Report No. 29373-PAK Pakistan Punjab Economic Report Towards a Medium-Term Development Strategy March 31, 2005 Poverty Reduction and Economic Management Unit South Asia Region Government of the Punjab Document of the World Bank TABLE OF CONTENTS EXECUTIVE SUMMARY ........................................................................................................... 7 CHAPTER 1: THE CHALLENGEAND THE ISSUES .......................................................... 15 1.1 CHARACTERISTICS OF INCOME ........................................................................................... 15 1.2 THEPOVERTYCHALLENGE................................................................................................. 18 1.3 CHARACTERISTICS OF THE LABOR FORCE .......................................................................... 20 1.4 THEEMPLOYMENT GROWTH AND CHALLENGE ................................................................. 21 CHAPTER2: PROMOTINGJOB-CREATIONINTHE NON-FARM SECTOR ..............24 2.1 DEVISING GROWTH A STRATEGY FOR SMALL AND MEDIUM ENTERPRISES ....................... 24 2.2 IMPROVING THE BUSINESS 28 2.3 STIMULATING FASTER GROWTH THE CONSTRUCTION SECTOR..................................... ENVIRONMENTFORFIRMS....................................................... IN 35 2.4 IMPROVING THE SYSTEM OF LAND-TITLING INTHE PUNJAB.............................................. 37 2.5 PROMOTING THE DEVELOPMENTCITIES AS GROWTH OF .................................... 2.6 TOWARDS BETTER STRATEGIES TO ACCELERATE GROWTH THE SERVICES SECTOR....43 ENGINES 39 OF CHAPTER 3: ACCELERATING GROWTHINTHE AGRICULTURE SECTOR ...........43 3.1 RECENT SECTOR PERFORMANCE AND PRIORITIES FORPUBLIC POLICY............................. 43 3.2 RAISING THE LEVEL EFFICACYOF PUBLIC EXPENDITURES AND ........................................ 46 3.3 AGRICULTURAL PRICES AND OUTPUT MARKETS ............................................................... 48 3.4 MODERNIZING WHOLESALE MARKETS ................................................................................ 51 3.6 UPGRADING THE AGRICULTUREINNOVATION SYSTEM: .................................................... 3.5 FACTOR MARKETSLAND WATER:....................................................................... FOR AND 54 58 CHAPTER 4: IMPROVING SERVICE DELIVERY AND SOCIAL OUTCOMES ............63 4.1 THESTATE OF EDUCATIONAND HEALTHTHE PUNJAB.................................................. IN 63 66 4.3 ACCOUNTABILITY OF POLICYMAKERS AND PROVIDERS: ................................................... 4.2 THEROLE OF PUBLIC SERVICES........................................................................................... 73 4.4 THEROLEOF THE PROVINCIAL GOVERNMENTUNDER DECENTRALIZATION ..................... 81 4.5 BETTERINFORMATION IS CRITICAL .................................................................................... 85 CHAPTER5: IMPROVING PROVINCIAL FINANCESFORDEVELOPMENT .............89 5.1 A FISCAL PROBLEM OF A DIFFERENT K~ND ........................................................................ 90 5.2 ADDRESSINGTHE PROBLEM................................................................................................ 95 5.3 MOVING FORWARD ........................................................................................................... 101 APPENDIX: A NOTE ON ESTIMATIONOF PROVINCIAL GDP: 1990/91-2001/02 REFERENCES ...............................................................................................................................105 115 STATISTICALAPPENDIX ..................................................................................................... 117 3 LIST OF FIGURESAND TABLES T.4BLE 1.1:GROWTH RATE OFREALGDP:PUNJABAND PAKISTAN:1991/92 -2001/02 ..................................... 16 TABLE1.2: POVERTYTRENDS: ALL-PAKISTAN ANDPUSJAB:1983/84 -2001/02.................................................. 18 TABLE 1.3:EXPENDITUREAND URBANIZATION INPUNJAB,BY DECILE:2001/02 .................... TABLE1.4:POVERTYAND VULNERABILITY INPUNJAB:2001/02 ..................... TABLE1.5:POVERTYHEADCOUNT REGIONSOFPUNJAB:1998/99 -2001/02 RATE: TABLE1.6:EMPLOYMENTBY GENDER AND PUBLIC/PRIVATE:2001/02 .................................................................. 20 TABLE1.7:EMPLOYMENTSECTOR, GENDER, BY AND SIZEOFFIRM:2001 /o TABLE1.8:EMPLOYMENTINPUNJABBYhhJOR INDUSTRYDIVISION: 1993/ TABLE1.9:CHANGE INMEDIANEARNINGS: PUNJABAND PAKISTAN:1993/ ................................. 21 TABLE1.10:SIZE OFTHE CIVILIANLABOR FORCE THEPUNJAB:1984/85 -2001/02 IN TABLE2.1: COMPOSITION OFTHE EMPLOYED LABORFORCE THE PUNJAB:1984/85 . ...................24 IN 2001/02 TABLE3.1:AGIIICULTUIW, EXPENDITURESPROVINCIALAGGDP: 1990/91 . IN 2001/02 ................................ 47 TABLE4.1:SELECTEDEDUCATION INDICATORS FORPUNJABAND PAI Revamp Other Arcane Regulatoy Legislation and Structures:Some of the regulatory laws are completely outdated, out of tune with technological developments, and beyond the technical competence of the enforcement machinery. In several instances, n e w products and instruments have become available that are better replacements and more effective mechanisms for achieving the objectives underlying the promulgation of existing laws or institutional and adrmnistrative arrangements for their enforcement. The provincial Boiler A c t 1923, for example, requires Grade-11 Boiler Inspectors recruited and trained to enforce a legislation that may have been relevant more than 80 years ago, and are today expected to inspect and certify boilers manufactured by multinationals. 16 One study of the surgcal instruments sector, Sector Strategy Report-StlrgicdInstwments, SMEDA, 1999/2000, suggests that only 17 percentof the firms operatinginthe sector were registeredwith the Labor Department. 17 Payable at 2 percent of taxable income and collected by the Income Tax authorities for transfer to the Workers Welfare Board. 18 Payable at 5 percent of profits and distributedamong the workers of the company up to a maximum of Rs. 3,000 per worker of the enterprise, with the undistributedbalancebeingdepositedin the Worker's Welfare Fund. 34 This regulatory function can easily be outsourced to universities and private firms providmg engineering services that could be pre-qualified o n the basis of well-drafted selection criteria. 2.42. S d a r l y , the government has Building and Electricity Inspectors to ensure the safety and security of private buildmgs used for public purposes, for example, cinema houses and restaurants. Objectives underlying the regulatory functions performed by these functionaries can be better achieved if such buildings are comprehensively insured through private insurance companies. Even inPalustan's economy, where the entire range of services on offer are stdl in the evolutionary stages of development, the private sector is able to provide products and services to supplant/replace the present role of the government's regulatory apparatus. Through this instrument the cinema owners can be spared the frequent visits of these government employees, who would be denied the opportunity for extortion on the basis of regulatory functions mandated to them. Moreover, the security and safety of the public using these buildings would also be assured, since the private insurance companies would ensure the proper construction and maintenance of the property. 2.3 STIMULATING FASTER GROWTHINTHE CONSTRUCTIONSECTOR 2.43. Construction has t h e potential to be one of the most dynamic sectors of Pahstan's economy.'g However, the shortage of housing units in the country i s estimated to stdl be around 5 rrdion units, indicating that this sector has considerable untapped growth potential. This i s especially so in the Punjab where the number of housingunits has increased by 2 percent per annum between 1980 and 1998, as compared with 3 percent p.a. for other provinces. With population growing at 2.5 percent per annum. (2.6 percent in other provinces), this sluggishgrowth in number of housingunits has led to increased crowding of houses in the Punjab (while in other provinces number of persons per housing unit have declined), despite somewhat faster growth in income. 2.44. Total national investment in housing, at around Rs. 150-160 billion per annum, contributes less than one percent of GDP.This i s due in large part to the relatively undeveloped state of housing finance. Historically, the state-owned Housing Buildmg Finance Corporation has provided the bulk of housingfinance. However, even at its peak, annual credit by formal financial institutions has never exceeded 1.5 percent of total housinginvestment, compared to around 25-50 percent in South East Asia and the United States respectively.20 A culture of d f u l default along with procedural complications encountered by financial institutions in encashing collateral, particularly of immovable property, have deterred the private sector from venturing into housing finance activities. While high interest rates throughout the 1990s have also restrained the demand for credt, prevahng interest rates are now at their lowest levels ever. 2.45. While addressing the biggest problem constraining the development of the housing finance sector-the dfficulties in enforcing foreclosure laws in the event of a default-is largely outside the purview of provincial governments, the Federal Govemment has recently taken several important steps in this regard. The improved performance of Banking Tribunals, along with enactment of the 2001 Finance Institutions (Recovery of Finances) Ordinance provide some hope that the challenges of foreclosure have been addressed effectively through strengthening of the legal framework and the 19Pakistan's PRSP notes that the employment elasticity of this sector with respect to GDP growth is more than twice that for the economy as a whole (overall GDP-employmentelasticity: 0.41; construction sector 0.87). 20Presentation by ConradD'Souza of the HousingDevelopment FinanceCorporationLimited, India at a HousingFinance Seminar in Karachi,August 2003. 35 debt recovery procedures.21 The eventual development of a secondary market in housing finance, leading to the introduction of tradable mortgage-backed instruments whereby primary lenders can securitize their mortgage holdings for re-sale to other investors, will give an additional fillip to this sector. 2.46. The Provincial Government can also play an important role in addressing a number of other key legal, fiscal, and administrative barriers hindering faster development of the housing and construction sector. These include: Unreliable Records of Rights due to Unclear L a n d Titles: T h e lack of adequate land information, access, and retrieval and verification system results in poor quality, reliabhty, and accuracy of the records of ownership rights.This dysfunctional nexus between land management and housingmarkets is a major impediment to the development of more efficient land markets. High Rates of StamD Dutv and other Levies: In urban areas of the Punjab, the costs of registering property transfers include a 1 percent registration fee, a 5 percent provincial stamp duty on the value of the property (compared to a nominal transfer fee of Rs.150per sq. yard and no stamp duty in Islamabad, and 3 percent stamp duty in Sindh), not to mention numerous other charges (professional charges, brokerage fees, court fees, etc). Outdated and Non-uniformBuilding and Zoning Laws: Multiple agencies administer different zoning and buildingby-laws (e.g. those relating to heights of buildings, ratio of floor to plot area, etc.) and building approval criteria within a city. For instance, in the case of Lahore, these include the LDX, Cantonment Board, and Tehsil MunicipalAdministrations. These zoning and buildmgby-laws need to be updated and made consistent across authorities within the province. Excessive L a n d Redation-Commercialization Charpes: Some local government taxes and fees raise the cost of investment for the construction sector. For instance, in the case of change of use of property, LDA levies a commercialization fee o f 20 percent of the value of the plot, even in areas that have already been declared as commercial zones. Lack of Adeauate Penalties for Non-Uthzation of Land: T h e lack of adequate penalties to discourage speculative acquisition of property means that even when all associated infrastructure has beenprovided, colonization of housing schemes proceeds at a slow pace. Not only is the investment on roads, electricity, gas, water supply and sanitation systems, etc. under-uthzed as a result, but also the assets created deteriorate because of long periods of non-use. To discourage speculation and release land for development, the government should consider raising the non-uthzation fees on vacant plots along with appropriate measures to address the issue of benaami holding of property (thereby also minimizing the problem of unclear titles).22 21 The Ordinance dlows financial institutions to repossess property without recourse to courts, and has also been successfully tested in the courts in two recent cases. Nevertheless, foreclosure may continue to be an issue in the case of home lending, since social and cultural values would also come into play, which would perceive financial institutionsas forcingdefenseless poor householdsto surrendertheir only shelter. 22 Benaamiliterally means "no name"; a benaamz transaction means any transaction in which property i s transferred to one person for a consideration paid or provided by another. Benaumi holding of property allows individuals to obfuscate the identity of the true owner of the property,or to undermineeffective implementationof landholdingceilings. 36 Rent Restriction Legslation and Disincentives to Develotinc Rental ProDertv: The excessive pro-tenant bias of the 1959 Urban Rent Restriction Ordinance, under which a Controller has the sole authority to determine the fair rent for residential premises as well as when eviction of tenants i s justified, discourages development of properties for rental purposes. Other major factors that serve as disincentives for renting out property include the high rate of stamp duty and registration fees (3 percent and 1 percent of value of contract respectively), and that property tax on rental property i s assessed at ten times the rate o n comparable owner-occupied properties, even though both may have the same access to public facilities and amenities [the current property tax differential i s much higher than in Karachi (2:l) or Islamabad (same rate applied)]. 2.47. Any potential loss of revenue from the reduction in property tax rate of rented properties is likely to be more than recovered from a reduction in evasion of property tax on rented properties, increased development of properties for rental purposes, and a slight revision in the tax of owner- occupied properties. In addition, disposal of land owned by government in prime commercial locations and being used for less productive purposes (or leased out to the private sector at nominal rents), will help better exploit the potential of this scarce asset, and thereby stimulate private construction activity, improve land utdlzation, and m o b h e revenues for all levels of government. RecommendationsforActions Within the Provincial Government's Control 2.48. T h e Provincial Government can help reduce the cost of investment in construction, and improve incentives for construction of property for rental purposes through a number of measures: 9 Rationalizethestamp dutyonpropertyrelatedtransactions 9 Prune change of use charges for property classified as ``commercialy'under zoningregulations 9 Rationalizedevelopmentchargesimposedby local waterand sanitation agencies for change ofuse ofproperty from `kesidential" to `%ommercial" 9 Imposealandnon-utilizationfee oranidleland tax 9 Promulgatealawabolishingthebenaamiholdingofproperty 9 Revamp theRentRestriction Ordinance. 9 Narrow the differentialinpropertytaxonrenterandowner-occupiedproperty 2.4 IMPROVINGTHE SYSTEM OF LAND-TITLING INTHEPUNJAB 2.49. L a n d i s a critical economic asset in any economy. A modern land admmistration system, with a secure, accurate, and accessible system of registering and recording land transactions, i s widely accepted as an important mechanism for empowering the poor. 2.50. Current Svstem of Record-of-Richts and Related Lesslation: T h e records-of-rights in land in Pakistan are of the fiscal variety. T h e person shown on the records i s responsible for paying land revenue or property tax and i s presumed to be the owner, unless it can be proved otherwise. However, the title to land i s only incidental, as Pakistani law does not adrmt to the provision of a 37 certificate from the government guaranteeing that the person mentioned in the records-of-rights i s the true owner. Inother words, entries in the records-of-rights can be challengedin the courts of law, as the registrar entering the transaction in official records only confirms the validlty and accuracy of the document, but does not provide any legally bindingassurance of title to the property. 2.51. Except for cases specifically provided for in the Registration A c t or the Transfer of Property Act, registration of land i s optional. For instance, under Islamic Law, oral declarations of gift (even if subsequently recorded on paper) do not have to be registered. Thus, the system of law allows for perfectly valid titles (Le. in a legal sense) to be created without the transaction being recorded anywhere. Any search by a buyer in the Revenue Record sand the Registrar's Office will not provide any clues to such a transaction having taken place. On the other hand, courts have repeatedly confirmed that registered documents do not guarantee title, but rather are open to scrutiny and possible rejection in courts. 2.52. The established system of conveyancing visualizes that the buyer must investigate the seller's title to the property-"let the buyer beware." While the transferor i s obliged to g v e a good title to the transferee, the government i s under no obligation to provide a guarantee to title, i.e., it does not furnish any assurance on ownership. Despite the exercise of due caution the title may still be defective. For instance, while the buyer may have satisfied himself of the authenticity of the transactions leading to the present transaction, it i s just possible that some of the documents of earlier transactions were defective on account o f forgery, lack o f consent, consent granted by a minor and hence not valid, etc. 2.53. Benefits o f State Guarantees of Title to ProDertv: Developing a system of guaranteeing land titles would not only confer great benefits to those h e c t l y involved in land transactions, but would likely also generate considerable economic externalities. It would help free up precious court-time presently engaged in settling title disputes, enable implementation of land reforms, facditate credit availabilitv to small farmers who could use secure titles as collateral. boost tax revenues bv reducing the scope for evasion of income tax, and Box 1:LandTitling Systemsin Other would also help reduce cost and time Countries overruns in development projects involving The Torrens System in Australia: Under this system, land acquisition. the state certifies the authenticity of title. Apart from the main advantage that it confers on the o w n e r a definitive title to t h e 2.54. The importance of modern land, this system greatly simplifies conveyancing procedures, technology and the introduction of a GIS circumvents the need for costly examination of records, a n d helps avoid problems created by genuine errors a n d mistakes are universally recognized as a way of in the past. Experience in several Australian states has shown enabling quick and efficient recordmg of that w h e n the central government created a database of all parcels of land a n d their respective owners they were able to title transfers. However, for the effective detect m a n y m o r e legitimate taxpayers. O v e r 50 countries, use of new technologies greater including Kenya, Uganda, Sudan, Morocco, T u n i s a n d Syria, consideration d have to be given to have now adopted the Torrens system. Singapore has also switched to a system of registration of titles rather than deeds. institutional roles and responsibdities, The English System: The L a n d Transfer A c t in development of human resources, work England reyuires a compulsory registration of title to land. procedures and processes and appropriate, T h e registered holder is regarded as the sole proprietor. T h e sustainable technology that has a well- difference of the English system from the Torrens system i s that, even though the former i s m u c h simpler, titles can b e defined development path. While setting up corrected in the event of fraud under t h e latter. In the Torrens the system may initially b e quite costly, for system the courts have limited, if any, jurisdiction over the records of titles. 