuation ofpast practices and policies. Realization of the prospects that exist willrequire effective actions and policy reforms to resolve the issues andproblems that now have become binding on the sector, particularly thoseimpeding productivity growth. The rest of the report will look at theseissues and constraints and suggests a future strategy for the agriculturesector.

16

Chapter II. Resource Constraints and Related Issues

2.01 How is the resource base affecting agriculture? Is the quality of

infrastructure - transportation, telecommunications, and power - constraining

the sector?

I: Land Resources and Constraints

2.02 This section describes Pakistan's soil resources and associated

problems. Charts 2.1 and 2.2 show the long-run trends in total, irrigated,

and rainfed area, and irrigated area by method (canal or tubewell).

Chart 2.1 Cultivated Area

2520

Million 1 5Hectares 1 0

56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92

Fiscal Year

ETotal Area Uirrigated ERainfed

Chart 2.2 Irrigated Area by method

201 5

Millio 1nHectares 1 0111

56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92

Fiscal Year

Mlrrigated CCanal ET.Wells

Source for Charts 2.1 and 2.2: Statistical Survey of Pakistan.

17

2.03 Total cultivated area has been more or less constant for quite sometime, and most expansion in irrigated area has with the reduction of rainfedarea. Since the early 1980s, irrigated area has hardly expanded at all. Thlisindicates that limits of both cultivable area and irrigated area have beenreached.

2.04 A comprehensive picture of Pakistan's soil resources is available fromthe Soil Survey of Pakistan. This has surveyed more than 700,000 squarekilometers of land - about 80 percent of Pakistan's total area. This includesa complete survey of Punjab, about 70 percent of Sindh, and over 80 percenteach of North West Frontier Province (NWFP) and Balochistan.

2.05 The survey integrated soil data with climatic, hydrological, plant, andother relevant information. Soils were then grouped into eight LandCapability Classes, according to their agricultural potential or relativesuitability for sustained agricultural use (Table 2.1). The first fourclasses are arable; the last four are forest or rangeland. Surveyed arableland accounted for 20 million hectares, and surveyed forestry and rangelandaccounted for 40 million hectares.

2.06 The rankings reflect limitations on the land, with class I having theleast limitations amongst the arable classes and class V having the leastlimitations amongst the forest/rangeland classes. The arable classes are alsodistinguished by whether they are irrigated or rain-fed. Information on thetype of limitation on the land is available. The rankings also reflect theresponsiveness to improved management - an important consideration if theyield gap with other countries is to be closed.

2.07 The Class I land has the highest potential for general agriculture.However, it is considered to be too highly permeable for rice. Class II landis slowly permeable, and prone to surface ponding. Some of the Class II soilis also clayey but has high potential if mechanically tilled. It also hashigh potential for rice. Other problems with this soil include sandiness andsalinity. Virtually all of the soil in Classes I and II is irrigated. Majorproblems with Class III land include salinity on irrigated land, and dry landthat is prone to flooding. Class IV land is poor or marginal. This soil issandy, or dry, or both, and has low potential for most crops. Crucially, theland capability surveys have found that negligible additional land isavailable for arable agriculture. The limits of land expansion appear to havebeen reached.

2.08 There are several constraints on land productivity in Pakistan. Theseinclude soil erosion, problems related to irrigation, and inappropriate use ofland.

Soil Erosion

2.09 Soil can be eroded by water or wind. Water erosion is most severe onhillsides, and along river banks. Some of the erosion is the result of long-term natural processes, but it has been accelerated by various abuses, notablythe depletion of natural vegetation and excessive tillage. Over 11 millionhectares are estimated (Mian and Mirza, 1993) to be affected by water erosion.Similar practices have also made wind erosion a serious problem, affecting

18

about 5 million hectares. Flood damage from soil erosion is gettingincreasingly serious. Natural vegetation, which forms the best waterabsorber, has been stripped. Meanwhile, man-made reservoirs, which shouldhelp control water flows, are impeded by sedimentation.

Table 2.1: Distribution of Land Classes by Province

Class Punjab Sindh NWFP Balo. Pakistan

M ha Percent of

arabl1eArable I-and (percent of each class by province) land

I. Very good 67 21 4 9 5.2 26

II. Good 53 33 8 6 7.0 35

III. Moderate 50 31 14 4 4.8 24

IV. Marginal 48 7 19 23 3.0 15

M ha Percent of

forest/Forest. and Rangeland (percent of each class by rangeprovince)

V. Moderate/Good 18 1 62 6 1.4 4

VI. Poor 30 6 17 43 15.4 38

VII. 18 12 13 46 23.2 58Unproductive

Note: 1. NWFP figures also include Tribal Areas2. Pakistan figures include Northern Areas.3. Source: Mian and Mirza (1993).

Irrigation-Related Problems

2.10 Salinity: Mian and Mirza (1993) suggest that waterlogging and salinitymay not be as serious as some scenarios had suggested. Certain types ofsalinity cannot be blamed on the canal irrigation system and result from themineral composition of the soil and the climate (salinity is more extensive inarid parts of the country). Canal irrigation did, however, exacerbate theproblem (known as secondary salinity), notably through seepage from the canalsystem and subsequent evaporation; rising water tables which draw up salinegroundwater; inadequate water to meet the leaching requirements of soils;insufficient attention to drainage around saline soils; and tubewells whichdraw on salty water.

19

2.11 Almost 8 percent of soil in Punjab and up to 15 percent of soil in Sindhis severely saline. While the problem does appear quite severe from theaggregate statistics, just less than 10 percent of the good quality class Iand class II soils are damaged. The cost of salinity in terms of lost yieldsis hard to evaluate. Yields have reportedly been reduced by about one-third inthe case of crops grown on slightly saline areas, and moderately affectedareas showed a two-thirds decrease compared with yields on normal land.Growth of any kind is difficult on highly saline soils.

2.12 Secondary salinity can be reversed with rehabilitation measures andreclamation measures have been undertaken. Application of gypsum to thesoils, additional water for leaching, and drainage (under SCARPs) have allbeen effective in reducing salinity. Of course, these measures should alwaysbe evaluated in terms of the benefits they provide in increased yieldsrelative to the cost of the measure. It is important to bear in mind that thereturns to addressing the problem of severe salinity on marginal lands arelikely to be quite low.

2.13 Waterlogging: Water-logging, too, argue Mian and Mirza (1993), maynot be as severe as commonly believed. As with salinity, inherentcharacteristics of the soil can be as important a determinant of thephenomenon as human activities such as irrigation. The major cause ofwaterlogging in the cultivated areas is excessive percolation from the canalsystem, which builds up the groundwater level. Also at fault are cultivationof water-intensive crops on permeable soils, obstruction of natural drainagechannels, and inefficient drainage.

2.14 The latest WAPDA figures indicates that, on average, about 2 millionhectares out of a surveyed area of 16 million hectares have a water tablewithin 1.5 meters from the surface - the Soil Survey's definition ofwaterlogged soil.

2.15 Other Problems: Soil nutrients have been depleted. Some get washedaway by irrigation water. Unchanged cropping patterns year after year are alsoto blame. Almost all soils are low in organic matter, and need continuousreplenishment through nitrogen and phosphate fertilizer. Soils are alsohighly deficient in zinc, and adequate zinc levels are particularly importantin rice production.

Inappropriate use of Soil Resources

2.16 Agricultural productivity in the future can be adversely affected ifsoil resources are not being used optimally. An example of short-termprofitability versus long-run sustainability is the general cultivation ofBasmati rice on the well-drained loamy soils of the northern Punjab, whichfrom a sustainability point of view are ideal for the cultivation of maize,sunflowers, groundnuts, and pulses. Basmati is grown because the climate ofthe region is uniquely well suited to Basmati rice, and more importantlyBasmati offers high returns. Water-intensive crops such as rice can contributeto rising water tables. Thus, there is an apparent conflict between short-runprofitability and long-run sustainability.

2.17 Another source of damage to soil resources is the standard fertilizerpackages recommended by research institutes - packages that do not take

20

account of local conditions. Infrastructure development has also played a role- the location of sugar mills has led to large scale cultivation of sugar onsandy soils, thus aggravating water logging problems.

2.18 Much land on steep slopes has been tilled, when it is really onlysuitable for forest. At the same time, arable land in the Indus plainsremains under irrigated forests. Efforts are made to bring marginal landunder arable use, at the expense of large tracts of arable land in the Indusbasin that remain under-utilized.

Forest and Rangeland

2.19 Pakistan has a very little forest area but has a vast area underrangelend or what is termed wasteland. This land is often in the vicinity ofvillages, and is often a common property resource. Problems typical to commonproperty resources are endemic, including soil erosion, overgrazing, anddestruction of trees and wildlife. Poor households are unusually dependent oncommon property resources. Large farms have begun to encroach on commonproperty, further marginalizing the poor. The degradation of this resourcehas been traced to ineffective management, and in turn, divergence betweenprivate and social costs is the root cause. Property rights to commonproperty are insecure, and sustainable community management has not beenencouraged. There is an institutional vacuum at the local level, astraditional institutions have broken down, and new community institutions havenot emerged.

II: Size. Tenure. and Productivity

2.20 The nature and distribution of owner and operator rights to land mayalso affect productivity. Analysis of the size distribution of farms reveals astrikingly uneven distribution of holding size (Table 2.2). Note that a farmconsists of all operated area, whether owned or rented in.

2.21 First, over one-quarter of farms are below 1 hectare in size, whichrepresents a huge number of marginal or near-marginal farms. Nearly half ofall farms are below 2 hectares in size, but occupy just 12 percent of totalland. At the other end of the spectrum, note that 7 percent of farms accountfor over 40 percent of total farm area. All the high-performing Asianeconomies with substantial agrarian sectors have widespread land holding,resulting either from traditional ownership patterns or land reform. Theevidence on how farm size and efficiency are related can be looked at from atleast two angles -- aggregate analysis, and farm-level evidence.

21

Table 2.2: Distribution of Land.

Size of Farm (ha) As e of number As % of Area

less than 0.5 13 1

0.5 to 1.0 14 3

1.0 to 2.0 21 8

2.0 to 3.0 16 10

3.0 to 5.0 17 17

Total Small Farms 81 39

5.0 to 10.0 (Medium) 12 21

10.0 to 20.0 5 16

20.0 to 60.0 2 14

larger than 60.0 0.3 10

Total Large Farms 7 40

Source: Census of Agriculture, 1990.

2.22 Aggregate Analysis: Quantifying the size-productivity relationship isdifficult. Many studies use output per acre as the measure of productivity,which ignores differences in use of inputs, cropping intensity, and landquality. The distinction between owned area and operated area is oftenignored. Nevertheless, a stylized fact from the empirical literature is thatthere is an inverse relationship between farm size and productivity. A recentsurvey17 cited a typical finding from Berry and Cline (1979). A comparison ofvalue added per cultivated area between small (between 5 and 10 ha) and large(over 20 ha) farmers in Punjab found that small farmers were over 2.5 times asproductive as large farmers. However, this was based on data from the late1960s, and the picture could have changed considerably since then. Somestudies did find a positive relationship between size and productivity(Byerlee, 1984, 1966; Akhter, 1986), but this can be explained by slowerdiffusion of green revolution technologies to small farmers. The largestcategory of farms (above 60 hectares) accounts for 0.3 percent of farms butoccupies over 10 percent of farm area; some of these farms represent absenteelandlords and underused lands and therefore give rise to a loss of efficiency.

1 Binswanger and Feder (1993).

22

2.23 Farm-Level Issues: Three important effects of size on productivitycan be distinguished. First, the small scale of operation and the use offamily labor gives small farmers an advantage in labor supervision over largerfarmers. Therefore, the quality of labor input is likely to be higher onsmall farms.

2.24 Second, however, small farmers can have restricted access to moderninputs, and this can hold back their productivity. Third, small and largefarmers may differ in their attitudes to risk and uncertainty -- large farmersmay be more willing and able to carry greater risks. The available farm-levelevidence on these three aspects of the relationship is summarized in Box 4.

2.25 The conclusion to be drawn is similar to that of Rosenzweig andBinswanger (1993): the lower supervision and labor costs arising from the useof family labor on small farms more than outweigh the scale, capital cost, andrisk diffusion advantages of bigger and wealthier farmers. Some resultssupport the view that the efficiency of the smallest, and presumably mostmarginal, farmers is low.

2.26 Tenancy: Tenancy plays an important role in explaining changes in thedistribution of holdings over time and can also have an important productivityimpact. The length of the time horizon for owners and tenants is bound todiffer, giving rise to differing attitudes towards long term investmento(especially natural resource management investments), and crops with longgestation lags. These effects are likely to be particularly strong forPakistan's sizable number of tenants-at-will, who can be evicted at any time,except during the cropping season. Insecurity of tenure also causes a biastowards the opening of new land (an unpriced resource), even if such land ismarginal land.

2.27 In addition, different types of tenancy may have different incentiveeffects. Since sharecroppers only receive half of the additional outputarising from extra effort, they may have weaker incentives than fixed-renttenants. In this regard, landlord supervision and cost-sharing can overcomethe disincentive effect of sharecropping. Nabi (1986) found that in a sampleof farmers in the Punjab, these mechanisms made sharecroppers as efficient asowner-cultivators. More generally, landlord supervision can overcome theproblem of differing time-horizons between owner and tenant. However,supervision is a costly activity, and may not be undertaken at all by absenteelandlords.

2.28 In practice, tenant farming has been in steep decline in Pakistan sincethe 1960s. Between 1960 and 1980 the number of tenants declined from 2million to 1.1 million. Mahmood (1994) finds that this decline continuedthrough the 1980s, with an increasing tendency to self-cultivate amongstlarger farmers. This indicates that larger farmers find it easier to self-cultivate than to rent to and supervise tenants. If this was the outcome of anon-distorted market, then this development would be desirable. But if policydistortions have promoted the tendency to self-cultivate, it is problematic.We argue in Chapter III that the extensive subsidy system has favored largefarmers over small farmers. Pakistan has also pursued a cheap tractorspolicy(relative to other countries), which has facilitated self-cultivation.The history of land reforms has likely put the possibility of further reformsin the minds of landlords, and made renting-out less profitable and more

23

risky. For all these reasons, the Pakistani land market has most likely notproduced the optimal distribution of owner and operator rights.

Box~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4~~~~~~~~~~~

Small Versu Larms, Labor SUPeCXrision. Access. .to !uts. ad T Sk.

The labor: supervision issue has a number of facets.-i Large farmers whochose to self-cultivate will have to hire' in labor. tThe 40inabil-ity. -tocompletelyy supervise this labor will give rise to a principal-agent- problem,more:severe than that faced by small f armers, who,rely more on family labor(Singh 1988). The alternative for large farmers isl-to rent out land to fixed_-rent tenants or sharecroppers. However, lMahmood. (l94) argues that the threatof land-to-the-tiller reforms has inhibited 'renting-out in Pakistan..Thi. scoupled with cheap machinery may have favored seIf-clAltivation a0and thebenefits of optimal labort supervision tare likely not being treaized.Landlords who rent out also face supervision problems),0 we discuss this :Singreater detail in the section on tenancy.

As for use of modern inputs, the, 1980 agricultural census showed xhoclear differences between small and large t farmers in adoption off fertilizerand plant protection (Mahmood etj al).] However,: tractor use and tubewell usewas much higher on medium and large farms intboth Punjab and Sindh in 1_980.While the productivity benefits of tractors are open tot dispute, it 0ei sgenerally recognized that access to tubewell water raises productivity., Theissue here is whether small farmers can rent in expensive lumpy inputs such-astractors and threshers.

Risk aversion could make small farmers less commercially oriented 0thanlarge farmers. However, Singh [1988] argues that small farmers are just llkelarge farmers in this ,respect. Both plant cashi crops, have a marketedsurplus, and hire labor, while tmarketed surplus does rise with ;farm si-ze:,0:even farmers with less than 10 hectares market over 25 percent-of crop output.tNot all of the sales of small farmers may be voluntary, however. ''A lack- ofstorage facilities may force some farmers to sell at the depressed: post-,harvest prices, only to have to buy back later at higher prices to meeIt foodneeds. The need to meet short-term debt obligations may have a simiklareffect. Large farmers are insulated from such distress 'sales through beingmore diversified, or having other assets to provide a buffer against shocks tospecific crops.

Evidence from other countries can help cast light on the effect of ri.skon investment decisions. For semi-arid India, Rosenzweigqand Binswanger fo-undthat it took very high levels of risk to reduce profitability of small farmersbelow that of large farmers.

24

2.29 Will Land Reform Help? To the extent that smaller farms are moreproductive than large farms by virtue of size alone, and if this relationshipcould be exploited by land redistribution, then clearly the size distributionof farms is inhibiting agricultural growth in Pakistan. Given this evidenceof likely impact of landholding size, and the adverse impact of the landmarket, the key question is whether land reform will help. Land reform islikely to be as problematic in the future as it was in the past (Box 5). Ill-conceived land and tenancy reforms have created a bias against renting-out,while poor implementation of the reforms resulted in little benefit to sittingtenants. An uncompensated land redistribution is politically infeasible, andeven if enacted, would be subject to widespread avoidance and evasion.However, a compensated land reform could be very expensive.

2.30 In addition, equity, as well as efficiency, issues arise in thedistribution of land. Successful land reform has played an important equityrole in East Asia. Comprehensive land reforms were undertaken in Korea andTaiwan, and the gains in productivity were considerable. 8However, these landreforms also advanced equity goals by establishing the principle of sharedgrowth and thus enhanced the legitimacy of governments and the popularownership of the growth process. Indonesia and Thailand had ownershippatterns with widespread land-holding to begin with. Even if the productivitydifference between different classes of farms is slight, it is important thatall farmers have equal access to land in a non-distorted land market.

2.31 Both the Korean and Taiwanese reforms were implemented by relativelyauthoritarian governments, partly seeking to neutralize the appeal ofcommunism. In Taiwan, the government seized land from landlords, who werecompensated with shares in state enterprises. Land was then sold to thetiller at favorable prices on favorable credit terms. Technologicalassistance was also provided to the new owners. The Korean reform had justnominal compensation, and land was distributed to 900,000 tenants, eliminatingtenancy at a stroke. On the other hand, the Philippines has a long record offailed land reforms, the most recent in the late 1980s. The reform law wasriddled with loopholes, and implementation and enforcement were very poor. Thereform was long-drawn out, and thus lacked the speed and finality ofsuccessful reforms elsewhere. The story with Pakistan's land reforms is verysimilar (see Box 5).

The East Asian Miracle, World Bank, 1993.

25

lj1sto;ry oQ loand RrQz i;n 124.a

Prior to 1959, land reforms were relatively conservative.. Reform:aimed

to provide security--of tenure and occupancy. rights to tenants,, rather thanredistributing land. In .41959 (under- martial :law), a ceiling of 500 irrigated

acres, or. I 0000 non.-irrigated acres,: was put on. land i holdings. Substantialexemptions were granted, however. Five thousand landowners ended up declaring

land in excess of 500 .acres,_ :fart below thet actual estimatte of the number ofsuch farms. About im hectaresof ..land was redistributed,, but, only about eight

percent of subsistence farmeri households kbenefited from the resale. Also, theresale was at well below the market price, and it is likely that many farmers

immediately sold the land back to the large farms.

The 1972 land reform was more far reaching. The ceilings on holdings

were reduced to 150 irrigated acres (G300non-irrigated), L or:to 15',000 Produce

Index Units, whichever. was greater. An additional exemption Eof 2 ,000.PIUs-was

allowed for mechanized farms, which .considerably diluted& the impact. of the

ceilings. Furthermore, the exemption for tractors encouraged massivemechanization of farms in the 1970s, displacing labor from agriculture..

No compensation was ptovided for Iexpropriated, , land, Land was

redistributed by the government according.tto various criteriat, but. vesting thispower in the state left the process open to influence from large landlords. A

market-based method of redistributing land would have been preferable. Both

reforms were undermined by the: granting of wide discretionary powers togovernment officials.

Tenants were granted security of tenure and freedom from eviction without

just cause. Enforcement of this provision was very lax and the¢ law: was easy to

circumvent. The law also regulated share-cropping contracts8and: there is someevidence that this provision helped sharecroppers by relieving then of the

obligation to perform non-farm labor for landlords.

Perhaps not surprisingly, the reform was a failure, and only about 0.5m

hectares was redistributed. Declarations of land holdings were very low, and

land holders used a variety of means to conceal the true size of their

holdings, with transfers to family members proving the biggest loophole in both

reforms. Since land was redistributed free, there was no guarantee that landwas going to the most productive farmers -- a market-based redistributivemechanisms would have helped identify productivity of farmers by their

willingness to pay for land.

In addition, neither reform did much help to landless laborers, with theredistributed land targeted at existing tenants. There is very little data on

how those who received land fared. As in other countries, the fact that

reform aimed to hand over land to sitting tenants may have led to tenants

being displaced by landlords fearful of future reform programs. Disputes overthe division of benefits from technical change also encouraged resumption of

rented-out land by landlords.

26

2.32 The potential costs and benefits of a future land reform should becarefully studied before further reforms are considered. But certain actionsare immediately called for. Distortions in the machinery and credit marketshave facilitated self-cultivation at the expense of tenants. These distortionsshould be removed. Tenants-at-will, who only have tenure for the length ofthe cropping season, need to be given greater security of tenure. Thesecurity of tenure provisions in the 1972 Land Reform law should be enforced.The operation of the land market could be improved by streamlining the land

title and registration process, and removing it from the potential for localabuse. Finally, unproductive land use could be penalized by aiming at fullcost recovery in service provision, with full pricing of irrigation water themost pressing need. A land tax could also be used to increase the efficiencyof land use.

III: Water Remourcee and Constraints

2.33 Pakistan's agriculture is almost completely irrigation-dependent.Irrigated land accounts for 76 percent of total agricultural land in Pakistan,against a corresponding figure of 25 percent for India and 35 percent forIndonesia. Irrigated land area increased from 12.5 million hectares in 1967-68 to nearly 17 million hectares in 1991-92, a rate of 1.5 percent per year.

There is limited scope for expanding Indus-irrigated land and at best, a 10percent expansion in water resources can be expected. Improvements in theefficiency of water use are therefore high priority.

2.34 Con_traints arising from system design A major problem with thesystem is the supply-based structure. The system was designed to distributewater with minimum human interference. There are few structures to regulate

canal flow, and the outlets run whenever a distributary or minor is running.No escapes are provided at the tail end of the system, and the surplus flows

have to be absorbed within the system. The surface and subsurface drainage

systems are inadequate. Construction of drains has not kept pace with

requirements, while infrastructure development has obstructed natural drainageflows.

2.35 A major constraint arises from the distribution of water over the year.River flows are highly seasonal. Roughly 85 percent of annual flows are inthe Kharif season (summer), and only 15 percent in the Rabi season (winter).Pakistan does not have enough reservoir capacity to allow it to smooth its

water consumption. Inadequate water availability at the beginning and end ofsummer, and in winter, is undoubtedly holding down cropping intensity. Thephysical capacity of the irrigation canal system is also a constraint duringperiods of exceptionally high demand for crop water. The system is operated onhistoric canal diversion patterns, which often bear little relationship tocurrent needs.

2.36 In response to problems with the canal system, groundwater use has risen

considerably, and has been a major factor in raising agricultural productionover the past 20 years. Groundwater use has been growing at 6 percent per

year. Tubewells not only provide additional water, but allow flexibility to

match surface water supplied with crop water requirements. Timeliness of water

27

delivery is therefore improved, and productivity rises. But opportunities forfurther exploiting groundwater are limited.

triucture of. the. Irriqxaton. aYste rT1 :00

The Indus Basin covers. 70%;. of the country's area, and0rms the jorwater source.j :, Total annual,iAnflows to the bas in. amount. tojust over 180 bn

cubic meters. IOf this, 130Ubn cub cm is diverted to canals; most of the restflows out to the sea. :The- flows to the:canals form the indus E13asinf Irrigation-System (IBIS) M addi tional SO bn cub cm is pumped from groundwater.

The IBIS. commands over 14m ha,4 and encorpasses the indus. ,ririver ,and,its'major ttributaries,1 three- major reserv-oirb, 19 barraes1, 12 lihnkcan5als.43canal commands, and over 107,000 watercourses. The, 4basin has; lat topograpy,poor natural drainage,0 0'POrouW .ols,5 and :a semi-arid climate.with 00high 0ievaporation.

River water is diverted by barrages: and weirs, into the main canals,. and-subsequently into branch canals, distributrlies, : and minors The flow to thef arm is delivered by the watercourses. The; .watercourses tare supplied :byoutlets (moghas): from the distributaries and minors. The system:was designedto cover a very wide area, and assumed dispersed farmers. The' imogha 1is:designed to allow a constant discharge that: self adjusts to variations inftheparent canal. Each watercourse is a miniature: irrigation system, coveringobetween 200 and 700 acres.

Farmers receive water proportional to theirt:] land holding.t water.. isrotated between farms on a weekly. basisi,, according to a fixed schedule(warabandi). Each watercourse command area is dividedinto 2S Sacre-squares.Each square has access to the publiic watercourse at ta::single pointt via a...,network of farm channels.

Each season, WAPDA makes an estimate, of water availability for the:-following season, based on consultation iwith the . Provincial tIrrigationt.Departments (PIDs). During the season, the PIDs communicate their water demand-to WAPDA at specific points in the system, at 10 day intervals and WAPDA thenreleases water from the reservoirs. The PIDs allocate and distribute waterwithin the canal command.

WAPDA is responsible for the operation and maintenance (o&M) of largemulti-purpose reservoirs and interprovincial link canals. O&M -of the-distribution system above the mogha is the responsibility of the. PIDs, and:farmers are responsible for O&M of watercourses and the field channels.' TheGovernment has full responsibility for providing drainage.

28

2.37 Constraints arisina from management; The canal irrigation system isplagued by low delivery efficiency. Due to age, overuse, and poormaintenance, the canal irrigation system is extremely inefficient. Averagedelivery efficiency is 35-40 percent from the canal head to the root zone,with most losses in watercourses. Most of these losses are not recovered viatubewells. Although Pakistan's efficiency appears high by internationalstandards (it is similar to India, and higher than that of Thailand), Pakistanreuses less water than other countries. This is because much of water seepsinto saline areas, and thus cannot be reused. The country is also moredependent on irrigation than other countries, so delivery efficiency is ofparticular importance to Pakistan.

2.38 The most pressing problem is inadequate operation and maintenance. Theirrigation and drainage system have been deteriorating, because of deferredmaintenance and utilization beyond design capacities. In the last four years,the shortfall between required and actual O&M expenditure has averaged 25percent (see Chapter IV).

2.39 On the revenue side, there is insufficient recovery of O&M expenditure,let alone capital costs. Charges are not linked to O&M needs; they areconsidered part of general provincial revenue. Recovery of operations andmaintenance expenditure (let alone capital cost) is completely insufficientTable 2.3) . Water charges are too low, and the scope for raising them can beseen in the willingness of farmers to pay much more for tubewell water, orindeed make direct payments to irrigation officials. Current charges are just5 percent of costs of production or farm income. With doubling of charges andfull cost recovery, charges would still be less than 10 percent of farmincome. The increased charge would amount to Rs 70 per acre-foot in thefield, versus a financial average return to water of Rs 700 per acre-foot, andprivate tubewell prices of Rs 100-400 per acre-foot.

Table 2.3: Operations and Maintenance Expenditure and Recoveries (Rs Million)

FY88 FY89 FY90 FY91 FY92

Expenditure 1704 1513 1617 1708 1985

Recovery 987 1028 962 1197 1105

Shortfall (%) -42 -32 -41 -30 -44

Source: World Bank (1993).

2.40 As noted earlier, waterlogging and salinity are by-products ofirrigation system. In other areas, falling, and not rising, water tables area problem. Due to explosive growth in groundwater use, there is danger ofexcessive lowering of water tables, and intrusion of saline water into freshwater aquifers.

2.41 The system is also characterized by inequitable distribution. Illegalpumping from canals is widespread, and in practice local water resources areoften controlled by a small number of politically powerful farmers. Water does

29

not reach users at the tail end of the system. The lack of water downstreamcreates an incentive to move upstream, where water is more plentiful.

The Rainfed Areas

2.42 The rain-fed area in Pakistan gets low average rainfall and is thusmostly arid. Total rainfed area has decreased as irrigated area has expanded.Rainfed area peaked at just under 8 million hectares in 1965-66. It thendeclined to around 4.5 million hectares in the early 1980s, before risingcloser to 5 million hectares in recent years. The main rainfed area stretchesfrom Northern Punjab into Southern NWFP. Wheat is the basic crop in theseareas; in the drier areas, rabi pulses (especially chick-peas) are alsoimportant. Livestock, too, is an important activity.

2.43 Yields in the rainfed areas naturally lag far behind the irrigatedareas. In terms of the land capability classes, no rainfed land is rated asvery good, and just 28,000 hectares is rated as good. About 1.3 millionhectares are rated as moderate, and 2.3 million hectares are rated as poor ormarginal. There are also about 1 million hectares that have moderate potentialfor flood-watered crops. The major constraints on the land are low anderratic moisture availability, shallowness and sandiness, and proneness toflooding.

2.44 Fertility conditions are also very poor. Soils in the rainfed areas arehighly deficient in fertilizer due to their low organic matter content. Mostsoils are also deficient in phosphorus. Experience of arid areas elsewhere(especially in Africa and India) shows that major technological packages arenot available for rainfed production. Although the relative size of aridagriculture is less in Pakistan than many other countries, more research isneeded for the development of technological packages in rainfed areas.

IY: Human Resources

2.45 The labor force was about 34 million in 1991-92, of which 25 million wasrural and 9 million urban. The size of the labor force corresponds to a lowparticipation rate of about 29 percent. This, in turn, can be traced toextremely low participation rates amongst women. Only about a quarter of theemployed labor force are wage earners, which partly reflects the extent towhich family labor is used in rural areas.

2.46 Pakistan's education levels compare unfavorably with other countries(Table 2.3). Spending levels are also very low. 19 Pakistan's primaryenrollment is well below the expected level based on its per capita income.There is a big gender difference too: the literacy rate was 45 percent for menand just 21 percent for women. This carries a cost, even if women do not work.

19 Education spending as a share of GDP was 2.3 percent in 1991-92, comparedto a UNESCO recommended level of 4 percent. Between 1970 and 1989, realexpenditure per primary pupil grew by 355 percent in Korea and 64 percent inMexico, while growing by just 13 percent between 1970 and 1985 in Pakistan.

30

In the East Asian experience, educated women who did not enter the labor forcewere able to educate their children at home, providing substantial indirectgains to the education of women. Such low basic literacy and numeracyseriously inhibit agricultural productivity, and complicate the task ofagricultural support services. Indeed, this could well be the bindingconstraint on rapid increases in agricultural productivity (and agriculturalproduction) over the next 10 to 15 years.

Table 2.4: Cross-country Comparison of Education Levels.

Country GNP $ per Literacy Net Primary Childrencapita, 1990 (%) enrollment reaching 4th

(%) year

Mexico 2680 81 88 81

Egypt 600 48 91 99

Indonesia 560 77 97 89

Pakistan 380 35 29 59

China 370 73 99 86

India 350 52 66 61

Source: UNESCO and UNICEF.

2.47 A study of total factor productivity in Pakistan20 over 1955-85 found animportant role for educational variables. Exploiting data from 35 districtsover 30 years, it was found that a 10 percent increase in rural male adultliteracy increased total factor productivity by 2.7 percent. This can becompared with the response to a 10 percent increase in the area underirrigation (2.4 percent), and a 10 percent increase in the share under high-yielding varieties (1.3 percent). Thus investments in education represent apotentially major source of productivity growth. The study also finds thatfalling investment in education can explain a significant part of thestagnation of productivity in agriculture after 1975.

2.48 Another survey (Ali and Byerlee, 1991) in Pakistan showed that manyfarmers did not display basic computational needs for optimal fertilizerusage. The deficiency was strongly correlated with the availability ofextension services. A study (Byerlee, 1994) in a rice-wheat area in Pakistanfound the major constraint in closing a large yield gap (40 percent) wasfarmers, technical knowledge and skills. There is a strong complementaritybetween education and the provision of extension services.

20 Rosegrant and Evenson (1993).

31

2.49 Quality of education is also a major concern. Butt (1984) found thatproductivity of farmers with secondary schooling (used as a proxy for quality)was significantly higher than productivity of farmers with just primaryeducation. Primary education increased productivity 7 percent, whilesecondary education increased productivity by 10.7 percent. Education alsotended to raise fertilizer use. Human capital improvements are particularlyimportant to the poor, whose major asset is their labor. Human capitaldeficiencies lower the return to labor and make it difficult for unskilledlaborers to move out of poverty.

V: Rural Infrastructure

2.50 Rural infrastructure covers many aspects. The most prominent is thetransportation network, which is central to the timely delivery of inputs andoutputs, as well as helping commercialization. Telecommunications have animportant role in commercialization, and in information dissemination.Finally, intensification of agriculture will involve increased energy usage,so constraints in this area must also be examined. Infrastructure deficienciesare an obvious obstacle to broad based growth.

2.51 Investment in rural areas has been another hallmark of the East Asiansuccess story. There has been a more even balance between rural and urbanpublic investment in roads, water, and sanitation facilities in Indonesia,Korea, and Thailand than in Pakistan. Rural investment provides importantpublic capital and raises the quality of human capital by facilitating healthimprovements.

2.52 The road volume in Pakistan is one of the lowest in the world. Fewerthan a third of Pakistan's 45,000 villages have access to wholesale tradingcenters through the network of all-weather roads. Certain areas remain cutoff from the rest of the country, not only during the rainy season and inwinter snows, but throughout the year. Overall rural road length increased by70 percent between 1981 and 1991, but remains inadequate. The road network isin poor condition - over half of the network is still unpaved, and more thantwo-thirds of the paved arterial roads do not have enough carriageway widthfor two lanes.

2.53 Despite the increased road density, Pakistan remains poorly suppliedwith road length relative to both area and population. For instance, theIndian Punjab has a road density of nearly 0.5 km per square km of area,whereas for the Pakistani Punjab, the figure is just 0.3 km. The country issimilarly lacking in other indicators. It should be noted that there is also55,000 km of good quality roads along canal banks. However, the public arenot permitted to use these roads. Upgrading these roads may be cheaper thanembarking on a new road-building program.

2.54 In 1988, the average distance between a village and various types ofinfrastructure was as follows: a railway station, 23 km; a metaled road, 8 km;a grain procurement center, 18 km; livestock and produce market, 20 km; apetrol/diesel pump, 17 km; and an extension field agent, 10 km. These

32

distances clearly require an adequate road network if agriculturalproductivity is not to be seriously hindered.

2.55 Evidence on rates of return to farm-to-market roads are available fromAsian Development Bank infrastructure project appraisals. Based on estimatesof transport cost savings and the cost of construction on maintenance, 57percent of projects (out of 447 proposals) have rates of return higher than 12percent -- the cutoff rate fixed by the government of Pakistan for approval ofprojects. Estimates showed that road upgrades could cut transport costs byover 30 percent.

. ~~~~~~~~~~~~~~~. . ..... .. _.. .B~~~~~2 ~ 4-~~~~~~~~~~~~~~~~~~~~~~~c ,. . e .. :

Th~~~~ Importul,~saneo raportation ~ .tts:~ .OWno:...,In a -carefu1 ttady--¢f ,Indian-.di ........................ ~i ge et 1 :~t~ -

-Goldstein,`-1993). found t. ta a 1 percent inct e in road pr' .ii i4 ould led1:to a 0.2 percent. increase inio.xput. Infrastructure provi'sn 'l. :'.important- indirect effects on output. C'C ercial banks s,were, mre.likely,'tloeat-e i ar¢*s -with d -: A:- 1 petcent -eiasprovisio '-roduce 'a O'.'B-.perce't-exp'niO'n''' in 'akn 'ser''- '' c.'- ''s'

,,- ','fPurtherore,,,,p'roved' i,nfrast rure t fnlonfarm incomc ge-'.'erating n opportunities., : which are: aruial to povert.yreductiort. The po rc2la ry, bef it.'tO the:: texnt that inf? ruatuSe lCetthe cost-of either tauriultural :puts '-or the g'' 'ds.- they' purha se'.-inf rastructure 'm,ay'] also .'e and .the- labor m'arket and' i'crese th deman-d dfounskilled labor. Good transport ean all.w the poor to take advantage ofi jbsmn urban areao:.. -oWever- it Is important t~o rtewbe -that any f. the

transportation ne ed of the poor will not be solved by imprements in roaalone. Sven rudimentary devices like wheelbarrows! carts, and bicycles wouldmake the life of-the poor much easier.