38 instance on account of indemnification losses, these costs would most likely be recovered through increased tax revenues. In addition, supplementing the initial fund with registration charges, once transactions start getting recorded can minimize budgetary outlays. Recommendations 2.55. While comprehensive cadastral surveys and registration of titles are admittedly costly affairs and difficult to undertake, they likely offer the only lasting solution to the problems discussed above. Those profiting from the present system d oppose an up-to-date, more accessible, system for recordmg land related information, so preventing reform efforts from being stifled by these vested interests drequire commitment and support from the highest level of government. 2.56. The government could initiate pilot projects in two districts of the country. These projects will enable the identification of the problems and difficulties likely to be encountered in setting up a system of registration of titles. As a first step in the long term strategy to design and implement a system of title registration there could be a requirement for the compulsory registration of all documents which relate to property, including sale agreements, declaration of gifts, powers o f attorney. Any party claiming title through adverse possession would be required, within six months of acquiring such title, to register his claim. Bemzmi transactions would have to be declared unlawful and the courts would have to refuse to recognize these. Any suit against any immovable property would also have to be registered with the Registrar. Provision exists in the Registration Act for such registration. This would have to be made compulsory. Simultaneously, the government could initiate the process of conversion of presumptive titles into exclusive titles. Draft lists could be prepared, that would be open to public inspection for a period o f G months or some such stipulated period during which any disputes or objections that are raised could be would be settled. While the process may be slow and painful, the long-term economic and social benefits that d result from establishment of a modern system of land tiding are likely to be immense. 2.5 PROMOTINGTHE DEVELOPMENT OF CITIES AS GROWTHENGINES 2.57. A recent UN study estimates that by 2007, more than half the world's population dlive in urban areas. Most of t h i s growth i s projected to come from smaller urban centers that currently have fewer than half d o n residents. Governments in some developing countries are increasingly focusing on development of cities-both smaller urban centers and mega-cities, with recognition that these areas contribute sipficantly to GDP and employment, and thus provide the best opportunity-to generate new employment opportunities and serve as growth engines for the economy. For instance, China now encourages rural urban migration, having realized that this has had a positive effect on the economy. To support these cities in generating economic benefits for the country, governments-national, sub-national, and local-undertake a series of actions aimed at supporting the city's comparative and competitive advantages, ensuring that they provide a sound local investment climate to businesses, and ensuring that it provides adequate and needed infrastructure at appropriate locations. The approach i s demand based, responding to business needs and i s premised o n partnership between the public and private sector. 2.58. With an overall population density more than 3% times that of the rest of Pakistan, the Punjab i s uniquely placed to locate cities at the center of i t s economic and development strategy. In addition to the don-plus cities of Lahore, Faisalabad, Rawalpindi, Multan, and Gujranwala and their adjoining areas, it has 10 other cities with over two hundred thousand residents, which could 39 potentially serve as mini-growth engines for the regional and provincial economy. Focus on these smaller urban centers in addition to the mega-cities provides greater scope to exploit local resource bases and comparative advantage, leading to the development of a more diversified and prosperous economic base overall. T h e example of Sialkot (Box 2) offers an excellent model to spur other cities to pursue s d a r objectives through innovative schemes and proposals relevant to their regions and specializations. 2.59. The overall approach would be to (i) ensure that the investment climate of the city / city region i s conducive to business needs (this could, for instance, mean ensuring well functioning infrastructure, addressing excessive regulation, ensuring good govemance), and (ii) focusing on the needs of key industrial sectors of the economy. This in turn would entail designing and implementing a collaborative work-plan of more focused and in-depth studies to examine the growth potential o f each region based on the full spectrum of potential private sector issues,23 and cultivating and pursuing successful public-private partnerships to help realize this latent potential, as government financing alone wdl be insufficient to meet the increased investment needs of these regions. 2.60. Such "competitiveness partnership" initiatives, consultative mechanisms, or public-private platforms for reform should ideally be initially launched in 1-2 pilot cities, selected based on displaying the strongest potential with respect to 4 dimensions: (1) an organized private sector with some entrepreneurs witling and able to provide leadership; (2) strong government support with political will (including support from provincial officials dealing with matters like taxation, business registration, trunk infrastructure, water supply and sanitation, roads, and other such key local infrastructure); (3) a sponsor with strong credibility with both private sector and government; and (4) instruments such as logistical facilities and seed funds for starting the initiative untd further funding becomes available. 2.61. The initial focus should be on policy and regulatory issues that affect the entire private sector, where broad consensus for reform exists, and where quick and visible wins are feasible, complemented by longer-term initiatives such as enhanced support for enterprise-based training. A carefully implemented media and communications strategy to enhance accountabdtty of group members to broad social interests should buttress a well-defined outreach strategy in support of an understandable and achievable vision statement. 23 K e y studies that are likely to b e required include:(i)development ofbaselineinformation; (ii) ofthe economic analysis potential of the local areas a n d priorities for government support; (iii) analysis of business needs within the area; a n d (iv) a thorough review of the economic services departments and the services they provide. Based on these studies a prioritization could b e undertaken (i) of the economic services that are required; and (ii) business investment climate needs. of 40 I Box 2: The Cluster Approach to RegionalDevelopment and Growth: The Example of Sialkot A key center for labor-intensive small-scale industrialenterprises, the Sialkot regional cluster is poised for accelerated growth. A relativelysmall city with an estimatedpopulationof approximatelyhalf a million, Sialkotis nonethelessinternationally renownedfor its export-intensive manufacturing sectors, in particular its three relatively well-developed industrial clusters producing surgical instruments, sports goods, and leathergarments. Annual exports from the regionare now close to $700 d o n , contributingnearly 7% of Pakistan's exports.The businesscommunity in Sialkot has successfully pursued a number of initiativesover the years: Sambrial Drv Port: As early as 1986, city residents collaborated to set up the Sialkot Dry Port Trust, the fust ever dry-port establishedin the private sector in Asia. The mainobjective of this venture was to providecustoms clearance through a one-window operationat the city's doorstep, thereby facilitatingparticipationby exporters and importers.The number of exporters subsequently increased substantially, enabling total exports to rise rapidly. The total annual value of exports from the port has increasedmore than twenty-foldsince its inception. CitvPackage:Since excessivetraveltime and inconveniencewas an importantfactor inhibitingpotentialclientsandvisitors, the - private sector raised funds to build better approach roads to the city. An innovative financing plan was agreed with the Punjab Government, whereby the Chamber of Commerce raised Rs. 130 million through deducting 0.25 percent of export earnings at source. While the government initially agreed to contribute Rs. 3 for every rupee raised, it eventually withdrew from the scheme after contributingRs.lOOmillion. Several key roads were nonethelessreconstructedand expandedthroughthe scheme, and drainage work undertaken, makinga dramatic differenceto the imageof Sialkot. Other self-helnschemes and nartnershins:Roadimprovementwork was undertakenin the Sialkot IndustrialEstatein 2001/02 through a Rs. 5 million self-help scheme, which in turn was matchedby a grant from the District Nazim and a contributionof Rs. 10 million from the city package in 2001-02. Similarly, when residents realized that the unplanned and uncontrolledgrowth of tanneries in populatedareas of the city posed serious a public healthrisk, the business community and the ProvincialEnvironment ProtectionAgency set up the Sialkot Tanneries Zone, to whichall tanneries would be relocated. Sialkot International AirDort initiative: The private sector's most ambitious undertaking to-date has been to initiate construction of an internationalairport on a build-own-operate basis. Members of Sialkot InternationalAirport Limited (SIAL) contributed 5 million each to raise a total of Rs. 800 d o n . In 2004, Rs. 200 million is expected to be raised through soliciting smaller contributions from 200 additional members. Landacquisition for the project was financedthrough funds advancedfrom the governmentat favorable terms.While initial estimates of the cost of the project were Rs 1.7 billion, this could rise to aroundRs 2.5 billion. SIAL is presently examining several options to meet the shortfall in funds through additional equity financing. In the meantime, constructionwork on a 3.6 kmrunway, equivalentin specificationto that of LahoreInternationalAirport, commencedin January 2003, and is expected to be completebyJuly2004allowingcargo flights to commence.Passenger operationsare expected to commenceinApril 2005. Addressing Impediments to Growth: While the business community in Sialkot has undertaken a number of commendable initiatives in the past few years, it alone cannot tacMe all the bottlenecks constraining growth. By joining hands with the private sector in the followingareas, the provincialgovernment canwork towards accelerated future development of this promisingregion: P Devefop human capital:Worlunginpartnershipwith the private sector, organizations such as the TechnicalandVocational Training Authority (TEVTA) and the Punjab Vocational Training Council (PVTC) can play an important role in stimulating employmentgenerationin this regionby designingtrainingcoursesthat develop skilledmanpower demandedby industry. P Roads infrastructure: Additional, relatively modest, investments in infrastructurecan play a key role in reducing the high logistical cost of doing business, and help improve, and not merely expand, the competitiveness of this growth cluster. The private sector has identified a number of schemes where investment by the provincial government to improve access to all the major cities in the region, and from the new airport canprovideswift returns. > Shared service facifities: Better connectivityin turn will allow sharing of common service facilities (e.g. training facilities, common bonded warehouses for raw materials, testing facilities, etc.) by industries across the region. A case in point is heat treatment and materials testing facilities, where the rudimentary public sector facilities that existed in the 1980s were found inadequate, and their use largely abandoned.These facilities need upgradmg, or better still, the government can help set up a modern faciltty in the privatesector. k Administrah'veprocedures:Followingthe recentand ongoinglocalgovernmentreforms, some administrativeconfusionhas arisen over the functional mandates of provincial and local government for maintenance of infrastructure.This needs to be addressed urgently, as it is perceived as a growing constraint to progress in improving access roads. The government can also focus its efforts on deregulationof systems andproceduresto reduce the cost of business. Source:For more detds, please see backgroundstudy: TheClusterApproachto RegionalDevelopmentand Gmwth: TheCaseofSialkot 41 2.6 TOWARDS BETTERSTRATEGIESTO ACCELERATEGROWTHOFTHE SERVICESSECTOR 2.62. As noted earlier, the non-farm sector accounts for the bulk of provincial GDP in the Punjab. In particular, the services sub-sector-comprising wholesale and retail trade, transport, storage, and communication, and public administration and defense-contributes about one-half of provincial GDP, and employs about one-third of the total labor force. This chapter has included a number o f recommendations on reducing unnecessary regulations which increase the cost of doing business, and hence act as a major disincentive to investment and private sector growth. In addtion, using survey data on perceptions of the investment climate by firms in both the formal and informal sectors, the report has highlighted how access to finance, infrastructure (electricity in particular), and availabhty o f slulled workers act as bindmgconstraints for firms operating in hfferent sectors. 2.63. Looking ahead, clearly more in-depth work i s needed-moving beyond t h i s fairly macro- analysis of the over-arching constraints to faster growth-to analyze specific policy actions and interventions that could help spur faster growth o f selected sub-sectors of the provincial economy. This is true not just of those selected sub-sectors contributing a high share of total GDP and employment (e.g. light engmeering, poultry, d a q , etc.), but also those that may b e relatively small at present, but nonetheless have considerable latent growth and employment potential (e.g. information technology, education and health services, IT-enabled services, etc.). In this regard, while manufacturing and construction-related sub-sectors have been the subject of a number of recent studies,24 very little attention has been drected to-date towards better-understandmg constraints faced by firms in various services-related sub-sectors. In thinkingabout a future program of work to help operationalize the provincial government's development vision, there i s an urgent need to strengthen the information base in this area. 24 For instance, see the recent studies b y SMEDA and other institutions analyzing the growth potential of various industrial sub-sectors, such as-for instance-poultry, dairy, electric fans, surgical goods, automobile parts, electronics, molds and dies, sugar, steel, marine fisheries, leather, ceramics, handlooms and other such texule-related sectors, etc. 42 CHAPTER 3: ACCELERATING GROWTH INTHE AGRICULTURE SECTOR 3.1. Agriculture i s a major contributor to the Punjab's economy, accounting for roughly 28 percent of its output and providing employment to over 40 percent of the work force. The performance of the sector i s therefore crucial to any strategy for generating higher incomes. While the agriculture sector in the Punjab has witnessed negligible growth in total employment during the 1990s (ref. Table 2.1), it nonetheless has considerable latent employment-generation potential. Examining data from several South and East Asian countries, a recent study concludes that converting one hectare of rice to vegetable cultivation for one season generates one year-round job (Ali and Abedullah 2002), indicating that increased agricultural &versification away from cereal crops to higher value and more labor-intensive activities-a central thrust o f this chapter-presents tremendous job-creation possibhties. This chapter reviews the recent performance of the agriculture sector in the Punjab, and identifies priorities for public policy to accelerate agricultural growth, enhance the sustainable use of natural resources, and reduce rural poverty. 3.1 RECENT SECTOR PERFORMANCEAND PRIORITIESFOR PUBLICPOLICY 3.2. Over the past 20 years, the agricultural sector has performed moderately well, growing at around 4-5 percent per Average Annual Growth (1991-2003) annum (higher than the average for Pakistan, South Asia, as well as low-income countries overall). Since the Green Revolution in the 1960s, agricultural growth has largely come from adoption of modern inputs, especially high yielding varieties, ferthzer and pesticides, combined with expansion of irrigated water supplies. The adoption of t h i s technology has reached high levels, with 80 percent of land irrigated, over 90 1.0% 2.0% 4.0% 6.0% percent of the area of the major crops under modern varieties Shares in Total Agriculture Value (sugar cane i s the exception), and ferthzer consumption of Added (2002103) over 100 kg/ha. Between 1970-2003, input growth accounted for one-half of growth in output, while total factor productivity (TFP) growth accounted for the remaining half, with technical change, infrastructure, and education being the major drivers of TFP growth (Ali and Byerlee, 2004). Loohng ahead, important considerations related to agriculture include: 0Livstk. BMin. Crp OForest. OMaj. Crp. mFishery 0 The Green Revolution strategy of input intensification in a few major crops, pursued since the 1960s has to a large extent been exhausted with h t e d prospects of increasing land area and water supplies, and rapidly diminishing returns, and in some cases serious environment and health implications to higher use of external inputs, especially pesticides (Box 3). 0 Productivity of the sector i s low in terms of efficiency of use of resources and inputs, and the rate o f TFP growth at 1.5 percent has been lower than in other countries and regions experiencing rapid technological transformation, such as northwest Inda, China, and Brazil.25 25Comparable estimates of TFP growth are 1.9 percent for the Indian Punjab 1970-95, 2.2 percent for China 1965-85, 1.9 percent in Taiwan 1951-80, and 1.7 percent in Brazil in 1968-87. 43 Water has become the major h t i n g factor to further growth yet the productivity of water used inagriculture has declined in the past three decades. Land has been degraded by the use o f poor quality water and continuous cereal mono-cropping. Together, these factors have held back the growth o f productivity-it i s estimated that in the absence of resource degradation, productivity growth in the Punjab would have been 42 percent higher than the actual growth during 1970-1990s. Agricultural exports have largely been provided by cotton and rice, and only recently has the Punjab begun to exploit the rapidly growing trade for high-value agricultural exports, despite a favorable ecology and geographic position. Growing income inequality and stagnating poverty levels have accompanied low agricultural growth. Part of the problem has been a high and growing concentration of land with 2.5 percent of large farmers (> 20 ha) owning one third o f the land while small farmers with less than 2 ha o w n only 12 percent of the land although they account for half of the total farm population. Box 3: Environmental and Health Problems ofAgricultural Intensification The most serious agriculturalpollution problemrelates to the rapid increasein pesticide usefrom less than 1000 tin 1980 to 70,000 t in 2002-most of which is used in the Punjab.Most pesticide i s insecticideand most of this i s applied to the cotton crop. The widespread use of generic and often dangerouspesticides on cotton has a number of potential health hazards, includingcontamination of workers who apply it (three quarters of producersuse a back-packsprayer without protective clothing), contamination of harvesters (allof whom are women), health hazards to plant workers, contaminationof soil and groundwater used for drinking, and contamination of consumers of agricultural products. Economic costs of these negative healtheffects have beenestimatedat over one billion Rupees annually. The sole reliance on pesticides to reduce pest losses has aggravatedsustainability problems as pest populations change and develop resistanceto commonly usedpesticides. This was manifestedin the recurringcotton crises of the past few years where pest losses sharply reduced the cottonharvest, threateningsupplies to the domestic textileindustry, causing estimatedlosses of Rupees 5.7 billion annually. The short run success in increasingcottonproductionover the past two decades has been translated into a major challenge to sustaining the industry. Despite the magnitude of the environmentaland healthproblemsassociatedwith highlevels of pesticide use, Pakistanis a latecomer to the practice of integratedpest management (IPM) and has beenreactiverather than proactivein fightingpests. 3.3. Because of its role in output, incomes, and employment, it i s clear that agriculture will be a priority sector in any strategy that the Punjab adopts to accelerate economic growth and to reduce poverty. However, it i s also clear that past approaches wdl not work for the future. 3.4. First,markets for agricultural products are rapidly changing. Not only does the sector need to respond to rapidly changing domestic markets with urbanization and m o r e affluent consumers, but also the Punjab with i t s irrigated agriculture, favorable c h a t e and geographic position, has excellent prospects to expand exports into high value and value-added products. With self-sufficiency inwheat and sugar, the growth strategy must shift increasingly towards fruits and vegetables, value-added products, and livestock products, and to promoting the export orientation of these products. 3.5. Second, with ever more h t e d prospects for expandmg land area and water supplies, growth strategies must increasingly rely on improved management and knowledge to increase the productivity o f existing resources. Technological progress and institutional reform i s central to better system-wide and location-specific management for enhanced input and resource use efficiency. Enhanced efficiency i s also essential for improving Pakistan's competitive position in world markets as it adjusts to WTO rules. 44 Box 4: Crop Yields inthe PakistaniPunjab,the IndianPunjab, and Egypt Despite the achievements noted, yields of most crops in Punjab are lower than international comparators (see below). 'Wheat and rice yields in Pakistan Punjab are less than half of what has been achieved in the comparable environment of Indian Punjab and Egypt. This not only suggests lack of appropriate technologies and inefficiency in using available technologies, but also indicates the enormous potentlal,which can be exploitedwithout many additionalresources. Tons / hectare (Data for 2002) Crop PakistaniPunjab IndianPunjab Egypt Wheat 2 5 4 5 6 2 Rm-fine (paddy) 2 5 Re-coarse (paddy) 3 3 5 3 9.4 Cotton @hum) 1 8 1 3 2 8 Sugarcane 45 1 65 2 90 9 Maze 2 1 2 7 7 7 Gram/chck peas 0 7 0 9 1.9 Rapeseed and mustard 1 0 1 2 Tomato 100 35 1 Potato 17 2 20 0 23 8 3.6. Third, sustainable agricultural growth depends on improved management of land and water resources to reduce natural resource degradation. This d require political commitment at the highest level to mount a concerted effort across a wide range of government departments, as well as close coordination with federal agencies and local governments on a priority basis. 3.7. Finally, a major challenge i s to ensure that the poor, both small farmers and the landless, participate in future growth. T h e move towards high-value products (which tend also to be more labor intensive) i s one way of generating employment. Targeting public resources on poor regions, such as southern Punjab, and poorer farmers, can also help towards this end. 3.8. Interms of specific commodities, these challenges suggest that the Punjab: Continue to ensure a moderate growth in wheat production, but recognize that under current yields and world prices it does not have a comparative advantage in wheat exports Improve competitiveness of rice and cotton, especially in view of the aggressive efforts by other countries in these markets. This applies especially to I n d a in Basmati rice, and to most major producers in cotton, that have sharply reduced their production costs by adopting Bt cotton. Judiciously seek opportunities in import substitution, especially oilseeds for more sustainable cropping systems, but only where they have a clear competitive advantage. Diversify production toward higher value horticultural and livestock products to meet rapidly expanding domestic demand and generate employment, and Exploit i t s excellent resource base and geographic advantage to tap into the rapidly expanding markets for high value products in Asia, the Middle East, and other highincome countries. The Government of the Punjab, especially the Department of Agriculture has already initiated many reforms to modernize public services to the future needs of the sector. T h e efforts to develop partnerships with the private sector in agricultural marketing and in promotion of conservation idlage, are especially commendable. 45 3.2 RAISINGTHE LEVEL EFFICACY PUBLICEXPENDITURES AND OF 3.9. Allocation o f public expenditures to and within agriculture i s one of the major ways that the government can influence growth and employment in the sector. Unfortunately, from the early 1990s to 2002, real total expenditures on agriculture and irrigation declined by 4.4 percent annually (Table 3.1). In 2002, total agriculture and irrigation expenditures combined accounted for only 2.1 percent of agriculture GDP, falling from 4.5 percent in 1991. These figures are very small in comparison to other developing countries and regions (Figure 3.1). For instance, in 1998, agriculture expenditures as a percentage of agriculture GDP was only about one quarter of the Asian average of 8 percent, and lower even than Nepal. Also contrasting to the steady decline in the Punjab, agricultural expenditures indeveloping countries as a whole grew at 3 percent during the 1990s (Fan and Rao, 2003). Figure3.1: Public Expenditures on Agriculture:Punjabvs. Other Countries:2000 14 12 'ij cln 10 e n 8 6 n O a 4 2 0 3.10. Irrigation expenditures have declined more rapidly than for agriculture. In fact more funds are spent on wheat subsidies to consumers than on crop, livestock, forestry and fisheries combined, for producers. The federal contribution to agriculture and irrigation has declined most rapidly, and made up only an estimated 8 percent of spending in the Punjab in 2001, before rising again in 2002. The largest decline in spending has been for development expendtures, which are now less than half their level in 1991 (Table 3.1). Development expenditures are especially important in the irrigation sector to ensure optimal exploitation of water resources. Most agricultural expenditures are allocated for non-developmental activities, mainly for establishment charges such as salaries and wages. The Punjab Government has made progress in recent years in increasing the ratio of operating costs to salaries. For example, the ratio in crop research has increased from 18 percent in 1996-97 to 24 percent in 2003-04, although s t i l l below the 30 percent ratio recommended for research. However, declining resources for O&M remain a problem in irrigation. 3.11. The foregoing trends suggest that, despite the rhetoric, the status of agriculture has declined within government priorities. Although the Government of the Punjab has taken a number of initiatives in recent years to use public resources in agriculture more efficiently, it still has major opportunities for reallocating expenditures, privatizing, and devolution, which would enhance the impact on growth and poverty reduction. A number of recommendations are offered that would improve the efficacy of public expenditures and move the province closer towards the attainment of i t s goals. 46 Table 3.1: Agricultural Expenditures in ProvincialAgGDP: 1990/91- 2001/02 - /l L x x Agriculture IX Irrigation Total Agric Exp. as Agr. exp. Estimated Total Develop Total Develop Total Develop Yo of Punjab as Oh of Federal as O/o Rsm Rsm Rsm Rsm Rsm Rsm Budget AgGDPa of Agric Exp. 1990/91 5,414 1,123 8,535 2,634 13,949 3,757 15 1 4 5 15 1 1991/92 6,604 2,406 8,451 2,529 15,055 4,935 14 6 4 3 23 6 1992/93 6,403 1,927 6,700 1,449 13,103 3,376 14 0 3 9 17 2 1993/94 5,595 1,243 4,996 643 10,591 1,886 12 7 3 1 20 9 1994/95 5,635 1,218 6,304 1,348 11,939 2,566 12 3 3 2 17 9 1995/96 5,529 1,121 7,877 1,396 13,406 2,517 12 3 3 4 18 3 1996/97 4,697 737 6,446 729 11,143 1,466 13 0 2 9 13 5 1997/98 4,877 851 5,226 1,492 10,103 2,343 117 2 4 14 1 1998/99 5,003 699 4,795 1,308 9,798 2,007 111 2 2 13 8 1999/00 4,615 514 4,506 1,201 9,121 1,715 8 9 1 9 113 2000/0l 4,502 395 3,975 970 8,477 1,365 9 1 1 8 8 2 2001/02 4,842 467 5,017 1,114 9,859 1,581 9 8 2 1 16.9 Growth Rate -2,.8 -13.8 -5.7 -5.9 -4.4 -9.2 Y " ulture, 11 /ob R s Mhon a Includes I spendingand district spending. Source: Pakistan Public Expenditure Management (WorldBank, 2004) > Increase public investment, particularly on irrigation, water management, and agricultural research: the Punjab's long run objective must be to increase agricultural O&M and development spendmg on core public goods toward intemational levels. For example, public investment in agricultural research in the Province (includmg federal expenditures) i s 0.26 percent of agricultural GDP, compared with an Asian average of 0.5 percent and an average for industrialized countries of 2.6 percent. Likewise, development expendltures in the irrigation sector are especially important to ensure optimal exploitation of water resources, one of the most critical natural endowments of the province. > Privatize commercial activities, livestock breeding farms, and veterinaty services, and focus instead on core public goods: Given the scarcity of overall public resources available, the first priority of the provincial government should be to ensure that the public sector i s only fundmg those activities that are clearly public goods, and to devolve other extraneous activities to the private sector. For example, the provincial Livestock Department currently spends Rs. 340 millton (i.e. over half its budget) on i t s livestock breedmg farms, which i s clearly a private sector activity. The Department should develop an action plan to devolve t h i s activity to the private sector, and focus instead on activities such as pedgree registration systems, capacity buildmg and strengthening of private breeders associations to facditate the transition. Similarly, veterinary services provided at the district level should also be rapidly privatized, and the public sector should instead focus on control of livestock disease epidemics. > Reduce the share ofestablishment costs out ofcurrent expenditures as an important step towards rationahzing public spending: Although some progress has been made, the share of establishment costs out of current expendttures for agricultural support services i s dlsproportionably high. For instance, despite declining public investment on irrigation, the Irrigation Department continues to employ a very large staff, thereby using up significant provincial budgetary non- development resources. > Improve costrecovely, especially throughgreater costrecovety for canal maintenance in conjunction with devolution to water user associations: The public sector needs to institute better cost recovery and sharing schemes in the provision of public services. In the Department of 47 Irrigation, for example, cost recovery for canal maintenance sull averages only 25 percent, and there i s clearly a major opportunity to save fiscal resources and improve canal maintenance by full cost recovery in conjunction with devolution to Water User Associations. S d a r l y , field operations services in the Department of Agriculture (which account for one-third of its budget) should be carefully scrutinized for potential for cost recovery, and for privatization. 3.3 AGRICULTURALPRICESAND OUTPUT MARKETS 3.12. Price Policv and Wheat Subsidv: Historically, the Pakistan federal and provincial governments intervened heavily in product markets. The public sector dominated marketing for major crops (especially wheat) from procurement and storage through public corporations such as Pakistan Agricultural Supply and Services Corporation (PASSCO), and the provincial Food Departments. These policies also set output prices well below border-parity prices, with the result that Pakistan had one of the highest rates of "taxation" of the agricultural sector in the world.26 Since the Punjab i s the major surplus province, these policies had greatest impact on the agricultural sector of that province. 3.13. Over the past decade, the government has taken major steps to liberalize markets and encourage private sector participation, so that wheat and sugarcane are the only major commodities where there i s significant state intervention. The Government of Pakistan remains concerned about price stabilization for wheat and continues to announce a support price, which i s backed through public procurement. Price policy of wheat i s also a major concern of the Government of the Punjab, since wheat procurement and storage has significant fiscal implications for provincial expenditures, and in fact, the Government of the Punjab pays the bulk of the wheat subsidy. 3.14. T h e stated intention of the current policy i s to ensure food security, and address deficiencies in s t i l l developing private markets, and to do so in ways that do not involve tradmg losses. The move toward `cascading prices' where procurement and issue prices are increased monthly after harvest to reflect storage charges, allowed the private sector more space in the wheat market in 2002-03. But, government procurement of wheat (50-60 percent of marketed surplus) still far exceeds the amount required for the strategic reserve and subsidies for wheat procurement and distribution remain large. In addition, a consistent policy is lacking with the Government of the Punjab implementingin 2004 i t s own price premium, restricting wheat movement across provinces, and aggressively competing with the private sector for p r o c ~ r e m e n t . ~ ~ 3.15. In recent years wheat subsidies in the Punjab have exceeded total expenditures by the Department of Agriculture. Subsidies for domestic wheat, which are almost entirely borne by the Government of the Punjab, reflect the fact that the difference between the price of procurement and the issue price to d s (currently 30 Rs per 40 kg) i s only about one third of the total costs of transport, storage, and handhng by PASSCO and the Punjab Food Corporation (about Rs 2400 per ton, or nearly Rs 100per 40kg).2*Inpart this reflects inefficiencies inpublic management, rather than a subsidy to consumers, since margins for private sector wheat trading are significantly lower (Salam and Muktar, 2003). Also alarming i s the rapid increase in real costs per ton procured by the public 26 SchiffandValdes, 1992. 27 In fact, the Governmentof Punjab's recentdecisionto add a premiumto the announcedfederalprocurementprice,will have to be supportedby a direct fiscaloutlays estimated at Rs 700 m, without any clear rational. 28 Although PASSCO only receives a direct subsidy on imported wheat, it passes on full costs to the provincial food departments,which have to subsidize the differencebetweentotal costs and the dowed margin. 48 sector, with a doubling of costs incurred by PASSCO, and an even higher increase in the Punjab F o o d Department (Figure 3.2). Figure 3.2. RealTransport and HandlingCosts for Wheat in Punjab: 1996/97 -2002/03 3,000 2,500 0 2 9 0 c? 2.000 1.500 1,000 1996-97 1997-98 1998-99 1999-00 200041 200142 200243 9 Rationalizingthe wheatprocurement andstoragepolicy 3.16. There may be justification for interventions to manage price risks of wheat in Pakistan, or at least, to mitigate their effects. High instabhty in wheat prices may be undesirable for promoting optimal production decisions, especially in the absence of market-based risk management instruments. M o r e importantly, the poor are most vulnerable to sharp price fluctuations, since wheat constitutes a large share of small farmers' incomes and poor consumers' expenditures. 3.17. The federal and provincial governments need to define a consistent policy for interventions in wheat markets that clearly articulates objectives and strategies. T h e Government of the Punjab, since it pays the major cost of the wheat subsidy, should lead this effort. 3.18. T h e first issue i s to clearly establish the objectives and targets for intervention in wheat markets. These should explicitly lay out a framework for interventions that: 0 Defines a tolerable level of price variabhty 0 Minimizes fiscal outlays, especially risks of unexpected fiscal impacts 0 Minimizes dstortions to long-run market equhbrium prices, 0 Maximizes private sector participation and competitive markets 0 Ensures that the poorest are the major beneficiaries of any fiscal expenditures. 3.19. In defining these objectives, it should be noted that it i s m u c h easier and cheaper to remove extreme price variation and confine movements to within a `tolerable band', than to fully s t a b h e prices. I t i s also much cheaper and more equitable to target subsidies, if there are any, on the poorest groups that are most vulnerable to price fluctuations. 