2.56 It is estimated (Qureshi, 1993b) that poorly maintained roads areraising transport costs by 30-40 percent. The distances between villages andmarketing and transportation points would not pose a problem if roads wereadequate, but the poor state of the rural road network severely inhibits thetimely transport of inputs and outputs. Such constraints are likely to beparticular important as Pakistan seeks to improve its position in the highvalue foods sector - such foods often being perishable fruits and vegetables.

2.57 Improved management of Pakistan's ports to facilitate timelytransportation of inputs and outputs is badly needed. The railway network isin seemingly inexorable decline. Maintenance has been neglected to such anextent that major rehabilitation investments are now necessary. The statebody, Pakistan Railways, is overstaffed and incurs heavy operational losses.Much needed reforms have been continually deferred.

33

2.58 Enrgy: Reliable energy supply is vital to the agriculture sector.

Tubewells only guarantee timely water delivery if their power source is

assured. The prompt harvesting of crops is particularly important in

Pakistan's double-cropping systems: the losses in wheat yields from late

harvesting of cotton are considerable. A more intensive livestock sector will

also require greater energy input for cooling milk and preservation of

perishable crops.

2.59 Pakistan faces significant energy supply constraints by comparison with

other developing countries. Converting all forms of energy consumption to

kilograms of oil equivalent, Pakistan's energy consumption per capita was 243

kg in 1991, below the South Asian average of 289 kg. Electricity power cuts

affect the rural sector particularly badly, for up to 10 hours per day.

Generators are in widespread use, but studies elsewhere (Lee and Anas (1989)

for example) have shown that the cost of power from generators exceeds that of

the electricity network by many times. Many problems can be traced to

Pakistan's energy sector policy, notably underpricing of energy and state

control, which has led to crowding- out of public and private capital

formation in the sector.

34

Chapter III. olicy nitortion.

This chapter has four sections. The first section looks at the effect onincentives of distortions arising from price and trade policy. The secondconsiders the resource transfer from agriculture and the related issue ofreform in agriculture taxation. The third section looks at the effect ofpolicy constraints on the functioning of input markets. The fourth and finalsection will sum up the effect of policy distortions on growth, poverty andthe environment.

l! The Incentiv- Effects of Price and Trade Policy

3.01 Few issues in Pakistan agriculture have received more attention than theimpact of price and trade policy. In the past, policy in Pakistan (as in othercountries) was dominated by the view that the interests of the people would bebest served by rapid industrialization and that agriculture would inevitablydecline. As it declined, however, it could provide a useful aid to industry.That was (and remains) entrenched in the underpricing of raw cotton to helpthe textile industry.

3.02 Food security is another important issue, and maintaining a low pricefor flour, while ensuring adequate production of wheat, has become a difficultbalancing act for the government. In the 1960s, the government thought that itcould handle commodity exports better than the private sector, so statecorporations were put in charge of cotton and rice exports. Another statedgoal was protecting the small farmer from the risk of exploitation by tradersand processors; for this reason, an elaborate procurement mechanism was put inplace for major crops.

3.03 Agricultural incentives can be distorted by sector-specific or economy-wide policies. Sector-specific policies would include distorted prices arisingfrom marketing or taxing by the government. Economy-wide policies includetrade and commercial policy that favors other sectors. A sectoral focus alonecan miss important linkages between these policies and incentives inagrirulture. Protection to other sectors relative to agriculture discriminatesagainst the sector and leads to a withdrawal of resources from agriculture infavor of protected sectors.

3.04 All major crops are covered by a guaranteed minimum price (GMP) orsupport price program. The setting of the GMP is, in theory at least, aconsultative process, that takes into account many factors including domesticand world demand and supply, costs of production, competing crops, andintersectoral considerations. The program is designed to combat price fallsimmediately following harvest, which could force farmers with limited storageto sell at depressed prices.

3.05 These agencies must purchase all quantities offered to them at the GMP,if the market price falls below the GMP (see box). Funding these purchases isoften problematic. Even if the GMP program has helped stabilize domesticprices for some commodities, such as wheat, it has associated costs.

35

Parastatals are not efficient in the handling of crops. They also colludewith processors or traders to share the gains from monopolistic marketpositions.

3.06 In Pakistan today, some distortions by direct marketing and taxingarrangements continue for some crops. Complicated regulations regarding theexport of cotton are in place. Anyone can export cotton, but it is subject toa (daily-adjusted) Minimum Export Price (MEP), below which no exports can takeplace. In addition, a benchmark MEP is set, and a variable export duty islevied on the gap between the MEP and the benchmark price. The variable dutyis imposed when international cotton prices rise (as they have recently) tokeep cotton in Pakistan. For wheat, the government maintains the farm pricebelow the trade price through subsidized imports.

Box8a

Marketing Arrangements for Major Crops

The Provincial Food Departments and the Pakistan Agricultural.Storage andServices Corporation (PASSCO) procure wheat at the support. price announced ibythe government. Procurement is voluntary and in recent years, tmore. and morewheat is being traded in the private sector. However, the government stillprocures between a quarter and a third of the'total crop. There is 5a -subs idyon imported wheat which is then released to flour rmills -at' the subsidizedprice.

An important feature of wheat policy is pan-territorial, pricing. Farmersanywhere in the country receive the same price for- their output. This saddlesthe government with all transport costs from surplus areas to markets. Thiscan also cause distortions in land use.

Private sector exporters are currently allowed 0to 0purchase cottondirectly from the open market. Both the state Cotton Export Corporation (CEC}and private exporters handle the export of cotton side by side. Private sectortraders dominate the domestic trade in cotton. An export dutyAis imp3sed toguarantee that raw cotton is available to textile mills.

The Rice Export Corporation of Pakistan (RECP) was established in 1974 tocarry out procurement, storage, quality control, and the export of rice.0 Riceprocurement has been voluntary since 1986.

Sugarcane owners sell their product to the mill at predetermined supportprices. The millers have a captive market for sugarcane. Sugar prices are kepthigh through import duties and other controls on imports.

36

3.07 In addition to these sector-specific policies, tariffs and quantitativerestrictions that protect other sectors affect the real exchange rate. Theexchange rate can be adjusted to take account of trade regime effect 21, andthe estimated influence of trade policy on the exchange rate in the past hasbeen large (Table 3.1). However, this is a partial approach that only takesaccount of one of the many influences on the real exchange rate (see Box 9).For the most recent calculation of incentives, we use the official exchangerate.

Table 3.1: Annual Average OvervaluationDue to Trade Policy Distortions

Overvaluation (%)

1972-75 25

1976-79 28

1980-83 25

1984-87 21

Source: Hamid, Nabi, and Nasim, Table 3.2

3.08 Once parity prices (world prices, possibly adjusted for transport coststo domestic locations) for crops have been calculated, indicators ofprotection and transfer out of agriculture are constructed in the usualfashion. The nominal rate of protection is the percentage by which theproducer price differs from the parity price. The transfer out of agricultureis the difference between the value-added in agriculture at actual prices andvalue-added at parity prices, adjusted for non-price transfers such assubsidies, investment, and taxation. The impact on output is calculated byapplying supply elasticities to the parity prices.

3.09 Let us first consider the effect of policy distortions on output prices(Table 3.2). We present the picture from 1960 through to the mid 1980s toprovide historical perspective, and we then provide estimates for 1991-92 and1992-93 to portray the current situation. Historically, agriculturalproducers have faced very large disincentives. The only major change in themid-1980s from the historical pattern was a significant fall in protection forsugar, and more modest falls in disprotection for cotton and irri rice. Largenominal disprotection persisted for wheat. More recently, note that there issome protection for coarse rice. Steep rates of disprotection remain forcotton and wheat, and sugarcane remains highly protected.

21 Dorosh and Valdes (1990) compare different methods of estimating the impacton the exchange rate of trade policy. Another approach to estimating thediscrimination against agriculture is adopted by Pursell and Gulati (1993) forthe Indian case. Since the discrimination arises from protection given toother sectors, this approach directly compares levels of protection inagriculture and industry. We use this approach later in this section.

37

:The ExchangeRt n Ah&cricu1ltuiaL~ Incentiv4s

sic mota u ur c !: he ve w

.since most jagricultu.ral'rommodities are. traded gqoods, the level of theexchange rate is an important d.deteerminantt of agricultural incentives. However,there are manyI. influences on the exchange rate. Distortions in the foreignexchange marketIcanh be ,accounted: ..for,by ..coinparingc,.the actual exchange ratie:with a purchaaing. power parityPPP rateI,Ior0 by :looking atI the black.. marketpremium. As Pakistan has -liberalizead itsforeign exchange market:0 on Ithecurrent Saccount, the divergence between:the actual exchange rate and the TPP'exchange rate (aorethemblackmarke.tpremium) has beome- less important einrecent yeats. H.eveit, convertibili.ty on the0 acapittI t0acon 's stillrestricted, anud some: divergence Ietween::.the PPP, ratepand the :actual rateremains.

Protectiont ft0r the ) -i t ided setor -n its own 0ausen real 0exchangerate appreciation and therefore: .results in lower: domestlc- prices, for tradedgoods. Most studies of agricultural incentives now attempt, tob use an exchange.rate that correct sfor:such distortions.. Hamid, INabi, 'and Nasim use a freetrade exchange rate,--which is -azough estimateo of the, xchange rate Ithat wouldprevail if tariffs,c quotast, :and, other,: restrictions were replaced by, a freetrade regime.

more recent studies- (Longmire and Debord , 1903: and Shabbir,t 1994) use anapproximation to update the etimates of overvalUation caused by: trade policydistortions. The formulai for:overvaluation is:

100o { (M + Tm) + (X -X1') }/ (M + x ) - 1]. where M --valu w of imports, TmImport Duties and customs -- Import-Rebateos x Value of Experts, and T Export Subsidies -Export Duties.

Longmire and Debord estimate thel overvaluation':at 145percent in 1990-:91V.we avoid the use of exchange :rate 4Scorrections6 in ou:r calculations for a numberof reasons. First, trade policyt is but one: influenice on ItheI exchange rate.,Other factors could lead to the rupee being undervalued, not overvalued. Mostnotably, if restrictions on capital 0flows were :relaxed, the rupee mayappreciate. Second, continued depreciation of the rupee:.-and ongoing tradereform have made trade 0reqim~e effects much lean important in recent years.

38

Table 3.2: Effect of Interventions on Agricultural PricesNominal Rates of Protection at Official and Free-Trade

Exchange Rates (percent).

Crop 1960-87 1984-87 1991-92 1992-93

Basmati -38 (-60) -59 (-69) -49 (-54) 17

Cotton -19 (-46) -14 (-36) -48 (-54) -18

Irri -29 (-51) -13 (-35) -22 (-30) 30

Sugarcane 24 (39) 10 (-18) 70 (53) 56

Wheat -10 (-42) -30 (-48) -31 (-39) -35

Sources: Hamid, Nabi, and Nasim (Table 5.2) for 1960-87 and 1984-87, Longmire andDebord (Table 9) for 1991-92, and Shabbir for 1992-93. Before 1992-93, we showprotection at both the official exchange rate and the free-trade adjusted exchangerate (in parentheses) . The 1992-93 data are at the official exchange rate. The 1991-92estimates are trend values and locations are Gujranwala (wheat and basmati), Multan(cotton), Faisalabad (sugarcane), and Larkana (irri).

- - - . .,, -, . . .E - E -. E - ~~~~. ... -. -. - . ...... . . . .,-. -.-

Are the Estimates for Wheat Realistia? -. .

Estimates of nominal protection ft rwheat P a hst 'aveiste't'lshown some of the highest levels: of. disprotectin ''i the rd. -w-Byerlee and Morris (1993) cauti,o,n that : these, estbiatesca' b somewhatmisleading. wheat is treated as 'an importable` and therefore the doimstti

price of wheat is comparedi-to .the import patypric,e.' . r

that the import parity price would prevail izna li eralized wheat- mrket'.'' Theimport parity price would prevail :in the. very- horttrun, :but,,-over tiiae, wheatsupply would rise in response to the higher prices (and onsumptin would

fall), and the domeStic: wheat' rkt wuld cl*ar -at -a price o wer than thimport parity price.: Pakistan t s high transport- osts- allow a substAntial

divergence between market-clearing and parity prices> - true measure ofnominal protection would compare the domestic price to the. market-clearing

price. Byerlee and Morris estimate- the- market cleatint ptiee f0r PAta:and using this estimate halves the estimate of nominal disprotection of Whet.

Taking Account of Innput Prices

3.10 It was recognized that depressing the price of agricultural output couldhave adverse incentive effects - hence, the extensive system of input

subsidies. As the green revolution progressed, the key inputs to modern

39

agriculture in Pakistan (irrigation water and fertilizer) became heavilysubsidized. Electricity and pesticides also were subsidized. In the early1980s, the pesticide subsidy was removed, and the fertilizer subsidy reduced.There is still a subsidy on diesel, electric tubewells, and the purchase ofseed. Subsidized credit schemes existed and, to the extent that agricultureuses imported inputs, it benefited from exchange rate overvaluation.

3.11 It is convenient to distinguish inputs by whether they are traded ornon-traded. Measures of nominal protection given can be adjusted to takeaccount of the fact that price and trade policy will make tradable inputscheaper, or more expensive, than their free-trade prices. The rate ofeffective protection takes this into account22. In the past, fertilizer,pesticides, and machinery have been subsidized, and a favorable price regimewas created for tractors (Box 11). Since Pakistan has now eliminatedsubsidies on most tradable inputs, rates of effective protection do not differthat much from rates of nominal protection (Table 3.3). The subsidy effect oftraded inputs is minuscule.

Table 3.3 Nominal and Effective Rates of Protection, 1991-92 Trend Values.

Item NRP (%) ERP (%)

Wheat, Gujranwala -39 -46

Basmati, Gujranwala -54 -59

Coarse Rice, Larkana -30 -39

Sugarcane, Faisalabad 53 70

Cotton, Multan -54 -63

Source: Longmire and Debord, Table 11. These measures were constructed using free-trade exchange rates.

2 It shows how the value added by Pakistani farmers at current farm prices(value of tradable outputs minus value of tradable inputs) compares with valueadded at social (parity) prices. The tradable inputs considered includedfertilizer, agro-chemicals, fuel, machinery, seed, and concentrate feed. In1991-92, both nitrogen and phosphate fertilizer were around 30 percent cheaperat domestic prices rather than world prices. Agro-chemicals were about 10percent more expensive, as was fuel.

40

Tractor Prices in Pakistan

A World Bank comparison-of tractor prices in 16 countries in 1985jfoundthat Pakistan and Turkey- had the cheapest tract4rs in- the world '(in USdollars, at exchange rates used in import transactiOnsO. -- The price perhorsepower was $136 in Pakistan, compared with ju'st over $175 in :India andBrazil, $200 in Indonesia and Sudan, $2-12 in Mexico and Egypt, and' over$300in Sub-Saharan Africa.

This is partly'because tractors imported into: the-'country as completelyknocked down kits'we-re subje¢t to relatively low import charges of lo percentand are assembled in a competitive industry, where the markup is official-lycontrolled. in addition, Pakistan: is able- to 0btin' the kits frominternational manufacturers, at very favorable prices.

However, unequal access to credit and mechanization subsidies could meanthat the benefits:of. low tractor prices are not widely distributed. -Interestrates on machinery remain -significantly lower: than parity interest rates: upto 40 percent lower-'in 1991. While Longmire- and Debord show that tractorprices have been slightly above world prices in recent years, it is possiblethat other countries: have decided :that even higher tractor prices weredesirable to avoid undue increases in farm size and,displacement of tenants.

3.11 It is also important to consider the role of non-traded inputs. Thecombined effects of output and input pricing can be summarized by the ProducerSubsidy Equivalent, which measures the subsidy to, or from, producersresulting from price policy (Tables 3.4 and 3.5). Important measured subsidieson non-traded inputs include the lack of complete recovery of operations andmaintenance costs on the irrigation system, and interest rate subsidies onloans. However, calculations do not take account of the capital costs of theirrigation system, or defaults on loans. Some argue that the capital costs ofthe irrigation system were recovered before the mid 1970s when revenue fromfarmers exceeded current costs. To highlight the role of non-traded inputs inPakistan, Table 3.4 presents producer subsidy equivalents and effective ratesof protection side by side. The most up-to-date estimates (Table 3.5) showthat levels of disprotection for export crops have fallen significantly inrecent years, and rice crops now appear somewhat protected. However sugarcanehas remained highly protected.

41

Table 3.4: Effective Rates of Protection and Producer Subsidy Equivalents

for Major Crops, 1991-92 Trend Values

Product ERP (%) PSE (I)

Wheat, Gujranwala -46 -15

Basmati, Gujranwala -59 -33

Irri, Larkana -39 -10

Cotton, Multan -63 -40

Sugarcane, Faisalabad 70 92

Source: Longmire and Debord, Table 16. These measures were constructed using the free

trade exchange rate. PSEs and ERPs are not directly comparable, but their respective

orders of magnitude can be used to indicate relative subsidy effects.

Table 3.5: Producer Subsidy Equivalents 1992-93 (percent).

Crop 1992-93

Wheat -30

Basmati 15

Irri 21

Cotton -21

Sugarcane 29

Source: Shabbir (1994)23. The estimates are at the official exchange rate.

3.12 Taking all transfers into account significantly reduces the extent of

price distortions. So much so, say Longmire and Debord, that discrimination

against agriculture was close to zero in 1991-92. This, however, ignores

serious intrasectoral distortions and intersectoral policy bias that adversely

affects the sector (see the next subsection).

3.13 While an aggregate estimate shows small discrimination against

agriculture, overall efficiency in resource allocation is impaired because of

significant disprotection for some crops and strong protection for others. All

estimates clearly show that there is significant disprotection for wheat and

cotton, whereas sugarcane is highly protected.

2 3Shabbir expresses the producer subsidy as a percentage of the value of

output at domestic prices. Others (e.g. Longmire and Debord) express the

subsidy as a percentage of the value of output at social prices.42

3.14 Moreover, even as price distortions have fallen, the efficacy of theprice and subsidy regime remains in question. Simply using subsidies ascompensation for reduced prices omits the crucial question of whether thesubsidies are going to the intended recipients. Considerable evidencesuggests that subsidies are not helping farmers, especially small farmers.While canal water may command a price one-fifth of its social price, whobenefits from this subsidy? While water is supposed to be equallydistributed, in practice, bribes and rent-seeking determine the distributionof water rights. Public procurement has similar abuses. Anecdotal evidencefrom the field suggests that procurement agents exploit farmers by absorbingmost of the difference between the market price and support price when supportprice regime is in effect. Panel studies (Haque, 1993) suggest that farmersare willing and able to undertake storage activities, reducing the need forpublic procurement. Input subsidies have also encountered major problems.

Sectoral Policy Bias: ComDarison with Other Countries

3.15 Given the similarity between agricultural policies in Pakistan andIndia, some useful insights can be gained from looking at the Indian incentivestructure. According to Gulati and Pursell (1993), at official exchange rates,Indian agriculture had an effective protection coefficient of 0.86 between1980-81 and 1987-87, while in 1986-87, manufacturing's effective protectioncoefficient was 1.34. Taking account of nontraded inputs reduced disprotectionof agriculture close to zero, but the protection of manufacturing relative toagriculture is indicative of substantial anti-agricultural bias, withconsequent misallocation of resources.

3.16 According to a recent Bank report (World Bank,1993), Pakistan also hasprotected industry relative to agriculture (Table 3.6), at least as revealedby import tax rates.

Table 3.6: Trade-Weighted Mean All-Inclusive Import Tax Rates, 1989-90

Whole Economy 70.1

Agriculture 46.7

Manufacturing 73.8

Consumer Goods 92.4

Intermediate Goods 71.8

Capital Goods 65.5

Source: World Bank (1993).

Of course, import tax rates are imposed rates and actual collection rates maybe less; in addition the protection may be latent, since some goods may stillbe produced cheaper domestically than at the world price. However, otherindicators also point to substantial protection for industry. The three mainindustrial sectors are chemicals, engineering, and textiles, which receiveaverage effective protection of 24 percent. In addition, the dispersion inrates of protection is huge. Within the three industrial subsectors, 70

43

percent of domestic resources are employed in inefficient and highly protectedindustries. This diversion of resources to inefficient sectors (and to rent-seeking) is at the expense of unprotected sectors, including agriculture.

3.17 Historically, over a broader range of countries, the effect ofprotection for other sectors on agriculture has been relatively high inPakistan. The effect of trade policy can be quantified by its impact onnominal rates of protection for agriculture (Table 3.7). For instance, in theearly 1980s, indirect policies had the effect of reducing the price receivedby Pakistani cotton farmers by 35 percent relative to the world price, whileproducers of Chile's primary export crop (grapes) were only losing 7 percentof the world price for the same reason. Even when measures of incentiveswithin agriculture show protection close to zero (consider the directprotection rates for Chile in 1980-84 and Pakistan now), protection given toother sectors can have a substantial effect on agricultural incentives. Mostof the difference in total nominal rates of protection for export cropsbetween Pakistan and Chile in 1980-84 arose from indirect policies.

Table 3.7.Direct, Indirect, and Total Nominal Rates of Protectionfor Exported Products, 1980-84 (percent).

Country Product Direct Indirect Total

Pakistan Cotton -7 -35 -42

Chile Grapes 0 -7 -7

Malaysia Rubber -18 -10 -28

Egypt Cotton -22 -14 -36

Source: Kreuger, Schiff, and Valdes (1988). Direct effects measure the percent bywhich producer prices diverged from prices at free trade (given the actual exchangerate and degree of industrial protection). Indirect effects take account of the impactof trade and macroeconomic policies on the real exchange rate, and the extent ofprotection afforded to non-agricultural commodities.

3.18 The high-performing East Asian economies had generally low levels ofdisprotection of agriculture. Both Korea and Malaysia had substantially lowerdisprotection of agriculture (in Korea, agriculture was protected). Thailand'sdisprotection of agriculture was similar to Pakistan's in the 1960s, butThailand subsequently reduced its disprotection while Pakistan was increasingthe burden on agriculture.

II: Resource Tranmfer and Tax Reform

3.19 There are significant resource flows out of agriculture, although muchof the transfer does not accrue to the government. The transfer due to outputpricing policy is the value of output at free trade prices, minus the value ofoutput at domestic prices. Subsidies through tradable input prices areevaluated in a similar manner. Agriculture also receives credit, water, and

44

electricity subsidies. As we noted earlier, two substantial transfers toagriculture are ignored: capital costs of the irrigation system, and defaultson loans. The total transfer out of agriculture was as large an 15 percent ofagricultural GDP in the early 1970s (Hamid, Nabi, and Nasim, p78). A morerecent estimate of transfers due to price interventions is shown in Table 3.8.The estimates are broken down by direct transfers and total transfers - thelatter accounts for the effect of the trade regime on the exchange rate.

Table 3.8: Transfers from Agriculture due to Output and Input PriceInterventions 1984-87 (Billion Rupees, 1985-86 prices)

Direct Transfers Total Transfers(Billion Rupees) (Billion Rupees)

Output Prices 9.8 21.0

Input Prices -2.6 -5.2

Net Transfers 7.2 15.8

Share of Agr GDP 6.4 13.6

Share of GDP (%) 1.6 3.4

Source: Hamid, Nabi, and Nasim, Table 7.2c.

3.20 In addition, nonprice transfers take place through (relatively small)land revenue and usher, and more significantly though public investment andsupport for research and extension. Hamid, Nabi, and Nasim estimate thattotal transfers into agriculture as a result of these factors averaged Rs 9.2billion (in 1985-86 prices) over 1984-87. This still leaves an overall nettransfer out of agriculture of over Rs 6.5 billion, or about 4 percent ofagricultural GDP.

3.21 The most recent estimate of the transfer from agriculture is in Shabbir(1994). This is for the five major crops, and includes the effect of outputpricing, assistance on fertilizer, credit, electricity, and canal water,public expenditure on research, extension, and education, and taxes. All thatresults in a transfer from agricultural producers of Rs 19 billion in 1992-93(Rs 10.9 billion in 1985-86 prices), or just over 6 percent of agriculturalgross product - much less than in previous years .2 4 The 6 percent figure isclose the consensus estimate of Haque (1993), who suggests a range of 5 to 8percent. One can question this, and argue that the actual figure would be evenlower if capital costs of irrigation, loan defaults, and subsidy on

24Shabbir also takes account of exchange rate overvaluation caused by tradepolicy distortions, which doubles the estimate of the transfer out ofagriculture. Given the difficulty of identifying the myriad differentinfluences on the exchange rate, we do not include this estimate here.

45

agricultural loans were included.25 One should note that the sectoralclassification is only one of many ways of looking at the tax burden, and itis more appropriate to assess tax burden by income levels. Nevertheless, theagriculture versus other sectors comparison plays a key role in the politicaleconomy of tax reform.

3.22 Keeping output prices lower than parity prices and offsetting this byinput subsidies is a particularly inefficient and inequitable way oftransferring resources from agriculture. The output-depressing effects of suchpolicy has been considerable, but looking at intersectoral transfers alonedoes not fully capture this. While free trade prices are used in thesecalculations, free trade quantities are not. Even if resources are directed toproductive use elsewhere, the price and trade regime create deadweight lossesthat are not reflected in intersectoral transfer calculations.

3.23 Much of the transfer out of agriculture does not accrue to thegovernment, but is dissipated as rents, notably in excess capacity in thetextiles and flour milling sectors. In turn, many of the subsidies toagriculture are also dissipated as rents, in the form of rent-seeking andinefficiency in procurement and input provision.

3.24 It is clear that revenue could be generated from agriculture in a farmore efficient fashion while advancing other worthwhile goals. Progressivedirect taxes could raise revenue in an efficient manner while facilitating thepursuit of equity. A land tax would raise revenue and increase the incentiveto use land as efficiently as possible. Outside the tax system, collectionrates for key services can also improve. In particular, the implicit subsidyin irrigation is of little benefit to most farmers, other than large farmersable to capture most of the subsidy; a better approach would be to raise watercharges and improve the quality of service.

Tax Reform Proposals

3.25 The interim government of 1993 introduced reforms in agricultural incomeand wealth tax which have largely been endorsed by the Task Force onAgriculture which submitted its report in February 1994. The income tax is inreality a presumptive tax based on the productive capability of land, assessedin terms of Produce Index Units (PIUs). The tax rate will be Rs 2 per PIUbetween 4,000 and 6,000 PIUs, and Rs 3 between 6,000 and 8,000 (with anexemption below 4,000, and a ceiling at 8,000 embodied in the land holdinglaws).

3.26 The rate is thus low: the maximum tax bill is Rs 10,000 - 2,000 PIUs atRs 2 and 2,000 PIUs at Rs 3 - or just $334. The 4,000 PIU exemption is high -anywhere from 75 to 150 acres, depending on location. This makes the tax baseextremely narrow and the revenue yield is extremely low (about Rs 50 million).Finally, the PIUs themselves are based on a decades-old assessment, and thetax base is now very outdated. It is also not clear that all loopholes have

2SThis finding of a small net transfer from agriculture is similar to Longmireand Debord's finding, cited earlier, that the net effect of policy onagricultural incentives is basically neutral.

46

.. ~~~~~~ ~~~~~~~~~~~~~~~~~~~~~ .. .... .. .....

The case for an optimal exL rt Lax-'

;; Whilej,in g'eneral, e ~port t es: re distortona .he theory for taxation of commodities,in:which a] country has: markwEt' pomei'.,Pakstan has market power in: beasmati rice1 ]'Ad' there wil be aop ' 1'-in both cases. Protection for o'ther sectors,, however,is alreadydimposi','a'implicit tariff e on the s * ani there' is-n o way of nwing w,hetan,the optimal tariffis geater,or -less -thanthis implicit' 'tariff. ...........

Moreover1 experience' has Mhown 'that' the dynamic: effectswof optimatariffs: can lead to'an erosion of market power as other competitors,emerge.This is already: 'idont: in Pakistat'- case f-r Ba ati rice,o wheret -ICalifornia have now entered the.market. While this cai p be xthe mishandling of Pakistani ..exports by' the Rice Export` Co rporatiodn itillustrates that the scope for an optimal tariff is -les,s,', than a', statipc'.analysis alone would suggest.

been closed. Once the PIU-based liability has been paid, agricultural incomecan still be used as a tax shelter. When land functions as a tax shelter,then land use is distorted.

3.27 The agriculture sector also has a wealth tax; agricultural land will bevalued at Rs 200 per PIU. There is a basic exemption of Rs 1 million, belowwhich no wealth tax is paid. There are also exemptions for a farm house,agricultural machinery, farm vehicles, and Rs 100,000 of agricultural land.Tax is then payable at a sliding scale of 0.59 to 2.5% in blocks of Rs400,000. Finally, agricultural wealth is not added to non-agricultural wealthfor the purposes of determining tax liability. Like income tax the wealth taxbase is narrow and outdated; exemptions are high, rates low; and, theprinciple of horizontal equity is again not satisfied.

3.28 Of paramount importance is the basic principle that all income should betaxed in the same manner, regardless of source. The size and direction ofresource flows between different sectors should not be relevant to anyindividual tax liability for a given amount of income. The revenue-generatingcapacity of agricultural taxation will increase with improved price policy andthe removal of distortions in input markets. Progressive direct taxes onincome and/or land would be desirable and such a system will have to replacethe current system of inefficient and inequitable resource transfer andcommodity-specific taxation, with the possible exception of taxes oncommodities in which Pakistan has market power (Box 12). Large farmers havevery low payment rates for services, which increases their gains from thecurrent subsidy regime. It is necessary to ensure economic pricing ofservices and full cost recovery; the latter will be facilitated by increasedincomes in agriculture.

47

III Policis Affecting Input Markets

3.29 Government policy is also creating constraints in input markets. Timelyavailability of fertilizer is essential, but phosphate, now being imported bythe public sector, is rarely delivered on time, and usually in insufficientquantities, with a resulting imbalance between nitrogen and phosphate use.While the recommended ratio is close to 1:1 for most crops, Pakistan's ratiois at best 3:1. Despite the fact that the import of phosphate is late yearafter year, the government seems unable to respond to the problem.

3.30 Fertilizer policy is imposing large hidden costs on farmers, includingsearch costs for scarce supplies, uncertainty about availability leading topanic buying, and depressed yields through lack of availability at therequired time.

3.31 The benefit of a liberalized input market is evident from theextraordinary growth in pesticide use after the subsidy was ended and entry tothe market was liberalized. Indeed, widespread pesticide use is widely heldto explain the dramatic growth in cotton yields in the 1980s.

3.32 Fertilizer pricing policy also has harmful intersectoral effects.Natural gas prices in Pakistan are held below world prices, with the objectiveof reducing the price paid by farmers for fertilizer. The problem is thatthis natural gas pricing policy reduces the availability of gas for use inindustry. In an energy-deficient country, this is an expensive way to offsetagricultural pricing policies.

3.33 Use of improved seed has been held back by problems of availability,accessibility, and quality. On-farm research (Byerlee, 1994) has estimatedthat use of old varieties of seed could be depressing yields by 15 percent.As in fertilizer, government policy is causing distortions. Private seedfirms have to compete with a large public sector producer which pricesuneconomically and runs losses. Private seed development is also held back bynon-existence of breeders' rights and lack of trademark protection.Enforcement of laws regarding seed quality is lax. Little work is now beingdone to develop seeds for fodder crops and high value food crops.

3.34 Factor markets have also been distorted by government policy. Apart fromtractors (see Box 11), the cost of agricultural mechanization has been furtherlowered by the access of large farmers to subsidized credit. Research fromPakistan and elsewhere has shown that mechanization has far greater labor-displacing than output enhancing effects. While some mechanization wasinevitable, policy induced lowering of tractor prices led to prematuretractorization and labor displacement.

Agricultural Credit

3.35 Research generally finds that higher credit use is correlated withhigher input use, and possibly higher output. Credit is also important inalleviating poverty, particularly in financing small-scale projects in therural non-farm sector. Export horticulture, a likely source of future growth,

48

requires substantial investment, both short-term and long-term, which thepresent financial system is incapable of meeting. Land mortgages are usuallynot sufficient to cover loan requirements, and the specialized nature ofhorticulture equipment lowers its value as security in the event of default.

3.36 Chart 3.1 shows the formal agricultural credit disbursed since 1981 bythe major lenders to the sector.

Chart 3.1: Credit Disbursement

20000

E15000.

a5000

1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993

Fiscal Year

--*-ADBP -U--Coops -U--Comm Banks -- C--Total

3.37 Lending by all agencies increased until 1987, and by the AgriculturalDevelopment Bank of Pakistan (ADBP) until 1990, followed by stagnation (Figure1). That, and the recent bailout of agricultural cooperatives by thegovernment are all signs of serious problems in the rural financial system inPakistan.

3.38 The formal credit sector lacks dynamism. Two institutions (ADBP, and theFederal Bank for Cooperatives) comprise virtually the entire formal sector.Both have failed to mobilize deposits, and rely on transfers to maintainlending capability. Deposits have been discouraged through interest rateceilings and the availability of more attractive government bonds. Cheapfinancing from the public sector gave the state banks no incentive to mobilizedeposits. Prudent financial regulation of the institutions is non-existent.

3.39 The 1972 Banking Reforms forced commercial banks to meet a target levelof lending to the agricultural sector. Quotas were also introduced forlending by size of farm, lending is at non-economic rates, for non-viableprojects and is directed to large farms, and the rural elite.

3.40 As in many developing countries, the credit system has poor enforcementand, so, many non-performing loans. Attempts at recovery can take years.Almost 30 percent of the US$100 million that farmers spend on tractors eachyear comes from loans that eventually turn bad. Thus distortions in thecredit market have wider effects: in this case, they lower the price oftractors and introduce a major distortion in factor prices.

49

3.41 Collateral requirements are a major obstacle to small farmers. Amongst

small farmers, credit access has increased much faster for owners and owners-cum-tenants than for tenant farmers. Nevertheless, there is a bizarre side tothe use of land as collateral - foreclosures are non-existent. This obviouslygives rise to perverse incentives.

3.42 The use of crops as collateral has been proposed as a partial solution

to the lack of credit. This, however, exposes the bank to much additional

risk - variability in the price of the crop and the chance that the purchasermight not pay for the crop. These risks are present for all crops, but are

particularly pronounced for horticultural export crops. Banks do not have the

technical capability to assess or price these risks, and would likelyconstrain lending even if the use of crops as collateral was permitted.

3.43 Informal finance, which is estimated to account for 70-80 percent ofagricultural credit, is geared towards meeting short-term credit and

consumption loans. This is partly because most of the informal sector is

illegal. Long-term investment and saving needs are correspondingly neglected.

Small farmers have generally not been drawn into the institutional credit

system - despite official statistics showing that most institutional creditgoes to small farmers.

3.44 Commission agents and merchants have remained an important source of

credit, even as formal credit provision has greatly expanded. They use tying

arrangements as a substitute for collateral, and interest rates are usually

much higher than in the formal system. Noninstitutional rates are far higher

than institutional rates. These higher rates partly reflect informal lenders'

high screening costs (Aleem, 1990). Since collateral is rarely available in

the informal market, lenders must undertake costly assessments of default risk

themselves.

3.45 Even in the informal market, rejection rates are high - over 50 percentin one village in Sind. While this would include some projects that would not

be viable at any positive rate of return, it also includes some projects that

with long gestation lags, or projects that would be viable from society's

perspective but not at the lender's required rate of return. In this samevillage, it emerged that the main cost to lenders arises from delinquent

payments, which the lender must pursue. Aleem also finds that market

equilibrium involves many lenders, each making a few small loans. Thus the

average cost per loan is high, reflected in informal sector interest rates.

3.46 Sources of the Problems: The rural finance system has been used as an

ineffective conduit for transfers to the poor. Rationed credit will inevitably

be captured by the better off. Credit was seen as a tool for offsetting the

transfers from agriculture resulting from output price policy. Some problems

with enforcement can be traced to inadequacies in the maintenance of property

rights.