3.20. Second, the government should recognize that there are a number of potential instruments for realizing i t s objectives, each of which raises several questions in implementation. 49 T h e current strategy of setting of minimum support prices for producers and maximum release prices for consumers, enforced by government or government-contracted procurement. (How to set theses prices to minimize fiscal outlays, and maximize private sector roles? Use of cascading prices, and spatially determined prices). Holding of a strategic reserve (by whom: public sector or private sector, and what size? the potential to use a tendering system for procurement of the reserve through the private sector?) Provisions of incentives/subsidies to the private sector to hold reserves and invest in storage capacity (how to do this without distorting markets and ensuring minimum fiscal outlays?) Use of border price bands and variable tariffs to protect against extreme movements in world prices (what criteria to set price bands, h o w to comply with WTO rules?) Strengthening of commodity exchanges with forward contract markets (how to regulate and enforce contracts, issues of standardization of grades, use of warehouse receipt systems?) Operation in c o m m o d t y futures and options markets by both public and private traders (which markets and types of contacts, analytical capacity to determine optimal hedging strategies?) Strengthening of market information systems, and crop forecasting (potential to use satellite imagery and crop models to enhance precision o f pre-harvest crop estimates?) Provision of safety nets to the most vulnerable consumers and producers, such as food stamps, and vouchers for production inputs (how to target? How to avoid distorting markets?). 3.21. At this time, public policy in Palustan only uses the first two of these instruments, which severely limits i t s effectiveness in reaching the set of objectives outhned above. In particular, the government maintains a major role for the public sector in actual buying, storage and selling of wheat, rather than regulating and influencing the market in desirable directions for producer and consumer welfare. I t also leads to inefficiency (high marketing margins), opportunities for rent seeking and-most of all-a high fiscal outlay. 3.22. In the short term, the authorities should clarify their objectives for market intervention and critically evaluate two options to improve performance. First, they should explore competitive tendering of procurement and storage to the private sector in order to minimize costs and enhance transparency. Second, they should set the margin for procurement and issue prices to cover full costs, and develop mechanisms for targeting any subsidies to the poorest and most vulnerable groups. 3.23. Over the longer term, the government should explore a wider range of alternatives, especially in light of the apparent achievement of self-sufficiency, and the need to diversify the agricultural sector. This will require considerable technical skdls and information for developing an appropriate strategy for wheat price risk management. The Government of the Punjab in collaboration with the federal government needs to establish a high level task force to develop a new approach to stabilizing wheat prices within the above framework.29 This task force should be backed by appropriate analytical capacity, both in modeling wheat markets and in crop forecasting based on crop models and satelhe imagery. 29This task force should go well beyond the 2001 task force to clarify objectives, and explore a wider set of instruments as developedhere. 50 3.4 MODERNIZING WHOLESALE MARKETS 3.24. Wholesale markets are a critical link in the marketing channel, connecting rural production areas to urban consumption centers. They break up large lots into smaller units to meet the demands of various buyers, contribute to the regularization of supplies, help stabhze prices, and link to export demands. Improved market operations ultimately reduce costs, improve food quality and standards for domestic consumers, and help meet export requirements. 3.25. Public sector domination of wholesale markets: Wholesale markets are controlled and by departments through Market Committees set up at the district level. Currently fruit and vegetable markets are managed by Market Committees under the Agriculture Department, food grain markets are controlled b y the Food Department, and livestock markets and abattoirs are supervised by the Livestock Department. These departments set different rules and regulations that create confusion for producers and consumers. Recently, the Punjab government has opened a new Agricultural Marketing Department in order to streamhe the procedures across the various types of agricultural markets. 3.26. The administrative structure to regulate output markets in the Punjab i s highly bureaucratic and lacks effective private sector involvement. This generates inefficiencies and administrative anomalies. For example, in the case of fruits and vegetables, commission agents often charge a higher commission or ad hoc charges from both sellers and buyers than they are legally allowed. In addition, the expansion of agricultural market infrastructure i s insufficient to meet demand. T h e net revenues generated by the Market Committees are not sufficient to expand markets and modernize information system, in line with the expansion of agricultural output and changing requirement of the agricultural sector.30 This adds congestion to already very crowded markets. The current regulatory system assures the status quo and provides little incentive to improve food quality and implement food safety standards, and discourages private investment in market infrastructure. 3.27. Lack of market infrastructure: In part due to the above constraints, farmers are also not well connected with markets. About one third of farmers in the province are located more than 15 km from any fruit and vegetable market, and eight percent have to travel more than five km to reach a metalled road. T h e weak market infrastructure, congestion in markets, and poor post-harvest handllng, increases the internal transport costs and post-harvest losses and reduces the quality of output. A recent estimate suggests that an average of 25 percent of fruit and vegetable production is lost before it reaches consumers, not including loss in value due to deterioration in quality.31 Therefore marketing i s considered a major cost of doing business in the agricultural sector and an important impedunent to private sector investment in farming and agribusiness. T h e available market infrastructure i s also biased toward major crops, with inadequate attention to high-value crops and livestock. With changmg consumer demands, this strategy i s being revised to shift focus towards the development of market chains for high-value products. 3.28. Poor market information: T h e market information system i s poorly developed. Market Committees are responsible for the collection and dissemination of domestic prices for different commodities. But the collection of price data i s unscientific and the dissemination system i s primitive (e.g., use o f chalk boards to post prices). As a result market supplies usually do not match demand, 30 In2003, a totd of Rs. 260 miUionwas generatedagainst Rs. 200 mdlion salary and operatingcosts 31SMEDA 2000. 51 and do not reflect the quality demanded by consumers. Gluts and shortages in agricultural commodity supplies are common and discourage long-term investment in the agricultural sector. 3.29. The poor infrastructure and inadequate information system lead to uncoordinated markets, resulting in high price difference for the same commodity across markets. For example, monthly average potato prices in Lahore can &verge as much 162 percent from the p r e v a h g price in Hyderabad, a m u c h higher difference than can be explained by the transportation and handling costs. 3.30. M s s e d omortunities for diversification: The Punjab has been a late entry into export markets for high value agricultural products. Recently, the Federal Government has established the Pakistan Horticulture Development and Export Board to facilitate the hking of domestic production to international markets through market information and technology dissemination. Efforts are being made to introduce Pahstani specialty fruits and vegetables abroad through trade fairs. However, much more needs to be done in developing appropriate cold storage fachty in and around aqorts, and providingrefrigerated bogies in Pakistan Railways to transport fruits and vegetables from inland to ports inorder to maintain quality. 3.31. Grades and standards as a maior bottleneck: Although neither official nor accredited, grades and standards for agricultural products are often used informally in Pakistan. For example, traders, assemblers, wholesalers, and consumers usually grade wheat based on size of grain, presence of insect damage, soil admixture/staining and discolored or diseased grains. However, global and domestic forces have created a need for more formalized and universally accepted standards. To respond to these demands, the federal government has set up the Palustan Standards and Quality Control Authority (PSQCA) to defme output and input standards including for food products. To encourage standards in cotton production and processing, Pakistan Cotton Standards Institute (PCSI) was established in 2001. T h e institute has defined the acceptable standards for various grades of lint, although its role in implementing these standards remains controversial among various stakeholders. 3.32. These standards, however, are poorly defined or implemented leaving ample space for corruption as well as distribution of poor quality and unsafe food. For example, 50-63 percent of the milk distributed in the country is adulterated.32 With high pesticide application, pesticide residues on vegetables have crossed the tolerance h t for food safety. This situation not only results in hygienically poor food in domestic markets, but also restricts agricultural exports to low price foreign markets. For example, in 2002, Pahstan was the second largest exporter of mangos in terms o f volume, but in fifth place in terms of value. Likewise, the price received for Pakistan Basmati rice in international markets has been lower than for Indian Basmati. Weak grades and standards, lack of credible sanitary and phytosanitary (SPS) laboratories and trained manpower to test and implement food safety standards, and consumer ignorance are the main constraints in improvinggrades and standards. Reforming WholesaleMarkets: Priorities for the Provincial Government ` 3.33. The present agricultural marketing system is a remnant of an era when self-sufficiency in cereals was the primary aim. As a result of urbanization and changing export demands, the characteristics, roles and needs of different stakeholders in the marketing system are changing. Consumers, both domestic and foreign, are demanding greater emphasis not only on food quality, 32SMEDA 1999. 52 but also on food safety. Domestic and export markets can be compromised if food of inferior quality or of a hazardous nature is produced and distributed. The marketing system also has to respond to a change from a focus on food-security to &versification. 9 Reform market regulations: Implement the planned revision of the Market Committee Act to encourage the greater involvement of the private sector in market development. T h i s will bring new private investment to modernize agricultural markets. 9 Improve themarketinformation system:The Government ofThe Punjab shouldrevamp the district-level market information system, by upgrading the collection and dissemination of market prices on scientific basis, training the Market Committee staff in collection of price data, and developing an action plan to electronically link all markets in the province (Box 5). At the same time, the government should work with the Horticulture Export Board and the private sector to develop an information system on international prices, quality standards, port charges, trade procedures, and production and processing technologies of horticultural products. Box 5: "Silicon Mali" Mali's success in establishinga market informationsystem earned it the title of "SiliconMali" by Forbes Magazine in 2002. M&'s market information system (Observatoke des March& Agricoles) is based on enumerators visiting 58 markets around Maliand recordingthe high and low prices for grains, crops, and livestock. They enter these on laptop computers and e-mail the informationby FM radio waves all solar-poweredequipment to other regional offices where data are compiledand reports prepared for differenttypes of producers. The system built up over a decade has made Maliangrain farmers more efficient, knowingwhen and where to sell, and for what price. With better information,the governmentcan now rely on the privatesector to shift surplusesto areas with shortages without resortingto foreign aid. Mali's informationsystem has become a model for the rest of West Africa, where such countries as Niger and Burkina Faso are setting up similar systems that will be linked together.Soon, farmers will be able to do moresellingacross nationalboundaries. Source: Sansoni, S. 2002: "Sdicon Mali." Forbes Global, February 4. Cited in World Bank. 2004. Agriculture InvestmentSourcebook.Washington, D.C.:World Bank. 9 Develop critical infrastructure: Establish cold storage facilities at major airports, especially Lahore, initially through public investment (or better, seek partnerships with private exporters) and operated to recover full cost. Later they can be completely transferred to the private sector. Initial government investment on refrigerator bogies and cold storages at alrports will encourage the development of cold storage chain for high value crops in the private sector at the vdlage and town levels, and boost the quality of agricultural exports. 9 Establish a state of the art laboratory for food safety testing: The development of a safe food industry,which can respond to consumers demand within and outside the country, requires public investment in laboratories to test for virus contamination and pesticide and nitrate residues. This laboratory, which has already been approved can be established on full cost recovery basis. A priority dbe to develop capacity to efficiently use the laboratory to efficiently serve the private sector. 9 Pilot traceability and riskmitigation:Introduce the preventive approach known as HACCP - Hazard Analysis and Critical Control Point - to manage hazards associated with identified stages within the c o m m o d q chain, such as fruits and vegetables for export. HACCP is increasingly becoming a requirement for export in high value markets. To prove that good practices are 53 adopted in agricultural production, traceabdity of products from specific farmer's field through the entire market chain, has become a necessary element in export markets. Initially outlays for developing appropriate food safety management systems can be quite high. In Bangladesh, for example, introduction of HACCP for the shrimp industry costs the equivalent of nine percent of annual sales for start up and 1.2 percent for recurrent operating costs. With a dynamic export industry, initial outlays are soon recuperated, as it promotes the effective handling and use of raw materials, as well as prevents excessive losses o f valuable natural resources. In time, small food enterprises can increase their markets and gain a step up the ladder to lucrative export markets (Nicolaides et al. 1997). Build capacity in grades and standards:The most serious constraint on implementing food quality and safety standards i s human capacity. There i s an urgent need to invest in capacity building for food testing. Involvement of producers and exporters in implementing food safety management systems can help to share some infrastructure and capacity buildingcosts. Priorities for dialogue with the Federal Government:Since there are a number of areas where the primary decision making responsibility rests with the Federal Government, the Government of the Punjab needs to actively pursue a dialogue with it in the following areas: Streamlining of laws to allow the private sector to establish cold storage at major ports. Legislation and capacity building to strengthen commodity markets to reduce risk and access finance, throughwarehouse receipts system and forward contracts. Streamlining of laws and regulations for the punishment of those supplyingadulterated, unsafe, and substandard food products, in a way that reduces the discretionary powers of food inspectors, yet improves the effectiveness of enforcement. Involvement of independent and credible testing laboratories in the food testing process would be helpful in this regard. Defining the minimumgrades and standards for agricultural raw and processed products for the domestic market, and aiming to have equivalence to the WTO recognized Codex Alunentarius standards for export markets. This requires developing a workable mechanism with the collaboration of growers, processors, and exporters for implementation of regulations that have already been developed, such as in cotton. 3.5 FACTOR MARKETS FOR LAND AND WATER: 3.34. Land:Irrigated land i s the most important productive asset in rural Punjab. L a n d distribution is highly skewed-2.5 percent of farmers with over 20 ha own one-third of the land while small farmers with less than 2 ha own only 12 percent of the land, although they account for half the total farm population. Over time, land ownership in the Punjab has become more concentrated with the Ginicoefficient of distributionworsening from0.43 in 1972to 0.56in2000. 3.35. Current land markets promote neither efficiency nor equity. First, land markets are very thin because of the high asset value of land in relation to use value, and the lack of fmancial institutions in rural areas to fachtate land acquisition. Second, an effective land market has failed to develop because of legal impedunents to the sale and purchase of land such as the right of shy% (first purchase), excessive bureaucratic intervention, and lack of a transparent and efficient land 54 administration system. Imperfections results from an antiquated land administration and registration system, high transactions costs in land transfers, inheritance laws that encourage fragmentation, an outdated legal framework, and inability to enforce existing laws (such as tenancy rights).Third, three land reforms since independence have generally failed to achieve their objectives, and have led to increased eviction of tenants, reduced employment and increased poverty (World Bank, 2002). These imperfections slow the exit from agriculture of inefficient and very small farms that cannot provide a livelihood, reduce incentives for very large holdings to sell off underuthzed land, and promote fragmentation rather than consolidation of land holdings, resulting in sub-optimal mechanization and irrigation operations. 3.36. X number of measures could improve functioning of land markets and encourage the move toward more efficient, equitable, and competitive farms. The major emphasis should be on improvingfunctioning of land markets and providmgincentives for large farmers with underuthed land to sell land. Unambiguous laws and an effective land recording system would promote the smooth transfer of titles and an efficient land market. The Punjab has already recognized the need to clarify property rights and to modernize the land registration and records system by initiating a pilot program to computerize land registration records. This program should be carefully evaluated and scaled up accordingly, taking account of successful experiences in other countries (e.g., Thailand). 