3.47 There are structural issues that would hinder the credit market even if

policy distortions were nonexistent. First, financial intermediation isinherently difficult in predominantly agricultural rural areas. Clients aredispersed, so the cost of serving them is high. Pakistan's poor

infrastructure does not help. Likewise, the cost of gathering informationabout projects is high.

50

3.48 Second, and more important, farmer's incomes are subject to commoninfluences. Insurance is difficult and liquidity needs are highly seasonal -everyone needs liquidity at the same time, and everyone wants to save at thesame time. Aleem (1990) estimates that seasonal needs accounted for 50percent of the total demand for credit of a typical farmer in his sample.

3.49 Banks need to be allowed to intermediate in different markets. This

creates an inherent tension: while information resides at the local level,which would suggest the usefulness of specialized credit provision, such a

specialized provider would be unable to diversify to meet insurance andliquidity needs.

IV! Iffect of Dimtortions on Growth and Poverty

3.50 In the past, the effect of price distortions on output was great. Anexample of the size of the losses is provided by Hamid, Nabi, and Nasim, whoestimate how much (at an annual rate) actual output departed from output atfree trade prices between 1984 and 1987, taking account of allinterventions.2 Estimated short-run losses ranged from 6 percent per year forirri rice and wheat, to 10 percent per year for cotton, and 13 percent peryear for Basmati rice. Of course, these results are specific to 1984-87, andsome of the distortions have been mitigated or even eliminated since then. Inaddition, the output-enhancing effects of rents from the irrigation system and

essentially free credit (due to non-payment of loans by some farmers) have notbeen taken into account. Exchange rate distortions have also been reducedsince the mid 1980s.

3.51 Cross-country comparisons highlight the adverse impact of the resourcetransfer on growth. High-performing Asian economies typically had low levelsof direct and indirect "taxation" (including transfer to other sectors) ofagriculture. This suggests that the issue is not the direction of the transfer- some transfer of resources from agriculture is probably inevitable. Rather,the concern is that the size of the burden be reasonable, and the method ofextraction is as efficient as possible.

3.52 To the extent that policy has reduced agricultural growth, then it hasmade poverty alleviation more difficult. in addition, the distribution of theburden of the resource transfer within the agriculture sector may be

regressive. The most important instrument to ensure progressivity of theburden (direct taxes) has not been available to policy makers until recently.Pricing policy will have a greater impact on large farmers, since theirmarketable surplus is larger. However, large farmers are also more likely tocapture subsidies and have access to scarce inputs, making the incidence ofsubsidies highly unequal, so it is likely that the overall burden of thetransfer is regressive.

26 These figures represent gross deadweight losses. Net deadweight losses(which take account of the withdrawal of resources from other sectors) wouldbe less.

51

3.53 Wheat provides a particularly clear case where policy is failing toachieve its intended purpose, and is holding back growth at the same time.Subsidized imports of wheat depresx wheat prices to farmers, and thus wheatproduction. The government releases wheat to millers at subsidized prices(see Chapter IV). This is supposed to benefit consumers, but since the priceof flour is market determined, consumers pay close to market-clearing pricefor flour. Millers absorb the gain from subsidized wheat as rent, and notsurprisingly, there is excess capacity in the milling industry. Even if thesubsidy was being passed on to consumers, this represents a very ineffectivemeans of helping the poor. All consumers, rich and poor, benefit. A targetedsubsidy could help the poor at much lower cost to the government, and withless distortionary impact on the economy.

52

Chapter IV: Public Private-Sector Role. Inatitutions. and Public

Exndituram

4.01 There are specific desirable roles for the public and private sectors in

Pakistan's agriculture, and those roles have specific implications for public

institutions and expenditures. Not only will the current role of public

institutions in agriculture have to change, so too will the pattern of the

government's public expenditure program.

I: What Should be the Public and Private Sector Roles in Agriculture?

4.02 The public sector's aim should be to provide an enabling environment for

private-sector agriculture, while assisting in reducing rural poverty, andensuring sustainable resource use. This means that government will have to

confine itself to certain functions (see box) . In Pakistan, it also meansthat the large public-sector role in input marketing, storage, supplies, andsimilar functions will have to be cut, freeing resources for other neededareas.

4.03 Government intervention, even in the legitimate areas outlined in the

box below, has sometimes serious problems. While the government may appear to

be intervening to correct a market failure for the public good, in factintervention may be serving private interests, including those of the public

officials themselves. In areas where government has no legitimate role, thosebenefiting from intervention become a strong force to maintain theintervention.

4.04 The government of Pakistan has, in fact, defined "public goods" or"market failure" too loosely and, so, the public-sector role has risen

unnecessarily. The need is for "less, but better government". Less governmentwill reduce the crowding-out of the private sector in areas such as inputprovision, and marketing of crops. Better government will raise theproductivity of the agriculture sector through improved public capital andsupport services. Scaling down government's role will require agriculturalmarket liberalization and privatization. The continued presence of subsidies

and regulations propping up state enterprises slows market development and

impedes the transmission of prices to agricultural producers (throughcorruption in procurement and so on).

4.05 There are some areas where an increased role for government is

necessary, specifically to strengthen market institutions. There is a glaringneed to strengthen individual property rights to land, which allows the marketeconomy to function correctly. Another key issue of institutional developmentis to decentralize agricultural support services - such as rural credit, wateruser groups, and farmer groups.

53

The nature arnd extent of- overnment i ter enti n should b , dictated by

the nature and extent of market ::failures in the gricu:ture sector, i I udingpublic goods,: externalities 4.... moral hazard p infant industry

situations, and monopolies. In agriculture, the clearest:~ examples ~of: publicgoods are most, infrastructure certain types of drainage, eduction, ba

research, and public information. But even in these-cases , delivery of someservices can be contracted out by the government to the private sector, localuser groups, or NGOs.

Externalities present a strong cate for government- :intervention. butnot always government ownership. Government regulation or: :taxation, may. benecessary for optimal use of aI product or :service. ;,(Moral hazard problemsarise where thequallity of a''product 00Xcannot 0be explicitly evaluated, creatingincentives for producers -to 4supply subs0t£andard q00Xuantit.ies. Where; thesubstandard product can have harmful side-effects i(forexample,'pesticides),this problem is 0particularlyt serious. The government, .may alsoe need:- tointervene to regulate market:monopolies or oligopolies. .In 4h4s case, it mustbe stressed that public control of the industry is not always necessary.

Finally, the government may have: a role in infant industries, 0whereeconomies of scale or a high degree of uncertainty make the private1 sectorreluctant to enter. This argument only justifies- a temporary governmentpresence in the industry, before the firm is divested.to:the private jsector.

There are clear areas where the need for government i-intervention ispressing. Specifically, the public sector should have ,a key role in growth.enhancing public goods, mainly in agricultural research, extension, pub'lichealth, and rural infrastructure, while correcting environmental externalitiesassociated with forestry. soil conservation, and integrated pest management.Adequate funding is required to maintain basic infrastructure and to provide aframework for generating, developing, adapting, and disseminating appropriateproduction technologies. In addition, public spending should be directedtowards poverty reduction through better rural primary education, healthservices, and targeted agricultural development programs.

54

I: Inatitutions and services

4.06 Public institutions have proliferated and the provision of most majorservices has a significant public sector presence (see Table 4.1). Theiractivities range from input supply, infrastructure provision, regulation,resource mobilization, and output price intervention across federal andprovincial levels of government. Clearly, the public sector is over-represented in the agricultural sector. There is, moreover, considerableinefficiency and duplication of functions within the institutions.

4.07 Over-representation of the public sector is most pronounced in inputsupplies (see the last section in this chapter). Public-sector entities thatstill handle major inputs are inefficient, inflexible, and often unresponsiveto market conditions. Pakistan suffers shortages of phosphatic fertilizeryear after year. This is usually blamed on transport problems, but this begsthe question of why the planning of imports cannot begin sooner.

4.08 Public presence in input provision (such as seed and fertilizer) hasstifled the growth of the private sector, leaving the private producer starvedof inputs at the right time or in the right place. This contributes to theinefficiency of other agricultural support services. For example, if seed orfertilizer is not available when needed, the impact of public research andextension is diluted.

4.09 There are inefficiencies and weaknesses, too, even where the publicsector has a legitimate role. Duplication of functions or poor coordinationbetween agencies are serious problems. In cotton, for instance, there is amultiplicity of research institutes, with overlapping functions. The federalPakistan Central Cotton Committee is responsible for research on cotton, whilePunjab and Sindh have cotton institutes. The Atomic Energy Commission, also,breeds cotton varieties. Both Federal and provincial research institutes workon the same commodities or areas with little or no coordination. The sameproblem exists with adaptive research within the provincial research andextension departments. Inadequate coordination between WAPDA and PIDs duringplanning and implementation of irrigation and drainage projects makes the taskof future O&M difficult, as do on-farm water management programs (whichinvolve the formation of user groups for improved water management) betweenMinistries of Agriculture and Water.

4.10 Property rights in agriculture are monitored by land managementinstitutions, which operate at the provincial level. The land records systemis cumbersome and outdated (Box 14). A streamlining of the system is a pre-requisite for smooth functioning of the land market.

4.11 Environmental protection agencies at both the federal and provinciallevel, still suffer from lack of trained manpower and have taken a long timeto become functional. For example, the Pakistan Environmental ProtectionCouncil (PEPC) was created in 1983, but met for the first time in 1993.Provincial environmental protection agencies have been active for some time.

55

However, federal-provincial linkages are still very weak, and coordinationamong the provinces is non-existent. Regulatory bodies, such as seedcertification, quarantine, market monitoring, and essential commoditysurveillance, are spread too thinly, and are not performing at the optimallevel. There are failures too in credit provision (Chapter III).

4.12 Nowhere, however, is the weakness and inadequacies in public sectorinstitutions more evident than in the areas of research and extension. Infact, deficiencies in research and extension have been highlighted as a factorexplaining lower productivity growth in Pakistan compared to India. Rosegrantand Evenson (1993) have shown that there was a dramatic decline in totalfactor productivity growth in Pakistan after 1975 that did not occur in India.They argue that this can be explained by the level of investment in publicresearch, extension, and literacy, which increased in India after 1975, whilefalling in Pakistan.

Box 14

Land Managemte,nt and Records System

Provincial Boards of Revenue (BoR) handle property rights and titling: ofland, and also settle land disputes. At the f armer level, the systemt- isadministered by the Patwari. The chain between the BoR and the patwari hasbecome more complicated over time. The patwari maintains'12 different types of(hand-written) land records, prepares three daily reports, maintains at -leastsix different registers, and updates a "Red .Book", whi0h- in Princ:iplet 0is adetailed survey of every farm in the district. The-system vests considerablepower in the patwari, who must be approachd every time:a? piece o f Iland is: tobe transferred or some property right0 is required to be established.Unsurprisingly, the system has become exploitative over-time.

The use of computers could in principle facilitate record, mainteknaaaand ensure immediate access to land records and agricultural data, Howeve'r..institutional deficiencies must also be addressed. There are ;no. permanenttitle deeds establishing land ownership, which necesssitates frequent visits tothe patwari. The absence of deeds severely impedes the functioning of theland market.

The patwari is also a primary data source, but his multifarious roleopens data collection and gathering to various biases. More generally, cropproduction data are subject to ongoing trial and error revisions, which makesagricultural policy and strategy formulation very difficult.

56

Table 4. AGRICULTURAL INSTITUTIONS

Activity/commodity Fedenri bia NWFP Sindh BaI1bisannINPUTSLand BOR BOR BOR BOR BORFertilizer MINFAC(FID) PAD(PADSC) PAD(ADA) PAD(SASO) PAD

NFC/NFMLSeed MINFAC(FSCD) PAD(PSC) PAD(ADA) PAD(SSC) PADWater MOW&PIWAPDA PID PID PID PID

IRSAOn-Farm Mant MINFAC(FMWC) PAD(OFWMD) PAD(OFWMD) PAD(OFWMD) PAD(OFWMD)Minor Irrig MOW&P PID PID PID PID

Plant Protection MINFAC(PPPD) PAD(PADSC) PAD(ADA) PAD(SASO) PADCredit SBD/ADBP/FBC PCD(PCDS) PCD(PCDS) PCD(PCBS) PCD(PCBS)

SUPPORT SERVICESExt. & Ad. ResAgri. MINFAC PAD(PAED) PAD(PAED) PAD(PAED) PAD(PAED)Forestry MINFAC PFD PFD PFD PFDLivestock MINFAC PLD PLD PLD PLDFisheries MINFAC/MFD PFD(PFSD) PFD(PFSD) PFD(PFSD) PFD(PFSD)

Machinery MINFAC PAD(EngD) PAD(EngD) PAD(EngD) PAD(EngD)Soil Conservation MINFAC PAD(SCD) PAD(SCD) PAD(SCD) PAD(SCD)Research MINFAC(PARC) PARB/AUFI AUP/ SARO/SAU/ BARB/ARIs

(AZRI) ARIs ARIs ARIs(PCCC)/KARINA

Economic PARC (AERU)/PIDE PERI/AUF AEC(PU) SRPO/AERC(KU)Training

Agri/Livestock MINFAC ATIs/AUP/PBAC ATl/AUP ATIs/SAU ATlslBACForestry MINFAC(PFI) FSs/AUF FSs FSs FSsIrrigation MIRIS PIRI (IWARSI)/(DRIP)

Forestry MINFAC(PDI) FSS FSS FSS FSS

INFRASTRUCTUREElectricity MOW&P/WAPDAFTM Roads MLG&RD LG&RDD LG&RDD LG&RDD LG&RDD

REGULATIONfINTERVENTIONSPrices AFCOMFertilizer NFDCIrrigation WAPDA PID PID PID PIDSeed Certif. MINFAC(FSCD)Electricity WAPDAQuality control Food & Drug Insp.Environment PEPA EPA EPA EPA EPA

RESOURCE MOBILIZATIONLand Revenue Collector Collector Collector CollectorUshr Collector Collector Collector CollectorOctroi DC/LBs DC/LBs DC/LBs DC/LBsAbiana/Drainage Cess Collector Collector Collector

AG. STATISTICS MINIAC(EW) PAD(CRD) PAD(SDIR) PAD(SDIR) PADAG. CENSUS SD(PACO)OUTPUTSProcurement PASSCO/RECP PDF PDF PDF PDF

GOP/CEC

57

ABREVIATI1S PDF Provincial Department of FoodPEPA Pakistan Environmental Protection AgencyPERI Punjab Economic Research Institute

ADA Agricultural Development Authority PFD Provincial Forest Departnent

ADBP Agricultural Development Bank of Pakistan PFI Pakistan Forest Institute

AERC/KU Applied Economic Research Center, PFsD Provincial Fisheries DepartmentKarachi University PID Provincial Irrigation Department

AMRI Agricultural Machinery Research Institute PIRI Punjab Irrigation Research Institute

APCOM Agricultural Prices Commission PLD Provincial Livestock Department

ARI Agricultural Research Institute PPPD Pakistan Plant Protection Department

ATI Agricultural Training Institute PSC Punjab Seed Corporation

AUP Agricultural University of Peshawar RECP Rice Export Corporation of Pakistan

AUF Agricultural university of Faisalabad SARO Sindh Agricultural Research Organization

AZRI Arid Zone Research Institute SASO Sindh Agricultural Services Organization

BAC Balochistan Agricultural College SAU Sindh Agricultural University

BARB Balochistan Agricultural Research Board SBP State Bank of Pakistan

BOR Board of Revenue SCD Soil Conservation Directorate

CEC Corton Export Corporation SD Statistics Division

Collector District Collector SDIR Statistics Directorate

CRD Crop Reporting Directorate SRPO Sindh Regional Planning Organization

DC District Council SSC Sindh Seed Corporation

DRIP Drainage and Reclamation Institute of WAPDA Water and Power Development AuthorityPakistan

EngD Engineering DirectorateEPA Environmental Protection AgencyEW Economic WingFBC Federal Bank of CooperativesID Fertilizer Impon Department

FS Forest SchoolFSCD Federal Seed Certification DepartentFTM Farm-to-MarketFWMC Federal Water Management CellGCP Ghee Corporation of PakistanIRSA Indus River System AuthorityIWARSI Intemational Waterloggitig and Research

InstituteKARINA Karrakoram Agricultural Research Institute

for Northem AreasLG&RDD Local Government and Rurl Developmnent

DeparmentMFD Marine Fisheries DepartmentMINFAC Ministry of Food. Agriculture and

CooperativesMIRI Mona Irrigation and Reclamation InstituteMLG&RD Ministry of Local Govenmment and Rural

DevelopmentMOW&P Ministry of Water and PowerNCBs National Commercialized BanksNFC National Fertilizer CorporationNFDC National Fertilizer Development CenterNFML National Fertilizer Marketing LimitedOFWMD On-Farm Water Management DirectontePACO Pakistan Agricultural Census OrganizationPAD Provincial Agriculture DepartmentPAED Provincial Agricultural Extension

DepartmentPARB Punjab Agriculturl Research BoardPARC Pakistan Agricultural Research CouncilPADSC Punjab Agriculturl Development and

Supplies CorportionPASSCO Pakistan Agricultural Services and Storage

CorporationPBAC Punjab Barani Agricultural CollegePCCC Pakistan Central Corton CommitteePCD Provincial Cooperatives Pepartment

58

4.13 Research: There is a lack of funding in general for research andoperational funds in particular. The desired ratio of wage to operational(non-wage) expenditure in agricultural research is 60:4027, but the prevailingratios in many of the provinces are in excess of 80:20. The effectiveness ofresearch has declined because of lack of material support, effective planning,monitoring, and evaluation of the research program. Researchers lack theproper equipment and journals (notably foreign journals), leading to a verypoor research environment. There is an apparent proliferation of researchinstitutes (universities, PARC, NARC, provincial research and monocropinstitutes), often with overlapping responsibility and duplication ofresearch.

4.14 Agricultural research broadly covers two areas -- plant breedingresearch and crop and resource management research (CRMR). Plant breedingresearch has been satisfactory, although the lag in getting research tofarmers is often long. In wheat, the average age of varieties in farmers'fields is about 11 years, compared to an average of about 7 years fordeveloping countries. This reflects failures in extension and seed marketing.There is much duplication of effort in breeding research, and economies ofscale could be realized if some breeding institutes at different levels ofgovernment were consolidated.