3.37. Land taxes based on the size of the land holdmg could be set to reflect productivity potential and discourage underuulization of land on large holdings. Reforms in other areas, such as reducing subsidies for irrigation (so that large land holders pay full O&M costs) and improving the efficiency of financial markets to fachtate land transactions, would also promote market-based land consolidation and redistribution. 3.38. Water: Agriculture presently uses about 95 percent of avdable water resources and water has become the major limitingfactor to further growth. I t i s therefore disquieting that the productivity of water used in agriculture has declined in the past three decades.33 Inefficient water management, inadequate maintenance funding, ad hoc bureaucratic and political influences, and lack of economic incentives to use water efficiently cause low productivity of water. Moreover, supply options are increasingly costly and approaching their physical h u t s , while managing demand (efficiency, productivity, pricing, water entitlements, etc.) has received too little attention. 3.39. Although there appears to be significant scope for additional surface water resources development, this i s M t e d if additional storage i s not developed. Among alternative supply options, groundwater development i s already nearing i t s practical hut, and water conservation may be able to add only about 5 percent to total supply. Ground water tables are falling because of over-pumping. Existinglaws place n o quantitative h t s on groundwater withdrawal by individual users. The lack o f these h t s , coupled with tying land rights with water rights, has serious equity implications, because it allows larger farmers with higher pumping capacity and deeper tube wells to have a disproportionate claim over water than others. 3.40. The poor delivery of irrigation and drainage services results primarily from deficiencies in the operation of the canal system, on which maintenance i s often deferred because of inadequate resources. Currently, abiunu (water charges) i s the principal means of sharing and recovering 0&M 33Ali and Byerlee, op. cit. 55 cost.34Abianu proceeds go directly to the Department of Finance, and there i s no du-ect link between recoveries and budget allocations for O&M and other system improvements. A cost-recovery gap has existed since 1970 because of the escalating cost of O&M, the stagnation of water rates and inefficient assessment and collection mechanisms (Figure 3.3). Figure 3.3: Irrigation Department Revenues and Expenditures: 1970/71- 2001/02 6,000 5,000 4,000 v) [1: -.--'E 63,000 2,000 1,000 0 3.41. The current regulatory framework impedes the development of water markets and does not provide incentives for efficient water use. Farmers or their organizations on a distributary or watercourse do not have a legal water entitlement or right-hence water markets that would help farmers decide on the best use of their water and land do not function. 3.42. M o r e than 90 percent of farmers with tube wells sell water, significantly improving access to water, particularly for small farmers, landless tenants and younger households who often lack the resources (or land and water rights) to install their own tube well. However, water purchasers do not have full access rights to the water, and are frequently denied access when water or energy supplies are scarce. 3.43. Water logging and salinity are the most serious outcome of inefficient irrigation and canal seepage caused by a poorly managed water distribution system and inappropriate irrigation practices. In spite of huge public investments in salinity control and reclamation, the problems of salinitylsodxity have become more acute in recent years, because of poorly maintained drainage systems and increased tapping of brackish groundwater for irrigation. Recommendations: 9 Improved water-use efficiency requires better maintenance, rehabifitation, and modernization of the irrigation and drainage systems. Equally important are on-farm conservation measures such as precision land leveling, zero tillage, bed-furrow irrigation, and other water conservation measures. Some of these practices have proven to b e effective in minimizing the impacts of recent drought years, and are being rapidly adopted. 34Farmerspay a share of the capital cost of watercourse lining (up to 50yo) and land leveling( S O 0 ) 56 9 Even with potential additional storage and plausible improvement in irrgation efficiency, the gap between water suppIy and demand will persist unless there is a move toward higher waterproductiviry through choice of croppingpatter-subs tituting high value and water efficient crops such as vegetables and fruits for high water consuming crops (e.g. rice and sugarcane)-as well as increased crop productivity. For example, an increase in irrigated wheat yields from a current level of 2.8 t/ha to an average o f 5.0 t/ha (already achieved in Mexico, In&an Punjab and Egypt) would allow demand for wheat in2025 to be achieved with 30 percent less water. 9 Sustainable water use requires an integrated approach to the use of surface and groundwater together with appropriate pricing mechanisms that reflect the scarcity of the resource. Efficient water markets, tradable water rights and physical infrastructure to transfer water and measure water flows are critical for farmers to maximize their incomes by allocating water according to i t s optimal use (whether on their own fields or sales to or purchases from other farmers). I t i s therefore important to introduce the necessary legal and infrastructure arrangements to facditate movement toward formal water markets. A first step would b e to move toward abiana based on the water volume actually supplied. This would require investments in flow measurement and metering devices and structures and a decentralized institutional set-up involvingFOs and AWBs for assessment and collection. 9 Empowering FOs through a secure system of water entitlements to manage water is a critical part of the incentive framework needed to improve both water efficiency and productivity. The progress on wider institutional reforms has been slow due to lack of ownership by irrigation departments, insufficient capacity and resistance to change, and ineffective monitoring and evaluation systems. Initial findings from Sindh, where the responsibility for O&M and abiana assessment and collection has been transferred to some 60 FOs, i s that both assessments and collections have substantially improved as compared with the pre-turnover years. T h e present use of khal and nehri panchayats as proxies for water course associations and FOs respectively could prove successful if Irrigation Management Transfer (IMTagreements clearly specify responsibllities and functions between PIDA, the AWB and the FOs, and in particular, ensure the effective empowerment of FOs. T h e Government of the Punjab has indicated that it wdl make the necessary changes in the financial regulations to allow FOs to retain their share of abiana to enable them to finance the operation and maintenance activities at the minor and distributary canal level for which they are responsible. T h e challenge for the Punjab i s to quickly develop a program of FO mobilization, training and capacity buildingto consolidate progress and deepen the pilot reforms. 9 To establish an active formalwater market, rights to water usemust be separated fromland for alternative use. An efficient market requires the prior existence of an effective legal institution of property rights, establishing the initial resource endowments of individuals. There is therefore a need to specify water withdrawal h t s by individuals involumetric terms. 9 Finally, increasedpublic andprivate investmentin theirrigation system isneeded to even maintain current efficiency levels. Curative measures WLLl be needed to control and reverse the effects of severe water loggmg and soil salinity. These include lining of canals in some saline ground water areas, improvement of watercourses, introduction of on-farm drainage and land-leveling, improvement of existing and creation of n e w drains including a trunk or spinal drain in the upper and middle basin, and conjunctive management of canal water and groundwater. There may also be 57 opportunities for expansion of the existing system, through small and medium schemes, that have a satisfactory benefit-cost ratio. However, increased public investment requires sharp re-allocation of current expenditures from non-development and establishment costs, and greatly enhanced cost recovery for O&M. 3.6 UPGRADINGTHE AGRICULTUREINNOVATION SYSTEM: 3.44. iistrong agricultural innovation system i s critical to modernizing agriculture, as a large part of future productivity gains, quality enhancement, and diversification must b e provided through generation, adaptation and dissemination of new technologies and information. While a fairly extensive agricultural R&D system was set up in Pakistan duringthe 1970s and 1980s, capacity of the agriculture technology system has declined sharply in recent years (Box 6). Box 6: Current Status of the Agricultural Research System in Punjab Decliningcanacitv in the research system:There is generalagreement that the capacity of the agricultural technology system has declined sharply inrecent years.X fundingshortfallis only part of the problemand reflects wider weaknesses in the system.Researchis managed through a highly centralized, bureaucratic and supply-driven approach. In addition, generally weak and unstableleadership has further underminedthe credibilityof these organizations.As a result, except in narrow areas such as crop breeding, researchcapacity is weak in priority areas for future growth, such as land and water management, post-harvestmanagement, social sciences and policy, horticulture and livestock,most of which have a publicgood natureand requirepublic funding. Decliningcapacity reflects severalweaknesses. First the framework of incentivesfor scientists is especiallyweak inthe provincialsystem. Within Punjab there are wide variations in the human resources policies, with institutes of the Palustan Atomic Energy Commission (PAEC) in Faisalabadhavingvirtually all the critical elements of a stronghumanresource policy and incentive framework, to provincial research agencies where lack of incentivesand qualified staff seriously limit the effectivenessof research.For example, AARI scientists receiveonly about halfof the packageof benefits of PAEC scientists locatedinFaisalabad.Likewiseonly one postinAARI is filed at the Grade 20 level (the DirectorGeneral) versus 68 scientists in the University ofAgriculture, Faisalabad. The rate of decline in research capacity is also accelerating due to an aging complement of research scientists resulting from a hiring freeze lasting more than 10 years, and a "brain drain" as many of the better scientists have left the system. Poor service rules relate to recruitmentof qualifiedresearchers,selectionof researchmanagers (still on a seniority basis), career pathfor scientists, andprovisionof a stimulatingworking environment. Finally, the cumbersome approval processes at the provinciallevel for training outside the country, results in many missedopportunities for strengtheningthe humanresourcecapacity. Good researchis also beingstifledby financial and bureaucraticrestrictionsandprotocols.Most are imbeddedintraditionalbureaucratic ways of doing business, but are inconsistent with the functions and flexibility needed to effectively manage a modernresearch system. Even simple rules that are accepted in nearly all research organizations, such as the ability to retainearned income, are not in place in the provincialinstitutes.In large part these problemsrelate to the continuedmanagementof researchas part of the provincialline ministries, accordingto normalcivil service bureaucratic rules. Punjab is one of the fewplaces in the world whereresearch is still organized wirhin lineministries, and certainly by far thelargest such system. Highly fra-mented research system: Provincialresearchon crops and on-farmwater is separated within the Department of Agriculture, and research on livestock, fisheries, and forestry in different departments.At the same time, various university programs and federal research organizations carry out research in the province on all of these topics, without any institutionalmechanisms for coordination and integration across the various agencies (i.e., between federal agencies, universities, and Provincial crop, livestock, irrigation and fisheries departments).This leads to a highly fragmentedresearch agenda that does not address emergingissues inan integratedmanner. Lack of an enabling environment for private sector R&D: The contribution of the private sector has expanded in the 1990's with some notable examples of creatingnew industries such as spring maize in the Punjab, and providinghybrid seed of crops such as sunflowers. But private sector investment depends critically on an enabling policy environment of Plant Varietal Protection CpW) and Biosafety regulations,bothof which haveyet to be implementedin Pakistan.These regulationsare also requiredunder internationaltreaties; PVPis a minimumrequirementof the TRIPS agreement of the W O , and the Biosafety regulations are requiredunder the Cartagena Protocol of the Convention on Biodiversity (CBD). Thelong delay (now approaching tenyears) in approval ofBiosafetyguideIinesmeans that Pakistan is at least fiveyears behind competing countries in the commercialization of transgenicinsect-resistantcotton which has provided a cost advantage of 20- 25% in countries as diverse as Mexico, China, USA and India, and resultedin substantial reductionin harmfulpesticideuse. Source:For more details, please see backgroundnote: TheP q a bAgn'cultm Innovation SyJtem. 58 3.45. While current spending on agricultural research i s very l 0 ~ , 3 5 it i s important to note that the immedate issue at hand i s not to increase funding, but rather to better u a z e available resources. The Punjab must introduce a more scientifically-based agricultural innovation system based o n interactions among all players, both public and private,-engaged in knowledge creation, those involved in its extension, as well as those engaged in i t s application at the farm level. 3.46. A comprehensive approach i s required on the part of the provincial government to tackle the key policy reforms needed: 9 Establish a competitive fund to support high priority research: The fund could provide operating costs to research by allocation on a competitive basis, using rigorous and transparent review procedures, to help ensure that high priority areas are funded., and would allow the GOP to tap the best scientific talent, whether in the GOP research agencies, agricultural universities, federal research agencies, or the private sector. It could even be used to seek out and test the most promising innovations from abroad, especially in high value products. Implementing this recommendation in turn requires three key actions: Establish a small hishlv- ..Drofessional funding bodv:This apex research body or council would be responsible for liaising with stakeholders to set priorities for the fund, implementing calls for proposals, setting up a rigorous and transparent peer review and selection process, awarding contracts, and monitoring and evaluating funded projects.36 I t should be small, but highly professional, with incentives and rules that can attract high quality staff. Seek out sources of fundins: The provincial government should be the f u s t source of funding by allocating the funding it currently provides in its budget to research institutes through the new apex funding body. While t h i s may mean that much of the current provincial research establishment that could not compete successfully d not receive funds, this d help ensure that provincial research fundtng i s put to optimal use. Other sources of funding would be the proposed USDAfund, and likely appropriations of funds under donor programs, such as the ADB agricultural adjustment loan. Set Driorities for the Fund:T h e funlng agency would liaise with key stakeholders in the public and private sectors to set key priorities for the fund, and would address emerging problems in crop and livestock lseases and pests, innovations to stimulate diversification toward high value products, and solutions to the serious land and water degradation issues. 9 Providean enablingpolicy environmentfor bothprivate andpublic sectorinnovation:An appropriate regulatory environment, in the form of Plant Varietal Rights and Biosafety Rules, is needed to further stimulate private R&D. While these regulatory issues are under federal control, the GOP must remain actively engaged to keep maximum pressure on federal bodies to urgently approve PVP laws and appropriate Bio-safety regulations that dbenefit both public and private research. With PIP, the public system should be encouraged to enter business agreements with the private sector, and stimulate the development of local private R&D companies 3.5Our estimates indicate that the provincial government spends only 0.15% of Punjab AgGDP on agriculturalresearch. Even adding estimates of research in universities and federalinstitutes allocated to the Punjab, the total spendingis sull only 0.26% AgGDP (ie. less than half of the average for Asia of 0.5%, and less than 10% that for developed countries of 2.6O/'). 36 Although the Punjab Agricultural ResearchBoard was constituted for this purpose in 1996 it never served its role, in part because it was not provided the appropriate business rules and flexibdity to operate, and it could not attract high quahty leadershipand human resources. 59 9 Develop a human resource and incentive framework for quality research: The research system i s only as good as the quality of its research leaders and scientists, and the incentive framework for stimulating innovation. An immediate priority i s to strengthen the human resource base of the provincial system through (i) selection o,f dynamic research leaders appointed on merit, rather than seniority, (ii) a transparent and open system of recruitment of scientists based on merit, (iii)establishing a career service stream that allows qualified scientists to advance based on well- defined promotion criteria not tied to available posts, (iv) overhauhg and standardzing the salary and benefits structure for scientists structures, (e.g., the NIAB model), and (v) immediate attention to supporting post-graduate training abroad, especially in fields where there i s little capacity within Pakistan (e.g., biotechnology, some areas of natural resources management, economic policy). 9 Provide autonomy and functional ffexibiliity for effective research management: The dynamic nature o f research requires flexibdity in the flow of funds to enable research managers to reallocate between line items and budget codes. Likewise, linkages with farmers, extension agents, academia, and external agencies, all of which are important for successful research, require that research managers have the authority to respond to opportunities for scientific exchanges of information, income generation, specialized training and materials exchange, without having to seek higher level approval, for what are everyday occurrences in a modern research system. Unul these issues are addressed, the impact of other efforts to improve the research system wdl be h t e d . 3.47. The legal framework for research developed by the Ministry of Science and Technology (Box 7),dif implemented by provincial research organizations, be an important step in recognizing the unique requirements of modern research systems, which differentiate it from other public sector functions. Over the long term, the Province must put in place processes for comprehensive restructuring and modernization within this new legal framework. This restructuring drequire a large redundancy of staff that are not suitably qualified for the new standards, and wdl result in a much smaller but highly qualified publicly funded system. Box 7: N e w Legal Framework for S&T Organizations The Ministry of S&T has developed a set of recommendationsregarding the legal framework and administration measures to enhance the effectivenessof R&D organizations in the country. These recommendations have been approved by Cabinet and are now being implemented through revisions of the Acts or Ordinances for R&D organizationsto providethe necessary autonomy and powers to their BOG, followinga standardizedtemplate. The mostimportantprovisions under the new frameworkare: a) Constitutionof a small (less then 15 persons) BOG of individuals serving in their personal capacity, with an electedChair. b) Open and transparent recruitment of the Chief Executive Office by the BOG for a futed term, against specified criteria. c) Empowerment of the BOG and CEO to approve rules and implement decisions with respect to human resources,financialbudgets, foreign travel, contractingetc, without referral to the parentMinistry. d) Provisionof establishmentand non-establishmentbudget,with flexibllitywithin these categories. e) Encouragement of income generation and partnerships with the private sector, within a profit sharing formula, that includesincentivesto staff 9 Uniform service conditions, with openrecruitmentof staffon merit. g) Rigorous procedures for review of new projects and preparation of a business plan that clearly identifies expected products and endusers 60 9 Improve coordination across research agencies and with technology transfer agents:The government should institutionalize a mechanism through which leaders/managers in the agriculture, livestock and fisheries research, extension, university and other key stakeholders can come together to identify opportunities for improving synergies and avoiding duplication of effort. The apex research body discussed above could be a vehicle for this. Similar coordmation mechanisms should be explored at the ago-ecological system to best link local stakeholders in the decision making at the system level (e.g. ABAD for the Bar& On-farm Water Management for rice-wheat; NUB for saline agriculture, University of Agriculture, Faisalabad, for mixed farming systems of central Punjab). Finally, the role of commodity groups that coordinate actions across agencies for major commodties should be strengthened as a vehicle for setting long-term goals and priorities for the commodlty in question (e.g., using a commodity chain approach that brings in major players from agro-processing and marketing, in order to address emerging market needs). 9 Promote a pro-active approach to improving effectiveness of decentralized extension: The transfer of proven technology adapted to local farmer condltions is of fundamental importance. The Punjab faces a major challenge in the reform o f the public extension model to promote high quality demand-driven services. A new extension policy needs to be developed to address issues of decentralization, empowerment, accountability, and sustainable funding. Provincial extension policv: The government should establish a policy that clearly lays out i t s roles and responsibihties in a decentralized system. These would include technical support and human resources development, monitoring and evaluating overall system performance, knowledge sharing among districts, and evaluation of innovative extension approaches, whether public or private, and identifying, piloting, and supportingscaling up of promisinginstitutional innovations. AdaDtive research and technolom transfer grants: T h e competitive fund above should establish a special window to fund innovative ideas for pilot testing and dissemination of new technologies, especially for higher value products that denhance diversification and growth in the sector, or address specific needs of the poorest farmers. Grants would encourage partnerships between research agencies, technology delivery agencies, especially from the private or NGO sector, and farmers, especially organized groups of farmers. H u m a n resources development: T h e major role of the Extension Wing under the provincial government i s to provide opportunities for continuous in service training for district extension staff. Given rapid changes in the agricultural sector, training programs d have to be considerably revamped to meet the changing needs of a market-oriented and diversified agriculture, such as post harvest handling, business and marketing sktlls, and s k d s in fachtating the emergence of strong producer organizations. 61 CHAPTER 4: IMPROVING SERVICEDELIVERYAND SOCIAL OUTCOMES 4.1. Increasing the opportunities to earn higher incomes i s but one part of the provincial government's strategy for improving the life of its citizens. Together with higher incomes must go the better availabdity of public services, especially education, health, access to clean drinkmg water, and sanitation, to name but the most essential. The chapter discusses generic issues of service delivery in a decentralized setting, summarizes some o f the reforms the Government of the Punjab i s currently considering (including those being implemented under the Punjab Education Sector Reform Program (F'ESRP)), and make suggestions as to next steps in the reform process. While this chapter discusses these issues in the context of primary education and of health, the discussion i s applicable to many other services. Table 4.1: Selected Education Indicators for Punjab and Pakistan: 2001/02 Puniab Overall Indicator Urban Rural Overall Pakistan 1.Literacy Rate (YoPopulation10Years andAbove) Male 71 51 57 58 Female 60 26 36 32 Overall 66 38 47 45 2. N e t Primary Enrolment Rate (YO) M a l e 57 44 47 46 Female 58 38 43 38 Overall 57 41 45 42 3. Dropout Rate (% children 10-18 yrs. leaving school before completing primary level) Male 13 19 17 15 Female 8 18 14 15 Overall 11 19 16 15 Source: 2001/02 PIIlS 4.1 THESTATE OF EDUCATION HEALTHTHE PUNJAB AND IN 4.2. Table 4.1 presents enrollment rates in the Punjab in comparison to the all-Pakistan average. In general, the Punjab has higher literacy and enrollment rates than the rest of the country. Notably, the differential between male and female literacy i s m u c h lower in the Punjab than elsewhere-a factor that augurs well for the future, as dbe discussed. Not only i s the current level of enrollment higher in the Punjab but also the trend in the past few years looks good.Figure 4.1 shows gross enrollment rates for t h e provinces. After a period of stagnation that, in turn, followed some improvement early in the decade, the last couple of years have seen enrollments pick up.The rest of the country, infact, has seen a decrease in gross enrollment rates from already low levels. Figure 4.1: Gross Enrollment Rates by Province: 1995/96 -2001/02 I 80 - 5170 5 - h ~ 0 6 5- 6 - FP Balochistan 1 50 I 1995196 1996197 I998199 200 1102 63 4.3. However, in comparison with other countries of the Punjab's level of income as well as others in the region, enrollment rates do not seem as impressive. N e t enrollment in Bangladesh, for example, i s fully twenty percentage points higher than in the Punjab and Sri Lanka has achieved virtually universal enrollments. T h e Punjab's performance looks m u c h more like that of the poorer states in Northern Indla than either the southern states or the country as a whole. 4.4. Moreover, behind this picture of overall improvement i s substantial variation between children from differing backgrounds. Figure 4.2 shows that enrollments are much higher among high income groups than low, ten percentage points higher for boys than p l s and vastly different between the Northern and Southern regions of the province. Figure 4.2: Enrollment Rates in the Punjab by Income, Gender, and Region: 2001/02 Male FCIlIdk Nonh Centrdl Source:Staff estimates based on the 2001/02 PIHS 4.5. Data on health status in the Punjab, as in Pahstan as a whole, i s seriously deficient. Data on mortality rates, for example, varies substantially across sources. Using h e c t survey questions in the Pakistan Demographic and Health Survey (PDHS), infant mortality was 104 per thousand in the Punjab compared to 94 nationwide in 1990-91. Usingmodel based estimates, the P I H S shows infant mortality in the period 1990-95 as 110 per thousand in the Punjab and 105 nationwide. For 1993-98 the P I H S reports 100 and 96 per thousand in the Punjab and nationwide, respectively. Official data shows a decline from 127 to 82 nationwide for the period 1991 to 2001/2. Data can be mustered, therefore, to make either the point that health status has stagnated almost completely in the 1990's (deteriorating and then recovering) or that enormous progress has been made-enough, infact, to be on track to achieve the Mdennium Development Goals. Discrepancies of this magnitude for the most basic (and unambiguously defined) measure of health status indicate the impossibility o f attributing effects of policy, or even of having a clear picture of reality. 4.6. Ignoringthe question of the usefulness of the data collected, one result common to all sources where the comparison can be made i s that the Punjab seems to have mortality rates consistently a little higher than in the rest of the country and progress, if any, at about the same pace. As a relatively 64 richer and, as noted above, better educated province, t h i s difference i s a real puzzle. Both income and education are usually highly correlated with better health. 4.7. Indeed, using the 1990/1 P D H S data (the only source large enough to be able to measure mortality-and even then only for Pakistan as a whole) Figure 4.3 shows that there are large differences in mortality rates between rich and poor. Children from relatively poor f a d e s are almost twice as likely to die before their second birthday than those from better-off f a d e s . T h i s result i s consistent with those in most countries. Figure 4.3: Mortality Rates of Children under 2 yrs, by Household `Wealth": 1991 0 14 0 12 0.1 0.08 0.06 0 04 0.02 0 0 50 100 150 200 250 Note: The figure is a non-parametricregression (a movingaverage) of mortality of children under two by a measure of living standards constructed as a weighted average of ownership of consumer durables and housing quality. The measureis more accurate at higherlevels of "wealth" so defined. Source:Staff estimates basedon the 1991PDHS. 4.8. The interpretation that progress in improving health has been stagnant over the past decade i s reinforced by data on nutritional status-measured by height for age (relative to a common standard)-from a survey done by the International Food Policy Research Institute in July 1986 - September 1989 and compared to a re-survey of the same f a d e s in 2001 by the Pakistan Rural Household Survey Q?RHS). In this comparison, there appeared to be no improvement over the fifteen year period, a deeply disturbing result. Oddly, on nutritional status, the Punjab i s somewhat better than the rest of the country, though, at 60 percent stunted (height for age being more than two standard deviations lower than norms) this is not m u c h comfort. Exacerbating the mystery around the accuracy of the mortality data, nutritional status i s also usually highly correlated with other measures o f health. 4.9. As a result of the very poor state of data collection on health, further results need to be treated with great skepticism. What can be confidently concluded, however, i s that if the causes of better health-in particular those causes that can be influenced by policy-are ever to be determined, it will require substantially better measurement of the ultimate goal of a healthier public, especially children. Further, data has only been collected in samples large enough to measure progress for large areas- provinces at best. If data i s to be used to aid decision malung in a decentralized setting, m u c h larger and more regular samples dneed to be collected. This pointwarrants continual repetition. 65 4.2 THEROLE OF PUBLIC SERVICES 4.10. This section asks two questions: First, what are the determinants of levels and trends in school attendance and health status and how much of a role does public policy play? Second, given the large (inhealth) and growing(ineducation) private sector, how shouldpublic priorities be set? 4.11. Determinants of Drimani school attendance: Analyses in Pakistan as a whole and in the Punjab offer some obvious possibilities of determinants that are generally observed in other countries. Table 4.2 shows the national and Punjab specific results for rural enrollments. Parents' education i s a strong determinant of enrollment, as are measures of income and wealth both of the f a d e s (as represented by their own consumption) as well as of the communities (as measured by overall quality of housing in villages). Interestingly, the gap between rich and poor is somewhat higher (and significantly so) in the Punjab relative to the country as a whole. 4.12. Most important, however, are the variables that can be influenced by education policy. N o t surprisingly, proximity to a primary school is an important determinant of attendance, though considerably less so for the Punjab than for the rest of the country. Proximity to a secondary school i s also related to higher enrollments in primary schools-an effect slightly higher for the Punjab. This has been observed in several countries37 of widely differing levels of income and i s often interpreted to mean that parents see primary education not as an end in itself but'as a means of obtaining secondary education-frequently required for employment in government service or the d t a r y . That this should be a stronger effect in the Punjab could well be related to the higher overall enrollment rate i n l c a t i n g more children whose parents may be considering education beyond primary for them. 4.13. Unfortunately, there are very few other policy-related variables available for analysis. Many of the features of schools that are often identified as being important for attracting students are not often measured in such a way as to be able to use in this sort of analysis. Various measures of "quality", either o f the infrastructure of schools or the absentee rates, effort and dedication of teachers are frequently mentioned as important. It i s impossible to determine from available data sets which of these alternatives may be important and which not. Quality on some measures i s certainly an issue as recent stules have noted that, of schools that are open at all (inwhich, it must be noted, the Punjab i s very m u c h better than the other provinces) the rate of absences of teachers in the Punjab i s somewhat higher than in the rest of Pakistan-24 percent versus 19 percent.38 4.14. This stdl leaves the question of to what extent i s the improvement in enrollment rates attributable to public policy?39 O n e reason for some skepticism concerning the public's role i s that the growth of the private sector has been dramatic in recent years. Figure 4.4 shows the massive increase in the formation of private schools in the Punjab over the 1990's. Also notable i s the balance between urban and rural areas. T h e few schools that had been opened prior to the 1990's had been primarily in urban areas. Since then, private schools in rural areas have been forming at almost 37 World Bank 1998,2002b. 38 World Bank 2002b. 39 I t i shoped the PESRP will provide an opportunity to assess this impact. The various policies introduced under this program will be subject to rigorous third-party assessments and impact evaluations. The new dimensions of public policy include introduction of incentives such as free textbooks in government primary schools, stipends for gxls in selected low literacy districts, recruitment of school-specific contract teachers, development of partnership agreements between the provinceand the districts, and contracts and partnerships with the private and NGO sectors to improve service delivery. 66 exactly the same rate. However, according to the 2000 private school census conducted by FBS, there i s h t e d growth of private schools in districts in Southern Punjab-districts like Rajanpur, Rahim Yar Khan, Lodhran, and Dera Ghazi Khan, which also have much higher poverty levels and lower literacy rates compared to dstricts in the rest of the province Table 4.2: Marginal Effects on The Probability ofAttendingSchool for Ages 6-14: 1999/00 (Results from Probit Regressions) " * 1 * " - * * * Independent Variables All Pakistan Rural Punjab Rural MarginalEffect T-ratio Marginal Effect T-ratio Quintile 2 0 064 3 95 0 046 167 Quintile 3 0.154 9.71 0.193 7.41 Quintile 4 0.196 12.3 0.213 8.13 Quintile 5 0.244 14.2 0.273 9.77 Rural Female * -0.241 -24.1 -0.191 -11.4 Age 0.316 18.9 0.346 11.91 Age squared -0.016 -19.4 -0.018 -12.6 Mother'sEducation Ever Attended School 0.249 9.44 0.221 5.64 Education>=Grade 6 0.116 1.96 0.135 1.56 Education>=Grade 11 *** 0.238 1.14 Father's Education Ever Attended School * 0.145 11.8 0.138 6.68 Education>=Grade 6 0.143 9.32 0.185 7.36 Education>=Grade 11 ** 0.115 4.19 0.092 1.40 Number of Children in Household 0.010 4.70 0.016 3.95 Amount of agric.land owned: Group 1 -0.017 -1.17 -0.033 -1.31 Group 2 0.024 1.33 0.013 0.43 Group 3 0.035 1.53 0.019 0.44 In PSU/Within 1km. from PSU Primary School 0.193 11.4 0.120 3.45 Primary, Middle & Sec. School ** 0.079 5.89 0.105 4.89 Bus Station * -0.007 -0.58 -0.008 -0.39 RailwayStation * 0.039 1.84 0.87 2.43 Shop 0.053 2.80 -0.033 -0.60 Market ** -0.045 -2.23 0.007 0.19 Bank -0.031 -2.02 -0.037 -1.42 Phone ** 0.030 2.46 0.057 2.65 Post Office -0.001 -0.09 -0.190 -1.0 District Capital ** 0.143 3.77 0.081 0.72 In PSU Hospital/Dispensaq * 0.012 1.00 -0.005 -0.22 HealthWorker 0.013 1.19 -0.025 -1.29 Drainage Fackty ** 0.036 3.07 0.024 1.25 hfotorableApproach Road * * 0.027 1.52 0.018 0.48 Mostly PuccaHouses in PSU 0.073 6.39 0.054 2.57 >=50% of Households with Electricity * 0.125 9.07 0.162 6.12 mdependent, conunuousvariable, for 0-1 vmables, it measures dscrete changem the probabkty for discrete change of vanable from 0 to 1 T-ratios pertam to underlymgProbit coefficients Specification for W-Palustan dudes dummy variables for every province (Punjab is the reference state) Source:World Bank 2002b 67 . Figure 4.4: Formation of Private Schools in Punjab: 1970 -2000 (Number) 6000 5000 4000 3000 2 0 0 0 1 0 0 0 0 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 Year of Formation I Note: 2001 incomplete:Source:FederalBureau of Statistics: PakistanCensus of PrivateSchools 4.15. This trend towards private education, while common across the country, i s particularly strong in the Punjab where, according to the CIET study,40 55 percent of children between the ages of 5 and 9 are in government schools, an extremely low number for t h i s age group by international experience. T h e comparable number for Pakistan as a whole i s 67 percent (implyingrates in the other major provinces in the high 70's given the Punjab's large contribution to the average). In urban Punjab, only about half of children in primary education are in government schools. Comparing the rapid growth of schools (and enrollments) in private schools, this leaves very little of the overall increase in attendance attributable to the public system. 