4.15 Crop management research emphasizes increases in productivity throughresearch on such issues as timing and method of application of input ratherthan type of input, and land preparation and harvesting. Resource managementresearch focuses on preservation of the natural resource base. For most cropsin Pakistan, it is difficult to find good examples of CRMR which have beentranslated to farmers' fields. Much evidence indicates that key inputs suchas water and fertilizer are inefficiently used.

4.16 Existing research has failed to increase input efficiency. Outcomesfrom well-controlled experiments need to be better tailored to farmerconditions. Issuing technical packages to large, heterogeneous groups offarmers is not enough and they have to be given the means to adapt thesepackages to their own circumstances. Crop and resource management researchlacks a systems perspective. Specialized research tends to ignore interactionsbetween different crops and different agronomic issues. Key long-term issuessuch as necessary natural resource investments are generally ignored.

4.17 Extension: Most studies indicate that the variables that consistentlyexplain farmer inefficiency are farmers' knowledge and skills. As emphasisswitches from input intensification to input efficiency, improvements ininformation and skills play a bigger role in increasing productivity. Theinformation burden on farmers will only increase as agriculture becomes morecommercialized and sustainability issues come to the fore. Extension thus hasa central role in improving productivity growth, and also in making the

27 Pakistan Public Expenditure Review, World Bank, 1992.

59

distribution of benefits from knowledge more equitable. Small farmers mayfind it more expensive to acquire knowledge, and so public extension has a

role in equalizing access to new methods.

4.18 Extension programs take most of the current agricultural budget of the

government but their impact is negligible. An elaborate extension system is

already in place for crops, but many farmers question its usefulness. There

are now over 5,000 village extension workers (EWs) but they often lack

adequate training, which makes their task of improving the management skills

of farmers very difficult. Extension workers with higher education and

communication skills are often likely to be assigned administrativeresponsibilities. The system is characterized by a one-way transfer of

technology to a few inadequately selected farmers. Women and small farmers

are often ignored. Feedback from farmers is poor.

4.19 organizational problems are severe (especially lack of accountability),

and operational funding for extension workers is low. For instance, in NWFP,

the ratio of salaries to operational expenditure actually deteriorated from

75:25 in 1982-83, to 85:15 in 1992-93. In addition, the salary budget is

spread too thinly over extension staff. In contrast to the system for crops,

livestock forestry and fisheries are neglected. Nor is it fully integrated

with water management extension - a clear obstacle to improving the efficiency

of water use.

4.20 Much effort was made in the 1980s to implement the training and visit

extension system (T&V). It appears that the system had only modest success.Measures of farmer contact do show an increase, and extension services may

have had impact on increasing pesticide use. However, Husain et al (1994)

could find little evidence that T&V had improved the quality of extension

advice. An increase in extension contact does not necessarily show that

extension has aided growth. In addition, contact is highly skewed towards

large farmers. A survey of extension contact in Punjab in 198628 showed that

60 percent of farmers with over 10 hectares of land had contact with the

extension service in the previous year, whereas only 24 percent of farmers

with 5 to 10 hectares had contact over the same period.

4.21 Moreover, some success of the T&V system can be explained by the poor

state of extension service before the system was introduced. By introducing

organizational discipline and some additional funding, the T&V system was able

to improve the previously deficient extension services. It is less clear,

however, that the system was the most cost effective use of resources. It

involves a centralized, hierarchical approach that limits feedback and

adaptability to local conditions. Overall, the quality of extension services

does not compare favorably with India (Box 15)

2 Byerlee (1994).

60

Box 15

Comparina Extension in Pakistan and India

There is some evidence that the extension service in Pakistani hasgreater problems than extension in India. Husain et al (1994) find thatalthough the T&V system has increased the quantity of extension advice(visits, -messages etc.)/ it does xnot appear to have increased farmers'technical knowledge or the rate of-adoption of new technology. This-contrastswith evidence from a comparable region of India. Two major problems specificto Pakistan are identified. First, implementation of T&V was lacking. Forexample, one survey (Malik, 1994) found that 20 percent of contact-farmers didnot know that they were contact farmers-, and hence the information flow neverwent beyond them.ta Second, messages were inappropriate to farmers'circumstances, and ignored the rational and often location-specific -tradeoffsthat farmers often make- For-instance, wheat messages continue to assume th-atwheat will be planted on-time, whereas farmers often delay planting wheat as aresult of interactions in double-cropping systems.1

4.22 With no-one happy with the performance of the extension service, majorreform of the service is a top priority. The traditional purpose of extensionas the disseminator of new technology needs to be rethought. In the earlygreen revolution stage there was an advantage to speeding the diffusion of theHYV seed and fertilizer package to farmers. However, with HYVs now widelydiffused, it is hard to believe that there is a technological package thatwould be profitable for farmers, but has not already been implemented by them.

4.23 The notion of extension as a top-down supply-driven process needs to berevised. The goal should instead be to create a demand for information amongstfarmers, a demand which could then be satisfied by extension workers. Theservice should concentrate more on participatory problem solving with farmersat the local level, which means substantially improving education levels offarmers and extension workers.

4.24 Research-Extension Linkage: The research-extension linkage is notstrong. For instance, in Punjab, the research and extension servicecollaborate in constructing a recommended crop package which is then issued tofarmers each year. However, such a package (issued in the form of a leaflet)is inevitably more difficult for poorly literate farmers to implement. Inaddition, issuing an all-embracing technical package represents an over-reliance on the recipe approach to delivering extension messages.Restructuring either research or extension will depend on complementaryreforms in the other. Crop and resource management research will require

61

greater on-site specificity and decentralization, but will only be effectiveif the extension service delivers site-specific information.

III: Public Zxpenditur-s

4.25 Both in terms of size and trend, there is evidence that agriculture is

being squeezed as a public expenditure priority. Between 1982 and 1994, totalprovincial and federal spending on agriculture decreased in real terms by anaverage of 0.57 percent per year. Over 55 percent of total government spendingon agriculture has been current expenditures. While this has risensignificantly in real terms (3.26 percent a year), development expenditure hasdeclined by 3.69 percent per year (Tables 4.2 and 4.3). Current expenditurehas risen from about half of development expenditure in 1982, to about one-

and-one-half times in 1994. Furthermore, this growth cannot be regarded asfavoring agriculture, because food subsidies (wheat, sugar, and edible oil)have accounted for almost half of it. Some of these subsidies (for wheat) are

actually meant to lower prices for consumers, that affect producers adversely.Development spending is also inflated by the inclusion of fertilizersubsidies. Between 1982 and 1994, constant rupee subsidy-free developmentexpenditure actually declined by 3.74 percent a year.

4.26 Although the share of agricultural expenditure in the budget by itselfis not a good indicator of either the government's commitment to agriculturaldevelopment, or the quality of the expenditure program, a comparison withother countries shows that agriculture in Pakistan has been under-represented.

Agricultural public expenditure, as a percentage of total public expenditure

has varied between 3 percent and 8.5 percent between 1982-94, with an averageof 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).

4.27 The most important question is the composition and efficiency ofagricultural expenditure. Subsidies form a large chunk of public expenditure,but it is not clear that they have achieved their stated purpose. Moreover,nonsubsidy expenditure is often inefficient in areas where public sector

involvement is required. Finally, spending is not directed to the appropriateareas: the government is engaged in activities that could be left to theprivate sector, while areas where public involvement is required areneglected.

Subaidies

4.28 The largest fiscal outlay made by the government in agriculture has beento subsidies. They fall under two categories - current subsidy which comnprisesmostly of food subsidy, and development subsidy which includes fertilizer, seedand tubewell subsidies. The provincial governments have borne a large portion

of the current subsidy, but development subsidies have almost exclusively been

federal.

62

Table 4.2

PAKISTAN: PUBLIC EXPENDITURE IN AGRICULTURE(Current)

Constant Rupees - millions

19t2-D3 196344 1954-t5 19S5-S6 191-7 1997-8 198 S-9 1989-90 1990-91 1991-92 1992-93 1993-94 Growth Rate

Crop Agricultlure 260.19 299.34 306.35 331.97 261.15 3U3.06 254.05 333.11 344.49 377.U 366.11 376.32 4.13

Irrigaion A Land Reclaim. 676.10 949.19 1039.23 1101.56 150.54 736.94 973.76 906.47 930.66 1252.12 1070.57 1302.07 4.42

Toul Pftijab IU0 == ] 1760 UZLA 1522 1599.90 O DL26 I4

SWCrop Agriculit6 U.21 122.89 109.14 127.14 134.31 144.52 140.20 157.79 166.63 265.25 156.02 153.63 6.74

Irruigior 264.78 304.17 329.89 353.50 394.24 405.46 411.71 373-90 352.54 355.49 356.87 343.55 2.70

TOl SiaLdh 42a2 5 12L2 642.65 65 0 72 67. 2 7 6t7 .2 4.3 6

NWFCrop Agruhuwe 70.51 65.66 114.41 115.60 120.35 120.6l 113.41 106.26 106.62 116.40 122.13 130.13 7.69

Iniptioo 96.76 206.55 129.00 147.65 222.00 216.06 244.95 235.67 236.70 237.77 253.06 234.87 13.44

Tooal NWFP 216.4 2 71 3.7 4 4 2 .7 45 0 S 6 4.2 4 tt3 4 4 474 .0 IJ

Crop Agricukt 56.63 65.96 73.29 S2.08 95.00 104.70 107.75 103.95 100.09 119.25 235.40 143.44 13.93

Irritation 71.21 65.53 66.41 U.61 64.72 73.36 76.97 66.36 60.01 73.19 73.52 76.13 0.63

TotalBocmiu ew J2 1LJ 2 3L U4.59 .2.21

Fed., o .mSudubsdie * 2744.00 1316.00 112.00 236.00 475L.00 2501.00 2229.00 1914.00 978.00 492.00 na

Wh A &SugarSubsidy - a 1039.00 1312.00 112.00 236.00 3076.00 1552.00 1179.00 1I30.00 978.00 492.00 -0.73

Edible Oil Subsidy m 1179.00 1704.55 0.00 0.00 122.00 1675.00 649.00 1050.00 U.34 0.00 0.00 -10.00

Total Fed Cw Exp. = A leo.46 291700 4147 260 00 23 9L3 =22 00

Total Natioul CurrentExpendirm a" 57.6t = mu a =1J W.L02 15 JW& 6343.60 5310 IIL 5099 La3*

Source: Economic Survey of Pakistan 1993-94 and Statistical Supplement 1992-93. The considerable fluctuation in certain series from year to year indicates that the data may not always be

reliable. The 1993-94 figures are Budget Projections and actual data could be considerably different. The growth rates for Federal items are calculated from 1984-85 to 1992-93. All growth figures

are calculated based on the first and last periods, except the figure for total National Expenditure growth which is calculated using a regression trend line on a 2-period moving average series.

Other subsectors like fisheries, veterinary etc. are not shown individually but are included in the total figures.

63

Table 4.3

PAKISTAN: PUBLIC EXPENDITURE IN AGRICULTURE(Development and Total)Constant Rupees - million

19S2-tl3 1963t4 1984-85 1985-86 1986-7 1987-t8 19U8-89 1989-90 1990-91 1991-92 1992-93 1993-94 Average Giow3h Rate

AViahwc & Ciedd 215.92 211.79 14.16 225.38 22699 273.86 124.76 135.57 157.1t 138.77 171.28 70.58 8x7.79 -3.95Inimon 424.67 28.29 322.24 272.59 402.51 350.38 12.13 237.11 595.62 545.90 410.55 186.77 366.54 -1.47Taa Punab 7579 , 64J29 2 I 3976 9 2I iZ6. 2 6 ; 1 L

Aprit a AVt Credit I.S5 64.37 71.79 76.51 97.66 13805 78.58 73.58 89.70 50.88 59.22 4808 80.83hgution 193.57 172.17 19.74 193.97 284.91 262.34 181.10 255.91 313.04 273.46 238.53 228.31 232.61 2.51Tow Simmdh 3 65 29t L7Ii 4 4 , 32L4 4 "4 amJ2NHELArirkhure 40.24 48.06 48.43 47.05 58.77 72.21 90.19 164.71 58.26 142.40 56.64 64.78 75.23 3.41lfr',g in 34.22 50.48 55.S2 51.82 64.16 92.37 60.92 105.39 124.50 145.22 189.46 172.00 88.64 29.70TOWl NWFP JJ 3 42 U 12Lj _1. 227.45 aLLJ .Lx I22 laCrop Apiculture 5991 53.62 56.56 43.15 56.51 41.01 42.45 41.33 60.24 24.16 41.56 5623 47.32 -1.26hrfimion lt18.50 122.54 10083 186.52 156.37 154.57 177.35 166.12 19944 368.59 415.75 24352 196.96 15.77TOWl laschiran 4U37 27.71 21. Mg9 L2oa 1=55 57 0V42 4 ao= X j_72

Fod& Agriculure 179.07 1713.52 1551.57 2349.41 1579.56 1337.06 1181.60 1163.87 809.71 1130.17 711368 47931 1400.66 -6.08bTmgawn 873.31 894 11 544.85 867.91 1011.44 608.18 363.68 920.79 775.46 312.07 334 15 507.59 682.36 -4.76FeoodSwrage 219.58 118.76 175.74 47.64 69.83 29.92 3.31 50.94 31.82 5.28 0.00 000 71.17FIt

el SubsdtnFeniliur Subsidy 1623.00 3.0 1137.12 1766.69 788.77 1389.24 1428.66 674.U4 678.41 1224.00 333.33 309.00 1045.31 -0.96TewniI Subsidy 2.00 13.33 13.33 11.74 12.59 6.40 4.71 4.42 na 0 00 0.00 Ni

TOl Federal J= al2 -40 43 3665 15 La&2 47 LM 279L05 = = 2 ;Z=

TOl National DevEaperdoa6a2aorr a2L 6fflS SS 5- =2L a 1-4 ~436 41 L S2 468B 22375402 -369X

SWcA: Economx Survy of Pakistan. 1993-94.

The growth rates for Federal items are calculated from 1984-85 to 1992-93. All growth figures are calculated based on the first and last periods,except the figure for total National Expenditure growth which is calculated using a regression trend line on a 2-period moving average series.Other subsectors like fisheries, veterinary etc. are not shown individually hut are included in the total figures.

64

4.29 The ratio of current subsidy to current expenditure for the provincialgovernments, 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 97percent (Table 4.4). Similarly, development subsidies borne almost exclusivelyby the federal government have also been high, from 19-45 percent ofdevelopment expenditure. Subsidies have fallen to some extent in recent years,both absolutely and relative to total expenditure (Table 4.5).

Table 4.4 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 na na na

83-84 2462 na na na na na84-85 3792 1149 0.30 2917 2744 0.9485-86 4409 1498 0.40 1498 1318 0.88

86-87 5025 1836 0.37 313 112 0.36

87-88 5129 2278 0.44 535 358 0.67

88-89 4173 1209 0.30 4919 4755 0.9789-90 2894 na na 3319 3150 0.9490-91 3950 975 0.25 2393 2229 0.93

91-92 4236 821 0.19 2085 1914 0.9292-93 4114 877 0.21 1196 978 0.82

93-9429 3474 963 0.28 540 492 0.91

82-83 1458 2 .00 4604 1625 0.35

83-84 1259 13 .01 4130 1233 0.30

84-85 1286 13 .01 3665 1151 0.31

85-86 1365 12 .01 5240 1778 0.34

86-87 1633 13 .01 3665 801 0.22

87-88 1638 6 .00 3621 1396 0.38

88-89 1187 5 .00 3150 1433 0.45

89-90 1422 4 .00 2941 652 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: Tables 4.2 and 4.3

4.30 Current subsidies, that have gone to support food procurement and pricestabilization have been the government's cornerstone for food autarky. This is

29 The 1993-94 figures in Tables 4.4 and 4.5 are Budget figures, and thussubject to major revision.

65

evidenced by the standard deviation of wheat production which has been lessthan 8 percent of the mean in 1978-1987 (Pinckney, 1989). Moreover, thegovernment has insulated consumers from major swings in wheat prices throughmassive procurement and marketing operations supported by the food subsidy - 24percent of production in 1970-80, and 31 percent thereafter (Alderman, 1993).In comparison, food procurement in Bangladesh and India have been below 10percent per year.

Table 4.5 Food Subsidy and Expenditure

(in Millions of Constant Rupees)

Year Total Total Current Subsidy as % Subsidy asSubsidy Expenditure Current % Agr. GDP

Expenditure

82-83 na 2301 na na83-84 na 2572 na na84-85 3893 6709 58 4.4285-86 2816 5907 48 3.0086-87 1948 5337 37 2.0187-88 2514 5542 45 2.5388-89 5960 9092 66 5.6389-90 2501 5563 45 2.2990-91 3204 6344 51 2.8091-92 2735 6321 43 2.1892-93 1855 5310 35 1.5493-94 1764 5408 33 1.45

4.31 Although the country has achieved a satisfactory level of food security,the policy of public intervention has come with a price. The level of wheatproduction has remained low because both farmers and traders were discouragedby low prices. Until recently, Pakistan had used the rationing system to ensurelow wheat flour prices to urban as well as some targeted rural consumers. Therationing system is now gone, but it has been replaced by a program of openmarket 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 controlledinternational trade of wheat by prohibiting export and importing irregularly toreplenish domestic stock. Thus, domestic prices of wheat have been lower thanimport parity prices. Meeting this price differential, along with the cost ofstorage and maintaining buffer stocks, has been a heavy fiscal drain for thegovernment. The cost of maintaining food subsidies has been high, ranging up to66 percent of current expenditure (Table 4.5).

4.32 This system is not achieving its stated goal because the price of flouris already market-determined. Millers are absorbing most of the subsidy onwheat as rent, and the milling industry has excess capacity. In 1993-94, the

66

retail price of flour was 95 percent of the import parity price of flour3 , sothe effect on the retail price of flour of the subsidy is minimal.

4.33 Food security is best ensured by raising the real income of households.Control of wheat prices runs counter to the goal of raising the real income ofpoor farmers. The government on one hand is squeezing the farmer's income bylowering the price of his output, and on the other hand trying to reduce thedamage by giving him some help through fertilizer and seed subsidies. However,the subsidy regime has been ineffective for a number of reasons.

4.34 Subsidies on such essential factors as fertilizer and seeds distort inputmarkets. Moreover, government activities in procurement, distribution andmarketing of inputs are almost always inefficient and unresponsive to the needsof buyers. For example, the cost of fertilizer distributed by governmentagencies in Pakistan has been much higher than the private sector, and thisinefficiency loss is borne by the public exchequer. Public seed corporationshave not been effective either. Most of these corporations have been running ata loss, as shown later in this chapter. The farmers have been deprived of thebenefits of competitively driven prices and an active commercial market.

4.35 Cheap inputs (in the short run) also encourage waste through overuse,particularly by poor farmers not well versed in agronomy. Fertilizer and seedsubsidies are essentially regressive since they benefit mostly the largerfarmers that use large amount of the subsidized inputs. Also, theadministration of the subsidy program encourages rent-seeking.

4.36 Other subsidies have outlived their usefulness. A subsidy on seedlingswas designed to spur the growth of private forestry. Rapid growth in forestryhas been achieved, but the principal beneficiary of the subsidy have been largefarmers, some of whom have evicted tenants in order to grow block plantations.

4.37 There are, also, a host of indirect subsidies that have serious impacton fiscal deficits. Qureshi (1993a) has shown that most of these indirectsubsidies (electricity, water and credit) have added anywhere from 3.64percent to 7.67 percent to the overall fiscal deficit.3 1Again, anycontribution that these subsidies have made towards assisting poor farmershas been neutralized by poor incentives and distortions, as noted in ChapterIII. Furthermore, the average consumers, who are the intended beneficiariesof government intervention, have not received significant benefits because ofleakage and inefficiency in state organizations. With regard to the wheatsubsidy, since the price of flour is already market-determined, the benefitto consumers of the wheat subsidy is minimal: it is in fact mostly a transferto millers.

30 The import parity price of flour is computed as the import parity price ofwheat plus milling and distribution charges.

31 The study omits another important subsidy -- seedlings in the forestry sub-sector.

67

in PakittAhi thei f etilizer.. subsidy has mr td theK l g ::isc OA: ~~~~~~~~~:S rersn vii: z e. 0 Are ti!4 i: :: "t : 4 I n Pakist W0- t he t::f-etilizerm tt.s1it. .-Ss;.; epr s .............................. .. ..... ... ....... ...... . t ...... ......

ooutlay of the g ernmnt. after tho. food tsusidyi;. - - e e aXlong period isa questio-le.. Thereis ea reakbl lack o c btween

0tertil-i-zet.-s*gi4.j;m$.00:A. eS.¢f ....~~~. ......... ... .. ... .. ......1t00t0x; ; ;.0

ThX fAt00 .. 0.fNese.,,neati.ve,, ,ca,'''"es,p,'' sug'est strongly t t fe.ti1i ............ strates tare .not affected by; the.t ........................... subsid:. .:.SVso th-i:e hget o ~el.. ........................ ....t o t.~ubsis4 hs .been -a .deed-. l os:: tc tbe~~ n5~t. ~ i4ieM i...on a ut:the:.. i u .-i i iand S i ::-::h .:fi-Y : : aee ete t i phase o:t cietelye t by 1ivi~boiO '~# b4arx : d dh... 4.1

- 0Xtj f- t 00 t ;0Ai i- - - - 4 i . i X ti: i i Xii -::: i.:-i i o - .2 i l[i . .... ...3.* :: : :T :: f i vet tilizer. Sii .......................... and .co.esumpt.im ..P;Ust. .. .a . x .... . ... . . ......... ......... .........

Q~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..... ..... .:....... . ... .......tdOO Or 404 .f ~ Fer $zurSubsiry a14 -C:-l ilit - i-..t: i. :-ki...

_S0 w ! F <S<-<<,~~~~.. . .... .... . .. ... .... . . .. ...... ......... ... . , ,. .. ....... . . . .. . .. . .....

0 100 .00 .0 j 0 0 ; ; ....... '', , 0 ' , ,.,,- '.'"'v""''' ' '"' ....... '''"'"'' 0 '4004 ... .' ..' .. ...........

fi~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. .... ................

finance.~~~~~~~~~~~~~~~~~~~~ Implici .usde ........ priin .... an rnses tonn

2000etitive 0.0te-owned enterprises cause similar problems.and.contribute

financial institutions. Some loans are not repaid, which implies an......... addit......(unintended) subsidy. Subsidized credit must be rationed, and unequal accessX . ...........

to0. 00nd and power almost guarantee that largefarmerswill.obtain.most.of.the

sub0idy.

32 Pearson correlation coefficient0between,subsidy0and0fertilizer.consumption.33 0Fertilizer.consumption0is measured in thousands.of.metric.tons.... . ... . ... .. . . . ... .. ... .. . . .6 8

800.00~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~................ . ......000.00~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.... ... . .. .... ...... ...... ...

400.00~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ . ..... .. ... I....

20 l04,.....0...0344 0445 I~~~~~~~~~~~~-A0 *~~~~~~4V *V.t~~~~~~~ 8..... . ... .. ............ # ...g ....

4.38 While explicit subsidies may appear on the surface to transfer income~~~~~~~~~~~......... - _ .... _%to the poorest groups and help the sector, they~~~~~~~~... .distor input.... andoupu

finnce wimeeplicit subsidiesthoug apericing polic srando transfers toncon-

competitive state-owned enterprises cause similar problems and contributeto low investment and farm productivity. The concentration of expendituretowards payroll and establishment costs and the diminution of operationalexpenditure is another sign of inefficiency.

4.39 In the past, farmers have received interest-free short term loans, aswell as low interest medium - and long-term loans from the nationalizedfinancial institutions. Some loans are not repaid, which implies an additional(unintended) subsidy. Subsidized credit must be rationed, and unequal accessto land and power almost guarantee that large farmers will obtain most of thesubsidy.

32 Pearson correlation coefficient between subsidy and fertilizer consumption.

31Fertilizer consumption is measured in thousands of metric tons.68

4.40 The shortfall between water charges and expenditure on irrigation alsorepresents a major subsidy. The active market for tubewell water shows thatfarmers are willing to pay more for water than the current price. Thisunderpricing has led to inefficient overuse of water, deterioration in thesystem resulting from the squeeze on O&M expenditure, and massive rentseeking, with most of the rents going to large farmers and irrigationofficials.

4.41 Electricity has also a big hidden subsidy. It is provided at aconcessional rate for operating electric tubewells. Electricity thefts arealso considerable. However, as the current power crisis in Pakistan shows,this subsidy is of dubious benefit to farmers. Rural areas can be withoutpower for up to 12 hours per day. Capital expenditure on electricitygeneration has been crowded out by the low rate of cost recovery.

4.42 Qureshi (1993a) estimates that concealed subsidies on credit,electricity, and water, which reached almost Rs 5 billion in 1990-91, are ahuge burden on the exchequer. Instead of helping the poor or promoting theefficient allocation of scarce resources, the subsidies have misallocatedresources and perhaps worsened rural income distribution.

Inadeauate Expenditure

4.43 Public expenditure has been insufficient in other areas. Naturalresource degradation arising from waterlogging and salinity has not beensufficiently addressed. To the extent that these problems are due toinadequate drainage (a public good), then the government can play an increasedrole. At the same time, the government has provided funds for the developmentand maintenance of public tubewells, even though most of the benefits oftubewells are privately appropriated (although there may be some publicbenefit of tubewells arising from lowering of water tables).

4.44 Particularly serious is the neglect of operations and maintenanceexpenditure on the irrigation system, which has consequently deteriorated(Table 4.6). The shortfall varies by province, and is as high as 37 percent inSind, and stems from low water rates and inadequate assessment and collectionof charges . 4 In addition, revenue does not go directly to the ProvincialIrrigation Departments, which creates poor collection incentives.

34 Moreover, the required levels did not include public tubewells, on theassumption that these would be privatized. But privatization of groundwatertubewells is slower than planned, and the O&M needs of these tubewells hasbeen considerable.

69

4.45 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.

Table 4.6: O&M Expenditure and Requirements (in Million Current Rupees)

Fiscal Year 1988 1989 1990 1991 1992

Requirements 1704 1875 2161 2408 2617

Expenditure 1704 1513 1617 1718 1985

Shortfall(*) 0 -19 -25 -29 -24

Source: World Bank (1993)

Inefficient Expenditure on Plic nterprises in Aariculture

4.46 Although the expenditure on public institutions is not included in the

budget, their losses are met mostly by borrowing from banks guaranteed by the

government, and sometimes by direct support. Guaranteed borrowing is

effectively part of the fiscal deficit and crowds out the private sector from

the credit market. 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

agriculture output. 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 Sind 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 are inefficient, with a high cost of

operation compared to the private sector. For example, a study (Government of

Pakistan, 1987) showed that the private rice mills operated at 40 percent less

cost than the RECP. Table 4.7 summarizes the operating profit and losses of

four major agriculture sector corporations in Pakistan.

4.47 Audited accounts of the Punjab Seed Corporation show that the company

made small profits in the early years and improved only in recent years. These

operating incomes (before tax) are before the seed subsidy given by the

Government of Punjab, which has run between three to five million rupees per

year. On the surface, this corporation does not seem to have done too badly in

the recent years. However, certain expenditure items have not been taken into

account, whose inclusion would in all probability depress the performance

picture. First, the seed farms acquired from the Punjab Agriculture Development

and Supplies Corporation (PADSC) have not been valued. The implicit land rent70

of these farms must be added to the operating costs to get a true estimate ofcosts. Second, the corporation's loans from the nationalized banks have beenunderwritten by the government at around 12.5 percent, which amounts to anindirect subsidy. Third, the corporation had borrowed working capital of Rs.50million from the Government of Punjab, but has not paid the interest on thisloan. Fourth, the corporation receives occasional grants from the Government ofPunjab for expansion of facilities.

Table 4.7 Consolidated Profit & Loss of Major Public Corporations in theAgriculture Sector (in Million Current Rupees)

1988 1989 1990 1991 1992 1993

PSC 1.74 4.38 18.99 14.65 21.66 31.00SASO - -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

4.48 Sind Agricultural Supplies Corporation distributes fertilizer and seedsto farmers in Sind. The corporation receives reimbursement of the distributioncosts on a predetermined formula, which is not sufficient. As a result, theorganization has had operating losses in all the years for which data wasavailable (1989-93).

4.49 Available data for the Punjab Agricultural Development and SuppliesCorporation also show net operating losses in all years between 1986 to 1993.The Pakistan Agricultural Storage and Services Corporation, which is engaged inprocuring, preserving, storing, and distributing food grains (especially wheat)has been operating with a budget close to 3 billion rupees. The operatingfinances of this organization shows major losses prior to 1987-89, which waspartly offset in 1990, 1991, and 1992. The corporation again went in the red in1993.

4.50 The role of the government in providing marketing services in an economynot characterized by market failure is extremely tenuous. A review of marketinginstitution by the World Bank in 1990'shas indicated that problems could arisewith parastatal marketing in developing countries. Some are evident inPakistan. First, under a system of controlled prices, inadequate marketingmargins are the primary reasons for the inadequacy of marketing servicesprovided 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 operationalcosts from the consumers. The prices set by the government are also imposed onthe private sector, either through administered prices or by the presence ofgovernment marketing agencies. This has resulted in inefficiency of supply,such as untimely availability, poor distribution in low demand areas, and soon.

35 Agricultural Marketing: The World Bank Experience, 1974-85 (1990),Operations Evaluation Department, The World Bank, Washington. D. C.

71

4.51 Second, for large parastatals, the cost of inefficient operation isreflected both in the price and the level of service in Pakistan. There isevidence that private sector prices have been less than state enterprises inmany instances of distribution of fertilizer and seeds. Even with the presenceof parastatals, private sector marketing channels have continued to servicefarmers, and are often preferred by them. This seriously calls into questionthe argument that expansion of public-sector marketing services is needed tocheck alleged exploitation by merchants. So, there is no justification for thepublic sector role in marketing and the record shows that parastatals have beenwasteful. The resources devoted to them could have been directed towardsexpenditure that would benefit agriculture, such as promoting the adoption ofproductivity-enhancing technology by farmers, building infrastructure to linkmarkets, and supporting private entrepreneurship.

72

Chapter V: The Strategy

5.01. How to sustain and even improve past growth in agriculture? Whatpolicies and programs are needed? These are two fundamental questions facingPakistani policy makers in agriculture. Answering them will mean firstdefining the appropriate role for government -- to encourage the developmentof a smoothly functioning market, through institutional and regulatory reformsthat facilitate private sector activities and market efficiency. Where marketfailure is not an issue, and government inefficiency is clearly evident, thestrategy will require that government's role must be reduced. This meansreform in some policies and strengthening the liberalization now underway.

5.02 Investment and public expenditure on agriculture will need to bereshaped. Government spending needs to focus on public goods and marketfailures, and not on activities better suited to the private sector. Inpoverty and the environment, the government will have an active role. Marketfriendly policy reform will not only ensure efficiency and growth, but alsohelp reduce poverty and protect the environment. In some cases, when marketoutcomes are inconsistent with poverty and environmental objectives,government intervention will be justified. Below is an outline of the neededpolicy and structural reforms summarized in a matrix at the end of theexecutive summary. This matrix is based on a first round of discussions, andwill be modified and further specified on the basis of future dialogue withthe government. It is intended that the measures listed in the matrix willform the basis for an action plan, to be jointly prepared with the government,for implementation of structural reforms in agriculture.

l: Policisa Affecting Incentives

5.03 Price Policy: In keeping with structural reforms, output pricesshould be market determined, so market signals will be transmitted to farmerswith least distortion. The best option for the government is to remove pricesupports and controls, because these distort market signals and have hugefiscal costs. Such liberalization will generally improve production incentivesfor crops in which Pakistan has a comparative advantage. The government aimof reducing inter- and intra- year price fluctuations can be achieved by othermore cost effective means, such as promoting on-farm storage, private-sectorstorage, and futures trading. A study could be initiated immediately todetermine cost-effective means for reducing agricultural price volatility inPakistan. The government will need to consider ending the subsidy on wheatimports. Since the price of flour is already market-determined and close toimport parity, this will have little effect on consumers, but producers mayget prices close to import parity. The government will need to ensure thatthere are no import restrictions on wheat and flour. It would be desirable toremove protection of sugarcane through high support prices and sugar importrestrictions, allowing diversion of scarce resources (such as water and land)to more efficient uses.

73

5.04 It is, however, crucial to ensure that there is an orderly transition tomarket-determined prices, while support prices are phased out. The Governmentwill need to ensure that a competitive and efficient private sector is readyto step in as it withdraws. Where competition is not possible (such as insugarcane processing), strict regulation will be necessary. Without adequatepolicing, the move to private sector orientation will be discredited and thewhole reform process will suffer.

5.05 Trade Policy: Trade reform will have to be completed speedily. Ingeneral, the government may not disprotect products in which Pakistan has acomparative advantage, while those in which it has no comparative advantageshould not have protection. In particular, taxes on cotton exports, duties onsugar imports, and quantitative restrictions on both will have to be removed;any loss of revenue would be offset by removing the wheat subsidy. The termsof trade for agriculture will have to be corrected by lowering industrialprotection. East Asian countries have benefited from having a much smallerbias against agriculture in their trade policies. As recommended by the recentBank report on trade policy, it would be desirable to replace the system ofhigh and dispersed duties and barriers with a uniform duty scheme. Ratherthan aim at self-sufficiency in industry, it is better to direct resources totheir most efficient use. Using agriculture's comparative advantage is thecheapest way to finance import requirements.

5.06 Fiscal Policy: At the same time, tax reform will need to ensure thatall sectors are treated equally as far as tax burden is concerned. Allspecial agricultural tax exemptions should be eliminated. This will becomeespecially relevant once intersectoral transfers -- now happening because ofprotective trade policy -- have been corrected. The system of agriculturalincome and wealth taxation that has been proposed shall be speedily put inplace and extended to meet the goals of equity and efficiency. The tax baseshould be periodically reassessed to ensure that the poor are not undulyaffected, while ensuring that taxes do not create intersectoral distortionand inequity.

II: Policies Affecting Input Markets

5.07 The government will need to halt the sort of micro-management that canbe better undertaken by the private sector. Regarding fertilizer, thegovernment has already moved towards privatizing urea production anddistribution, and this will have to be extended, by making the NationalFertilizer Corporation (NFC) more commercially oriented. The NFC could bedivested to the private sector. Privatizing the import of phosphaticfertilizer will have to be speeded up. Given the strategic importance offertilizer, the government will ensure that private firms are given the samefacilities as the public sector had. A smooth transition to a market dominatedby the private sector may require that public sector imports are phased out.Also, the ability of the private sector to take over imports effectively willdepend on it having a level playing field with the government during thetransition. Policy reform in other sectors will likely require decontrol of

74

natural gas prices, and fertilizer prices will likely rise, reflecting fulleconomic pricing of inputs.

5.08 There is no need for government presence in the marketing of improved(certified) seed. Government seed corporations should be run along strictcommercial lines and be privatized as soon as possible. Private-sector seedcompanies must have to be provided with a level playing field. Qualitycontrol should stay in the public sector and the Seed Certification Departmentwill need to be fully funded and strengthened further as use of improved seedincreases from the present low level -- 5 to 6 percent for wheat and rice, andover 50 percent for cotton.

5.09 Irrigation should undergo a demand-based decentralization, through thedevelopment of water user associations (WUAs) at distributary level. TheseWUAs would be participatory farmer organizations that will assumeresponsibility for downstream operations and maintenance of irrigationsystems. Upstream at canal command would be the responsibility ofcommercially oriented public utilities (PUs), and explicit contractualobligations would exist between the PUs and the WUAs. At provincial level,autonomous water authorities would be responsible for major provincialstorages, link canals, off-farm and provincial drains, and flood control andmanagement. Farmers, in turn, should be entitled to expect timely andefficient water delivery from the WUAs.

5.10 Water user associations have been successful in other countries. In thePhilippines, more than 5600 communal irrigation schemes with full legal statuscover almost half of the irrigated land. They conduct operations andmaintenance and operate on a self-financing basis. In Nepal, 70 percent ofall irrigation is now under farmer control, and experimentation with farmermanagement is increasing in Central and South America.

5.11 Irrigation charges can be raised to reflect the cost of provision, thequality of service provided, and the cost of competing alternatives. Disputesbetween the different groups in the irrigation system will be resolved bypowerful regulatory authorities, free from political influence. A legal andinstitutional framework for the market exchange of water rights will also beestablished. Off-farm drainage, a public good, will continue to be theresponsibility of government. Costs of drainage can be recovered fromfarmers.