4.16. Several factors may account for the rapid increase in private schools. O n e may b e an overall dissatisfaction with the public system. T h e CIET study shows 53 percent of households satisfied with government education-virtually the same as the 55 percent national average. Dissatisfaction of 47 percent could well be fueling the exodus. However, two other factors are related to the steady increase in female education over time. First, overall demand for education should be rising- parental education being a primary driver. Second, more subtly, recent research41 has used the fact that proximity to schools is also an important determinant of attendance and that having educated women in rural areas increases the pool of potential teachers for the private schools to draw upon. These women are a large fraction of the private teaching force: over 70 percent in the Punjab in contrast to about 50 percent nationwide. So, on both the demand (increases due to educated mothers) and the supply (more potential teachers), increases in female education generate a continuing momentum for education overall and for private education in particular. 4.17. That same research also puts into doubt some often-heard reasons for low enrollments. The vast majority o f private schools established in rural Punjab-well over 90 percent-are coeducational. I t i s not a reluctance to send q l s to coeducational schools that has been standing in the way of progress. Private schools must cater to the wishes of the parents in order to attract business and therefore have a better sense of parents' demand. Also interesting in that research is: 40CIET,2003. 41Xndrabi et d,2002. 68 quality as measured by functional infrastructure (toilets, etc.) and maintenance i s also better than inpublic schools overall costs of private schools (total cost of provision-not fees to students) i s slightly lower than in public schools-Rs. 1000vs. Rs. 1200per year, respectively. The median monthly tuition at private schools in rural Punjab are quite low at Rs. 51 (the lowest inPakistan) Parents appear to b e quite sensitive to various measures of quality in that higher fees are associated with observable school characteristics such as better student-teacher ratios and better qualifications of teachers. T h i s indicates that parents can be good "shoppers" for education and pay attention to quality. 4.18. Exactly how many students enrolled in these schools had previously come from public schools and how many are new students are hard to determine but are critical pieces of information. If expansion or improvement of the public system simply reverses this process-extra public investment will not translate into more overall enrollment if it simply returns students to the public that had left for private education. The only advantage o f expanding the lower price (to students, not to society as a whole) option of public education would be if the benefits would disproportionately accrue to the poor. 4.19. While it i s certainly the case that a higher fraction of children from better-off f a d e s attend private schools, private enrollment had been increasing rapidly in both rural areas and among poorer f a d e s . Table 4.3 shows that even among the poorest quintde, one-fifths of all students are attending private primary schools, negating the popular misconception that private schools in the Punjab cater only to the rich. Table 4.3: Educational Enrollments in Punjab by Consumption Group: 2001/02 Consumption Gross Share in Total in quintile enrollment rate government school government school Poorest 56 80 45 2 72 75 54 3 86 70 60 4 95 65 61 k c h e s t 105 58 61 4.20. I t i s important to know whether enrollment expansion is due to private or public schools. Planning needs to take into account the possibdq that n e w government schools could provide competition for private schools-not in itself a bad thingif there are sufficient numbers of students to attend both-but which could simply displace private schools and lead to m u c h less net increase in enrollment than would otherwise be expected. One possibhty for ensuring complementarity between public and private schools systems i s to encourage partnership between them. Recognizing this, the provincial government has restructured the Punjab Education Foundation into an autonomous body as a vehicle to promote and develop such partnerships. In addition, different models for strengthening School Councils (SC) are being implemented. Inthe first phase, PESRP has established a model for contracting out services to NGOs worktng in collaboration with SCs, which will be implementedin six districts. Current research is planned to study this model (Box 8). 69 I Box 8: Studying Public-Private Partnerships in Educationinthe Punjab Following the denationalizationof private schooling in 1979, there has been significantgrowth in private schools in Pakistan. However, relatively little analytic work, if any, has been carried out to-date examining the causes and consequences of this phenomenon, largely due to the lack of reliable nationwidedata on private schools. For instance, researches and policy-makers do not yet have a good understanding of school-choice decisions by households (inparticular relating to factors such as availability, quality, costs, role of female teachers, etc. across different types of schools). Inparticular, no outcome variables are available related to learningachievements of students in the private sector. As part of the ongoing collaboration between the provincial education authorities and the World Bank in the context of the Punjab Education Sector Reform Program (see Box 8), an extensive study on Public-PrivatePartnerships in the Education Sector is beingcarried out to help fd this knowledge gap. Since public-privatepartnerships constitute an important element of the provincial government's education reform program, the study aims to deepen policy-makers' understandingof the role of the private sector in education. Specifically the study will examine (i) quality of education provided in private rural schools, both in absolute terms the and relative to that of public schools, (ii) fee-structure charged in private schools and its relationshipwith school quality, (ii) the the relevant market structure for (private) education, and the extent to which private schools compete with other (private or public) schools, (iv) factors that influence the school-choice decision of the household, and numerous other such important questions of policy interest. I t will be carried out in two main phases: (a) a careful evaluation of school choice with regard to school quality using detailed data on householdinputs and learningoutcomes in schools, and @) a randomizedintervention includingboth demand side and supply side incentivesto evaluate the outcomes of potentialpolicy changes. The study concentrates on village units endowed both with public and private schools. 120 villages have been sampled from 3 districts (httock, Faisalabad, and Rahim Yar Khan), where fieldwork will be carried out. A list of all households with children of school-going age w d be compiled for all selected villages, following which a household survey will be administered to selected households alongwith school surveys and tests of children in the 4th grade in Mathematics, English and Urdu. Inaddition, test will also be administered to teachers to determine a measure of teacher quality. Tests and household surveys will be administered every s i x months, until a longitudinal data set comprising 3 rounds i s available for a panel of interviewed households and children. The household data collected will be matched with detailed school-level information on fee structure, teacher qualifications, ownership structure, etc. Analysis of these combined data will help yield valuable information on school choice and relative performance among schools in the same village. The results of this analysis w d then be usedin discussionswith various stakeholdersto agree on an appropriate intervention to be studied in the randomized experiments administered under the second phase of the project. Sources: For more details, see Das et al(2003)Improving ScboolQaafip in Pakistan:A Concqt Note, World Bank, Washington DC. 4.21. Incorporatingthe private sector into the planning of the public sector implies that information be collected on the basis of the population of students and not just from the Department of Education's internal management information systems. I t i s not sufficient to keep track of children who happen to attend public schools. I t is necessary to know the progress of all children. While the Punjab Education Management Information System (EMIS) collects data for government schools on an annual basis, s d a r instruments for collecting information on private schools are not yet available. T h e provincial government i s considering conducting an annual private school census to collect updated information on the total student population on a regular basis. 4.22, Determinants of health status and public policy's role in it are even harder to assess than in education. This i s not only because of the lack of relevant information but also because health involves a m u c h more complex combination of determining factors. Education requires a teacher and children in classrooms on a regular basis and some educational materials-most of which are concerns of the Ministryof Education. Health status, in contrast, i s determined by a large number of factors that lie outside of the Ministry of Health. Information to correctly assess these various potential contributors i s simply not available so the following discussion i s merely suggestive. 4.23. The P D H S i s the only source of data that i s both large enough and broad enough to attempt an explanation of infant and child mortality. Table 4.4 presents the basic results of a regression analysis based on data from the whole country. N o t e that there i s no information on the presence or absence of public health facilities in the same data set so a comparison with other policy options 70 cannot be made. However, several very strong correlations do appear. First i s the effect of mother's education, further emphasizing the benefits of expanding female education. Second, the impact of "wealth" i s strong and reinforces the implication of Figure 4.3 above. Table 4.4: Determinants of Health status in Pakistan: 1990/91 Independent variable Coefficient (Standard Error) Wealth - 068 ( 028) Secondary/ higher education -.701 ( 189) Pipedwater - 249 ( 140) Flush toilet -.061 ( 125) Pit todet - 622 ( 129) rural - 099 (.113) ~ n an n that were not significant Source:Staff estlmates based on the 1990/91 PDHS 4.24. Finally, the impact of having any sanitation fachty at all, i s very strong on child survival, even stronger than source of water. Sanitation habits are not directly addressable with infrastructure- related interventions42 but require behavioral change in f a d e s . Achieving behavior change has important public good characteristics (health education-the dissemination of information that i s not associated with a commercial product that would be advertised) as well as direct effects on other village members,43 that is, with large externalities. Further, infectious lsease and deaths from diarrhea among children are almost always more concentrated among the poor than any other lsease condition.44 Therefore, addressing the sanitation problem has a strong claim on public resources on both equity and efficiency grounds. 4.25. What of the effect of public health care fachties? This is hard to say. It i s possible that better sanitation practices are brought about by contact with public health fachties or that the other determinants are simply proxies for access to health care. This cannot b e directly tested. However, there i s reason to doubt the effect of public care. First, very few people, even among the poorest use public health care fachties (Figure 4.5). T h e CIET study shows very low use of public facilities and that "very vulnerable" households were not any more likely to use them than other households. Some of t h i s i s due to fewer people who are "very vulnerable" being within 5 km of a public facility. The current pattern of public subsidy, then, i s not pro-poor. 4.26. Second, overall changes in mortality (either the modest estimates from comparing the PDHS to PIHS or the more dramatic estimate using official records as the baseline) are consistent with changes in the other important determinants. Over the decade there has been increased coverage of water and sanitation,45 increased female education and decreased ferulity (usually associated with lower mortality since high risk mothers-very young, very old, those with largest numbers of children and those that feel themselves to be at high risk-tend to reduce ferulity first when overall rates decline). Together these could account for the observed improvements. Finally, in I n l a and Bangladesh, where information on access to public fachties i s available in the same type of surveys, their presence is not, generally, associated with lower mortality. Perhaps the most comparable case is the Indian state of Punjab. There analyses show no such relationship. ?*Water and Sanitation Program, SouthAsia Region,The World Bank. 43Hughes and Dunleavy, 2000. 44World Bank, 1998, 2002b. 45World Bank 2002b. 71 Figure 4.5: Percent of Households who usually use Government Health Facilities Sindh Baluchistan NWFP Punjab isiamabad Pakistan Source:CIET 2003 4.27. why might sanitation work when health care does not and what does this mean for public policy? Sanitation, as noted, i s characterized as having large external benefits and disproportionately affects the poor. Given the very large private sector in health care, the substitution possibhties with the public sector are also potentially substantial. In fact, analysis of substitution in Paktstan as a whole shows that any increase in the cost of attendmg public facdities that would lead to a 10 percent reduction in public facilities i s associated with an increase in the use of private facilities that amounts to 7 o f those 10 percentage points.46The substitutabllity seems high and therefore increases in public facilities could lead to m u c h lower-than-expected increases in net coverage of health care uulization due to reductions in private sector use.47 4.28. Simply not enough i s known about the contributions of different policies on health status and they cannot be determined with currently available data-in our case as old as 13 years. More, and timelier, information that can be used by health planners and health care providers i s a top priority. Further, even more than in education, the kinds of information that need to be collected must be from population-based surveys and not simply from fachty based information. Since so little of the sector i s covered by public fachties, the focus of attention and data collection should be on what most people are really doing about their health and health care-information that cannot, even in principle, be collected at public health centers. `6 Aldermanand Gertler, 1989. 47 There i s substantial controversy over the "qudty" of care in both public and private facdities. This i s a grossly under- researched topic givenits importance. Recent evidence (Das and Hammer, 2004) seems to indicatethat inurbanIndia, poor people are treated by low quality providers in both public and private facilities (with public MBBS doctors a little, but not much, better than private providers without an MBBS) while richer peopleare treated by higher quality doctors in both the public and private sectors (the privatebeinga little, but not much, better than the public).No such study exists for Pakistan, let alone the Punjab. 72 4.3 ACCOUNTABILITYOF POLICYMAKERSAND PROVIDERS: 4.29. The World Development Report 2004: "Making Services Work for Poor People" identified the core factors involved in making sure that services do, in fact, work for poor people. These are a set of "relationships of accountabhty", or, more prosaically, making sure that service providers feel committed and have incentives to provide services effectively (Figure 4.6). Figure 4.6: Relationships of Accountability in Service Delivery Services Source:World DevelopmentReport 2004 4.30. The basic point i s that the goal i s to improve services to clients, particularly poor clients, which serve their needs and preferences. People are concerned with outcomes. They want educated children, clean and convenient water and better health (and attentive health care). They do not care what managerial system gets it to them. In a market for an ordinary good, people pay for it directly, of course, but in doing this they make sure that they are getting the type and quality that they want. They are closely monitoring the seller-it i s their money after all-and have implicitly or explicitly the abhty to make sure the seller delivers. They can complain (an explicit sanction) or they can threaten not to return or to tell their friends (an implicit sanction). T h e seller has every reason to satisfy the wishes of the consumer since that i s how they earn their living. This i s a powerful incentive. In the dlagram, it is one form of the "short route" of accountability. 4.31. This fable i s relevant to basic services because even if services are provided by the government, the same degree of care on the part of providers to satisfy the needs of clients needs to be maintained. However, this now takes two steps (the "long route" of accountabhty). First, the government must be aware and have the incentive to satis@ the desires of the clients, particularly poorer clients. This relationship of accountabhty is "voice" in the diagram-more commonly known as politics. Service delivery should always be seen as inherently political. In democracies, it i s the politicians that usually respond to voters' wishes, and they must pass on these wishes to the policy-makers. 4.32. Second, policymakers must be able to ensure that providers have the incentives to satis@ the needs of clients. Policy makers need to be able to monitor and induce provider organizations (such as the Ministries of Health or Education) and, through them, f r o n t h e providers 73 such as teachers and doctors, to perform their duties properly. High rates of absenteeism are an indication that this i s not an easy task. This relationship of accountabhty i s labeled "compact" in the diagram. "Voice" and "compact" together form the "long route" of accountabhty. For the "long route" to work well, both policy makers and providers feel responsible to supply the same things that people demand. Providers in the public sector must be accountable to the policymakers who, in turn, mustbe accountable to the public. 