5.12 A careful study of the costs and benefits of land reform could beinitiated. However, some immediate measures would be highly desirable tocorrect land market distortions, which include low machinery prices (ofteneffectively zero through the use of delinquent loans to buy machinery), andunequal access to credit (which makes it difficult for small farmers to expandtheir holdings) . These reforms would increase opportunities available forfarming, and for wage labor on employment-intensive farms. The land titleprocess can be modernized and streamlined, including the establishment of asystem of permanent title deeds to land. Security of tenure shall be assured,without creating further disincentives to rent-out land. This wouldfacilitate long-term (especially natural resource management) investments.

75

5.13 It will be desirable to end directed credit in any form. Groups findingit hard to gain access to credit could perhaps be helped by a scheme tounderwrite the setup costs of credit. Such assistance would be one-time andreduce the transaction costs, and the best option is not to have any recurrentsubsidy. It is necessary to ensure high loan repayment rates to ensuresustainability and any such scheme should not interfere with liberalization ofinterest rates. The government will need to consider developing aninstitutional and legal framework to allow efficient lending to agriculture,unhindered by highly restrictive collateral requirements and seasonal creditregulations.

I1I: Government IxDenditure

5.14 Most of Pakistan's agriculture has entered a post-green revolution stageof development that requires new strategies to enhance input efficiency andmaintain and improve the quality of the resource base. By most measures ofproductivity, Pakistan's institutions have not evolved to meet this challenge.Reform in public spending is central to these needed institutionalimprovements.

5.15 Support for research should continue, but expenditure needs to berestructured so that salaries do not soak up most of the available funds.Research institutions will be made more autonomous, salary restrictionslifted, and other funding sources mobilized. Training of researchprofessionals will include an understanding of the problems of farmingsystems, including irrigated agriculture. Greater importance should beattached to research in cropping systems. For instance, a build-up of pestand disease problems due to continuous "monocropping" of the prevalent doublecropping systems such as, rice-wheat or cotton-wheat, calls into question theviability of such systems. A pertinent example for the rice-wheat system isthat of Phalaris minor which has become a major weed in wheat in the rice-wheat area over the past two decade. Thus, research needs to focus onsolutions to such problems, even on cotton, by focusing on farming systems.Universities will continue to play a key role in manpower training forresearch. Publicly funded research will stress growth enhancing public goods,environmental impact, and poverty reduction.

5.16 Research which can profitably be financed by the private sector, neednot be underwritten by the limited public budget. Joint public-privatefunding of research is also desirable. Private and public sector researchfinancing could, where possible, be on a competitive basis, meaning thatfunding to institutions would be on the basis of performance. Monitoring ofresearch needs to improve, though this will have to be consistent with greaterdecentralization. Coordination between different research institutions willhave to be improved, and unnecessary duplication avoided. Producers, NGOs,and other participants from the private sector should be encouraged to takepart in setting research priorities. Transfer of technology will be anintegral part of the mandate of research institutions. Coordination betweenprovincial research agencies will have to be encouraged.

76

5.17 Both productivity and sustainability can be enhanced by improvements incrop and resource-management research. Greater attention to local conditionscan be achieved by defining major agro-ecological zones and subsystems inthese zones. These zones and systems, rather than political boundaries, wouldthen be the basis for organizing crop and resource-management research.Research should be directed at monitoring changes in the resource base at thefarm level. There is a special need for improvements in the major croppingsystems, overcoming problems such as delayed planting, and depletion of soilorganic matter. Adaptive research (which would be a part of the extensionservice, outlined next) would provide site- and season-specificrecommendations and information to adapt them to each farmer's needs.Incentives could be provided for greater participation of farmers in definingresearch priorities.

5.18 The extension service may have to be reduced in size. The cadre ofextension agents at Union Council level may have to be gradually scaled down.Instead of too many extension agents with too little training, there will befewer, better qualified agents. It would be desirable for extension to beextended to cover women farmers. Additional public funding should becontingent on an improvement in organizational capability.

5.19 Some extension is already being undertaken by the private sectorcompanies. These and NGOs could be encouraged to increase the provision ofextension services. A system of advisory services by adaptive researchinstitutions to medium and large farmers on payment of fees could beencouraged.

5.20 Operational management of extension services needs to be decentralized,and farmers should play a bigger part in control and evaluation of theseservices. Extension services will have to be geared to problems of allfarming systems (including livestock, forestry, and water management) and notjust major crops. It will be desirable to have a more diversified approach toextension -- one that responds to varying needs of farmers, and uses variousavailable sources of extension services, including the private sector.

5.21 The "contact farmer" approach to extension will have to be reconsidered.A group approach to message delivery would be broadly based, it would alsocomplement groups in other areas, notably in water and credit provision. Sucha system would rely more on audio-visual means, extensive on-farmdemonstrations, and farm fairs.

5.22 Greater investment in formal schooling will also be critical in raisingtechnical efficiency and productivity. Education raises a person's ability toacquire and process information, and to respond to changes in the environment.Increased training will also prepare farmers for the increased organizationalburden which will be placed on them through the formation of water-userassociations and other user groups.

5.23 The government's role in rural infrastructure provision needs to bestrengthened. Increased revenue from the tax reforms and savings from therationalization of public expenditures on agriculture could be used to financeroad work. Farm to village roads would improve the distribution of inputs andmarketability of outputs. The transition to high value (often perishable)

77

foods, would be greatly facilitated by better roads. Greater public spendingis also needed to improve rural communications the provision of energy foragriculture.

5.24 The government's new role will require institutional changes, butgovernment institutions can adapt to the new environment, so no independentinstitutional reform should be needed. Some institutions (such as outputmarketing agencies) will have to be phased out as the reform programprogresses.

IVz Pov-rty

5.25 Speeding up rural poverty alleviation means rapid growth of agriculture,which in turn depends on government policy. Price reforms that improve theincentives to farmers are specially urgent in this regard. Moreover, for agiven level of growth, reforms are required to maximize the associated povertyreduction. In a labor-abundant economy, subsidies on capital (such as tractorsand combine harvesters) do not make economic sense because they distort factormarkets and lead to labor displacement. Incentives to mechanize, which leadto labor displacement without a worthwhile increase in output, will have beremoved.

5.26 Provinces will have to be encouraged to better target their developmentspending towards the rural poor. This could be done by first identifyingdistricts with the biggest concentration of poor people, and giving themhigher priority in the allocation of public investments, such as rural roads,and health and education facilities.

5.27 Rural credit policy should ensure that viable small-scale ruralinvestment opportunities can be efficiently financed. Directed credit,quotas, and interest-rate ceilings should end and the private sector should berelied on for credit provision wherever possible. If commercial banks areunwilling to provide credit (because of high transaction costs) thencommunity-based lending institutions (along the lines of Grameen Bank) shouldbe encouraged.

5.28 As noted, the government needs to consider removing land-marketdistortions which impair efficiency as well as restrict access of the poor toland.

5.29 Participatory community-based organizations also offer great promise inpoverty reduction. These can be integrated with credit provision, andinstitution building in rural communities will be complementary with increasedreliance on user groups in water and extensions services.

Vs Invironment

5.30 The best option for the government is to place more emphasis on naturalresource management problems in agriculture. Policy interventions for naturalresource management and the environment will be based on three principles:

78

price adjustments for scarce natural resources in order to provideappropriate conservation incentives;

* identifying regulatory mechanisms which could be effective in addressingmarket failure, bearing in mind the poor record of existing regulatoryagencies; and

* restructuring public expenditures to focus on natural resourcemanagement priorities.

5.31 The government needs to phase out policies that give the wrong signalsto private agents for use of natural resources. Take, for example, pricing ofirrigation water. Although increasingly scarce water is providedsubstantially below its economic price. Water pricing should be resolved aspart of the overall reform of the irrigation sector. Factor price distortionsthat lead to labor displacement can increase environmental stress (farming onmarginal land or rural-urban migration) . It will be helpful to remove suchdistortions.

5.32 Related to pricing is the removal of subsidies. Witness the saplingsubsidy, whose rationale has disappeared with the rapid planting of trees onprivate farmland. Abolition would release resources for worthwhile naturalresource management projects.

5.33 In environment and natural resource management, market failure is morelikely to be a problem. Many market failures, such as the excessiveapplication of harmful pesticides, will require public regulation. Increasedpesticide use has created growing resistance among pests, and destroyednatural predators. Integrated pest management would be more effective andenvironmentally friendly, as well as consistent with the demands of Pakistan'sexport markets. An effective institutional mechanism for transmittingknowledge about Integrated pest management is essential.

5.34 Increasing the power of regulators will be counterproductive if theagencies still protect the powerful, are rent-seekers, or enforce the lawselectively. It may, therefore, be wise to include NGOs in Pakistan'senvironmental strategy. Much of the awareness and some of policy work on theenvironment has been undertaken by NGOs, and they have had notable successesin the civil courts. They will be supported wherever appropriate and accordingto their comparative advantage.

5.35 Lack of property rights and institutions to manage common propertyresources can inflict on-site damage and create negative externalities.Successful watershed management projects need to be extended. Project designshould be sensitive to the creation of community management institutions toaddress common property resource management problems. Interventions shouldtake the form of providing incentives for the adoption of sustainable resourcemanagement techniques. New technologies which can enhance the physical statusof common property resources should be encouraged.

79

VI: Core Elements of the Strategy

5.36 While this entire strategy for the future will have to be implemented toensure sustainability, some elements have higher priority.

Moving Towards Market

5.37 Those policy reforms that support the ongoing structural adjustmentshould be top of the list. These reforms include all actions at governmentwithdrawal from activities better suited to the private sector, and those thatallow market forces to operate unhindered, allocating scarce resourcesefficiently. These actions will include withdrawing from agricultural cropmarkets (such as wheat, rice, cotton, and sugarcane) and leaving prices to bedetermined solely by market forces. Reforms also include the withdrawal ofthe government from input market, such as imports of fertilizer. Withliberalization of trade and tariff policy, the incentives for the sector willimprove and distortionary policy-induced intersectoral transfers out ofagriculture must end. To make the tax system equitable and non-distortionary,agriculture must contribute a fair share to government revenue, which willalso advance the of internal balance and stabilization. These reforms willhelp adjustment (including stabilization), as well as improve agriculturalefficiency and growth.

Addressing the Irrigation Crisis

5.38 Irrigation is the life-blood of agriculture in Pakistan. To address thepresent crisis in irrigation management, market-determined incentives (waterprices) must be allowed to direct the irrigation system. The scope for raisingwater charges is considerable, and this measure can be implemented quickly. Inthe longer term, decentralization of water provision is essential andpreparatory steps can begin. The Government will have to draw up enforceableproperty rights to water. Property rights and legalized water markets willmake the opportunity cost of water transparent, leading to greater efficiencyin use. Administering institutions will be restructured along commerciallines.

5.39 The best option is to develop autonomous, commercially-oriented publicutilities on a canal command basis, ensuring cost recovery of all currentoperations and maintenance and future capital expenditures. Since thegovernment cannot collect all costs of operations and maintenance from thefarmers, it is crucial that farmers themselves become responsible foroperations and maintenance. Water user associations will be vital to the newmarket-oriented irrigation system. These will better maintain the system and,moreover, water will be used more efficiently to generate greater farm outputand income.

80

Reforming Extension

5.40 Given the potential of increasing productivity of most crops by 30 - 50percent, effective diffusion of existing technology and of best farmingpractices is paramount. Reform of extension services will have to take placein four key areas:

* closer links with research institutions;

* reducing the number of front-line extension workers and replacing themwith fewer, better trained workers more responsive to the needs offarming systems and not just given crops;

* opening consulting services by adaptive research institutes to betteroff farmers on a cost-sharing basis; and

* extensive use of mass media and other group approaches for basicmessages about available technology and better farming systems.

Correcting Distortions in the Land Market

5.41 A full-fledged land reform is difficult and can only be considered aftera complete study of costs and benefits. Some important measures, however,need to be implemented now. Foremost is rectifying the insecurity of tenureof many farmers, especially tenants-at-will. Security of tenure would improveresponsiveness to incentives and create better incentives for long- terminvestments. Property rights can also be reinforced by improving andstreamlining land registration. A system of permanent title deeds will beestablished. Finally, the government needs to consider eliminating artificialincentives to large holders, such as low machinery prices and unequal accessto credit.

81

Bibliography

Akmal, Muhammad, 1993. "The Production and Consumption of Livestock Foods in

Pakistan: A Look Into the Future". Pakistan Institute of Development Economics

Working Paper.

Alderman, Harold, 1993. "Issues for Food Security in Pakistan", in

Agricultural Strategies in the 1990s, Agha Sajjad Haider et al, eds., Pakistan

Association of Agricultural Social Scientists.

Alderman, Harold and Marito Garcia, 1993. "Poverty, Household Food Security,

and Nutrition in Rural Pakistan", International Food Policy Research Institute

Research Report.

Ali, Mubarak, and John C. Flinn, 1989. "Profit Efficiency Among Basmati Rice

Producers in Pakistan Punjab", American Journal of Agricultural Economics,

May,pp 303-310

Ali, Mubarak, and Derek Byerlee, 1991. "Economic Efficiency of Small Farmers

in a Changing World: A Survey of Recent Evidence", Journal of International

Development, Vol. 3, No. 1, pp 1-27

Binswanger, Hans and Gershon Feder, 1993. "Power, Distortions, Revolt, and

Reform in Agricultural Land Relations". World Bank Policy Research Working

Paper.

Blarcom, Knudsen, and Nash, 1993. "The Reform of Public Expenditure for

Agriculture", World Bank Discussion Paper No. 216.

Byerlee, Derek, 1994. "Agricultural Productivity in Pakistan: Problems and

Potential", prepared for World Bank Agricultural Sector Review.

Government of Pakistan, 1987. "Report on Paddy Rice Marketing in Pakistan".

Agricultural Prices Commission.

Goldstein, Ellen, 1993. "The Impact of Rural Infrastructure on Rural Povery:

Lessons for South Asia", Internal Discussion Paper, South Asia Region.

Gulati, Ashok, and Garry Pursell, 1993. "Liberalizing Indian Agriculture: An

Agenda for Reform". World Bank Policy Research Working Paper WPS 1172.

Hamid, Naved, Ijaz Nabi and Anjum Nasim, 1990. "Trade, Exchange Rate, and

Agricultural Pricing Policies in Pakistan". World Bank.

Haque, Nadeem Ul, 1993. "Agricultural Pricing Policy in Pakistan",

International Food Policy Research Institute Working Paper.

Ilahi, Nadeem, 1994. "Agriculture, The Rural Economy, and Poverty in

Pakistan", prepared for World Bank Agricultural Sector Review.

82

Krueger, Ann O., Maurice Schiff, and Alberto Valdes, 1988. "Agricultural

Incentives in Developing Countries: Measuring the Effect of Sectoral and

Economywide Policies". World Bank Economic Review, Vol.2, No. 3, 255-271.

Lee, Kyu Sik and Alex Anas, 1989. "Manufacturers' Responses to Infrastructural

Deficiencies in Nigeria: Private Alternatives and Policy Options". World Bank

Working Paper, WPS 325, December

Longmire, Jim and Pascale Debord, 1993. "Agricultural Pricing and Comparative

Advantage in Pakistan: An Update to 1991-92". Prepared for South Asia Region,

Agricultural Operations, World Bank.

Mahmood, Moazam, 1994. "A Macro Analysis of Time Change in the Distribution of

Agricultural Land in Pakistan", prepared for World Bank Agricultural Sector

Review.

Malik, Sohail, 1991. "Poverty in Pakistan: 1984-85 to 1987-88", International

Food Policy Research Institute Working Paper.

Malik, Sohail, 1994. "Agricultural Extension". Unpublished Background Paperfor Sector Study, Pakistan Agricultural Research Council.

Mian, Dr, Alim, and M. Yasin Mirza, 1993. "Pakistan's Soil Resources",

National Conservation Strategy Sector Paper 4, Environment and Urban Affairs

Division, Government of Pakistan, in collaboration with the World Conservation

Union.

National Commission on Agriculture, Ministry of Food and Agriculture,

Government of Pakistan, March 1988.

Noman, Omar, 1994. "Sustainable Development: What does it mean, and how can it

be implemented?", prepared for World Bank Agricultural Sector Review.

Pinckney, Thomas, 1989. "The Demand for Public Storage of Wheat in Pakistan",

Research Report No. 77, International Food Policy Research Institute.

Qureshi, Sarfraz Khan, 1993a. "Fiscal Impact of Government Interventions in

Agriculture".

Qureshi, Sarfraz Khan, 1993b. "Role of Rural Infrastructure and Financing

Requirements for Rural Roads and Education".

Rosegrant, Mark W., and Robert Evenson, 1993. "Agricultural Productivity

Growth in Pakistan and India: A Comparative Analysis", presented at Pakistan

Institute of Development Economists Ninth A.G.M.

Saleemi, A.R., 1994. "Agronomy, Crop Production, and Sources of Productivity",

prepared for World Bank Agricultural Sector Review.

83

Shabbir, Tayyeb, 1994. "Agricultural Prices in Pakistan", prepared for World

Bank Agricultural Sector Review.

Singh, Inderjit, 1988. "Small Farmers in South Asia: Their Characteristics,Productivity, and Efficiency", World Bank Discussion Paper No. 31.

Srinivasan, T.N., 1993. "Poverty Alleviation in South Asia", Yale UniversityWorking Paper.

Task Force on Agriculture, Ministry of Finance, Department of Development and

Economic Affairs, Government of Pakistan, December 1993.

World Bank, 1993. The East Asian Miracle: Economic Growth and Public PolicyPolicy Research Report.

World Bank, 1992. Islamic Republic of Pakistan: Changes in Trade and Domestic

Taxation for Reform of the Incentive System and Fiscal Adjustment.

World Bank, 1994. "Pakistan Irrigation and Drainage: Issues and Options".

Agricultural Operations Division, South Asia Region.

84

Report No. 1 3092-PAK Pakistan A Strategy for Sustainable Agricultural Growth - World Bank ·...

Documents

Transcript of Report No. 1 3092-PAK Pakistan A Strategy for Sustainable Agricultural Growth - World Bank ·...