4.33. FinaII' monitoringof the performance o f providers is often difficult to do through ordinary managerial supervision. S d a r l y , implementing systems of rewards and punishments for performance based on this monitoring i s not easy. Depending on their capacity, (and even rich and technologically advanced countries often find this difficult) governments might find that reliance on parents, patients or community organizations can be helpful in ensuring better performance-either as monitors or as direct demanders o f services. T h i s i s called "client power" in the diagram. This i s a second meaning of the "short route", Le. not relying on the market (also a short route but only by default)-but as a deliberate policy initiative to take advantage of local information and enforcement capacities. 4.34. These services are provided by the state for good reason-market outcomes are not always sufficient. Poor people may not obtain education without state intervention, even if it is in the long term interests of the children. Primary education, in particular, i s considered to have benefits to the community at large (externalities). In health there i s a wide range of activities with differing degrees o f market failure-from the pure public goods of pest control for which there cannot be a private market to control of infectious disease to relatively inexpensive curative care which i s close to a pure private good. Public sector priorities should balance the abhty of markets to deliver particular services to the ability of governments to deliver them given problems that appear along the "long route" of accountabhty. The remainder of this section examines ways in which each of these relationships of accountabhty can be made to work better in the Punjab and, for example, as i s already being done under the PESRP. T h e following section examines the appropriate role of the provincial government that will further improve these relationships in the devolved system. 4.35. Voice: Policy makers, in the devolved system the nazimeen, must believe that improved services serve their own interests. Under the PESRP, all district govemments have entered into terms of partnership (see B o x 8) with the provincial government in a major effort to improve education service delivery. In the course of consultations for the reform program, a majority of Zdla Nazimeen stressed the high priority they accorded to improving education outcomes. There are several examples of this, including the Zilla N a z i m and UnionNazimeen in Faisalabad who feel responsible for service delivery, and consider success to be important for their political futures.48 In Rahim Yar K h a n as well, both health care provision and educational improvement are given high priority for the same reason-the N a z i m and councilors believe that better services amount to good politics. 4.36. One major expected benefit of devolution was that local governments will both know more about the specific needs of their constituents and be more likely to feel accountable to people. But devolution, by itself, does not ensure this dhappen. If politicians in Faisalabad and Rahim Yar K h a n do, in fact, reap political benefits from an emphasis on improved service delivery, their experience i s likely to be repeated. T h e devolution study indicates that Nazimeen do find that 48World Bank 2004c 74 listening to citizens i s more important after devolution-certainly a promising result. However, d e p e n l n g on local c o n l t i o n s and history, improved communication may not be the result, local politicians may be subject to political capture by elites and services may not reach the poor. To counter this possibility, actions on the part of the provincial government, as discussed below, and more open, public, and informed debates on issues are necessary. Box 9: The Punjab Education Sector Reform Program Followingextensive consultationswith students, parents, teachers, and local officials, includingin-depthmeetings with district education teams and NGOs, the Government of the Punjab has recentlyembarked upon an extensive reform program to enhance access and quality of educationin the province. In 2004 the World Bank extended US$lOO million,the first of a series of three IDA adjustment credits,in support of this reformeffort.The government's medium-termreformprogramhas three mainpillars: (11PubIicfinance reforms: The government has increased public expenditures on pro-poor services with a focus on educationby increasingprovincialpublic sector expenditures with conditionalgrants for districtgovernments. 611Devolution reforms: In line with the ongoing devolution initiative,increaseddecision makingauthority on resourceallocation has been passed on to district governments. The provincial government is also developing an implementation framework that clarifies the roles of service providers and users and institutionalizesperformance-basedbudgeting. (iii)Education sector reforms:To improveeducationsector performance,the government is takingmeasuresto strengthen education accountability by establishing transparent criteria for teacher recruitment and deployment. These reforms support the effective use of School Councils (SCs) to strengthen parents and children's participation in the process and to generate improvementsin monitoringand evaluationof the educationsector. An integral part of the reforms is the clarification of the respective roles of the various tiers of government under the decentralizationprogram and the introductionof incentivesfor districtgovernmentsto improve service delivery.These reforms are embodied in the Terms of Partnerships (TOP) agreement between the provincial and district governments,initiated for the fxst time in Pakistan.The TOP improves the transparency and efficiency in the financing and management of the educationsector by clearly defining the responsibilities and obligations of districts, province and SCs in planning, implementingand monitoringthe education system and use of funds. The TOP also provides financialand performancebased incentives for districts to improvetheir education service delivery. Duringthe first year, all 34 districts receivedearmarkedgrants for educationafter signing the TOP. The size of the secondand outer year grants is based on needand is also conditionalon the attainment of educationperformanceoutcomes and an increasein district expenditures on education agreed under the TOP. Finally, the TOP agreement also increases client participationin service delivery by requiringdistrict and provincialgovernmentsto take all necessary administrativeand financial actions to assure that the SCs are effective partnersof the educationsystem. 4.37. ComDact: Throughout South Asia problems such as absenteeism or discourteous behavior on the part of providers are common. In India and Bangladesh, for example, surprise visits found that 42 percent and 35 percent of medical providers and 25 percent and 18 percent of teachers, respectively, were not at their assigned posts. S d a r (though less consistently measured) results appear to be true for Pakistan. Further, the PIHS indicates that for urban Punjab, 21 percent of people avoid government health facilities due to discourteous treatment by the staff. Clearly, neither absenteeism nor lscourtesy was the intention of policymakers but reflects their difficulties in ensuring appropriate performance by frontline providers. How to improve the monitoring and administrative capacity of policy makers i s a high priority. 4.38. Reforms in improving the "compact" can vary substantially. Under the PESRP, six district governments have entered tripartite contract arrangements with NGOs like the National and the Punjab Rural Support Programs to directly manage government primary schools in partnership with the SCs. This i s an interesting example of using explicit contracts to engage the services of either the non-profit or for-profit private providers. O n e advantage of explicit contracts i s that accountabhty can be much stronger to the policy-maker. Instead of government officials feeling like they need to 75 make excuses for deficient performance, the explicit separation of roles of policy maker and provider allows the former to want to find cases of bad performance in order to correct them.49 A second advantage i s that payments in contracts can be made dependent on the actual performance of the provider, improving the effort put in to achieve these goals. T h e model being used in the Punjab i s based on a similar model developed for the health sector (and subsequently for schools as well) in Rahim Yar Khan. Another example i s in Bangladesh, where health educators were issued performance contracts to teach the use of Oral Rehydration Therapy to mothers. Bonuses were paid for the number of mothers who could explain how to handle diarrhea (withORT) as measured by independent evaluators-not program administrators or the providers themselves. This led to both an increase in mothers' knowledge and, as a means to this end, more use of hands-on education techniques instead of classroom learning. The change in teaching technique was not devised by program managers, it was a result of the teachers themselves wanting to make sure the information got across since this wouldincrease their bonus.50 4.39. This sort of contract may not always work. Another major benefit of devolution i s potentially that different districts can experiment with alternative methods of dealing with their problems. Differences in local conditions or attitudes may make the appropriate response vary in each case, or, on the other hand, an experiment may be successful in one district and be considered a model to emulate by others. Recognizing this, the Government of the Punjab has allowed for considerable diversity in the partnership models for SCs (Box lo), and some districts in the province have already begun experimentation with alternative ways of doing business. T h e CARE initiative in Lahore, for example, i s a way of usingNGO's as managers of government schools (Box 10). As and when such examples are successful, other district governments take them up with suitable modifications as needed, with the view to replicating these successes within their respective jurisdxtions. Box 10: Community Management Schools Models in the Punjab ModelA: 1. A selected NGO enters into an agreement with the district government to take over 400 schools in various union councils. 2. The NGO now becomes responsiblefor the operationand management of these schools. 3. The NGO also forms SCs in linewith the guidelines of the EducationDepartment. 4. The capacitybuildingof the council is also the responsibility of the NGO. The PESRPpays for the field of the NGO and also places a reasonablesum at the disposal of the Councilwhich it will spend on improvementof the school. Model B: 1. A selectedNGO enters into an agreementwith the district government to form SCs in 400 schools in various union councils. 2. The operation and management of these schools remainwith the districtgovernments. 3. The NGO forms SCs in line with the guidelines of the EducationDepartment, and carries out capacity building of these councils. The PESRPpays for the fielding of the NGO and also places a reasonablesum at the disposalof the Councilwhich it will spend on improvement of the school. Note: Five districtgovernmentshave opted for ModelA, while one has opted for ModelB. -19World Bank 2004b, p. 98 51)ibid,p.102 76 Box 11:NGO Management of Public Schools:The CARE Initiative in Lahore The Government of Punjab launched a scheme in 1998 to improve management of government schools by encouragingNGOs with a proven track record in the education sector to "adopt" schools of the Metropolitan Corporation Lahore (MCL). The charitable trust, Cooperationfor the Advancement Rehabditation and Education (CARE) was amongst the fxst NGOs inLahore to be approachedwith an initial offer to run ten schools. Since this initialpilot, CARE's partnershipwith the Lahore city government has subsequentlyexpanded to include the responsibilityto rehabilitate, finance, and manage 165 MCL schools across the city, with 840 teachers and a total enrollment of over 80,000 students. When it fust adopted the MCL schools, CARE was confronted with many challenges: in addition to the problems of staff shortages and teacher absenteeism, these schools were characterized by poor teaching methods, low student pass rates, high dropouts, and deterioratinginfrastructure. The schools lacked basic amenities such as clean drinking water, latrines, blackboards, chalk, etc. To improve school infrastructure,CARE made major repairs as well as constructed additionalclassrooms, provided furniture, improved sanitary conditions by repairingand/or constructingtoilets, set up functional laboratories for better understanding of basic scientific concepts, as well as made a small library or book resource room in every school. CARE chose to work in close partnership with MCL head teachers and union leaders to uplift the standardof educationin schools, despite the government staffs initialskepticism and suspicion.A CARE-hired teacher was appointedas an "Internal Coordinator" in each school to work in tandem with the MCL Head teacher to monitor the performance of teachers. Weekly supervisions of schools were carried out by a volunteer "External Coordinator", whose main role was to conduct quality checks to ensure regular attendance of teachers and students, monitor performance of staff, ensure that copies were checked properly and tests were administered and recordedevery month. Inaddition, CARE appointed an "Academic Coordinator" to streamline the curriculum and teaching practices to improve the standard of educationin schools, and a Teacher Training Center was set up to equip teachers with modern teaching methodologes. CARE's guiding philosophy is rooted in the belief that education i s the birth right of every child and should never be denied to anyone, as well as to combat the notion that the poor do not "want" or "value" education. In addition to running schools across Lahore, it has also initiated a merit scholarship program for its outstanding students to finance their higher education.This program, which was started two years ago, covers the academic and in some cases, living expenditures of 82 students enrolled in such prestigious institutions as King Edward Medical College, IOnnaird College and Government College. Finances for the NGO are generated from donations, with 98 percent of these coming from contributions within Pakistan. Sources:CAREwebsite: www.care.ory.pk,"Empowerment through Education" AsianJournal of ManagementCases, Volume 1 Number 1:Jan-June 2004. Pgs. 39-60, as well as various newspaperarticles, reviews, and interviews. 4.40. Reforms of the agreements between policy makers and providers can be done within the public sector as well. In Rahim Yar Khan, the district experimented with an alternative organization for the functioning of the Basic Health Units (BHU).Instead of keeping a Medical Officer (MO) responsible for a single BHU, each was given responsibdity for three BHUs, at an enhanced salary of course, that they would visit each day in rotation. In addition a revolving drug fund was established that ensured a more reliable supply of pharmaceuticals. People would know when the MO was to be present and were more confident of being treated with adequate supplies. T h e use of public BHUs increased dramatically over the first five months of the experiment. It i s too soon to say if this i s an unambiguous success or if the greater use of public facilities translates into better health status of the public. But the dramatic increase in use-more than doublingthe monthlyvisits from 74,000 inJune 2003 to an average of 175,000 for each of the following five months-is an early marker of success. Again, whether this can be adapted to other districts is, and should be, decided based on local conditions-experimentation that devolution allows. 4.41. Client Power: For good reasons the discussion o f "compact" used several examples from the health sector. Since many of the market failures in health are so severe, it i s not possible to solve them without a substantial government role and improving the "long route" of accountability i s necessary. There i s no alternative to the government providing (or contracting with full payment for) pure public goods. Similarly, ensuring that immunizations are complete does not require demand- driven fixed facilities (that are often difficult to manage) but i s not likely to b e accomplished without drect public involvement. W h e n it comes to increasing client-power, however, there are many more possibilities in the education sector. While it i s hard to attribute better health to a particular health 77 care episode, parents-even fiterate ones-can tell if the teacher i s showing up to work on a regular basis or if their children are learning anything. In fact, they are in a better position than government hierarchies to do SO. 4.42. Client power can improve services in several ways. In one sense the simplest i s to recruit clients as aides in monitoring the performance of providers. By ensuring that complaints will be heard and addressed, information from clients can be used to improve services. In another sense, this i s not simple at all and requires a tradltion of expecting services to function properly and of access to government ears. In the Punjab, only ten percent of respondents in the CIET survey knew h o w to complain about government health services. Increased devolution may make such lines of communication clearer but that remains to be seen. 4.43. The second way to empower clients i s to provide them with more choice and competition among facilities. T h e growth of the private sector in both health and education already shows that people have options to public facilities if they are dissatisfied. However, competition and choice can be incorporated into government programs to improve efficiency of service delivery. A particularly successful case i s the Female Secondary School Assistance Program in Bangladesh. Here, money was deposited in bank accounts in the grls'.names and could b e used to pay for any secondary school- government, NGO or for-profit-provided she stayed in school, maintained a passing grade and remained unmarried. Secondary school enrollments rose 10 percent per year-mostly due to female attendance-for the decade after the program started.51 M o r e interesting, many of the barriers that were thought to keep girls away from school and that the public system had a very hard time overcoming-separate latrines for boys and girls, more female teachers-were solved very quickly once the schools had the incentive of attracting more girls and their tuition money. 4.44. The Punjab government's stipend program for girls enrolled ingrades 6-8 i s designed to attract more girls to schools, and i s being implemented in the 15 lowest-literacy districts in the province. It shares some of the features of the Bangladesh program, but in i t s first phase lacks the dimension of encouraging competition between public and private schools.52 T h e stipend i s paid to f a d e s if they keep their 9 1 s in school (maintaining at least 80 percent attendance rate). But they do not have the choice of where to send them. If the main reason that girls are not attending school i s because of financial constraints such as cost of transportation, the program might work.53 If, however, characteristics of schools such as female teachers or headmasters who make sure absenteeism i s kept to a minimum or any other dimension of quality on which schools might compete, the lack of portability of the scholarship across schools h u t s the effectiveness it would have on improving school performance. T h e government plans to subject this program to a rigorous evaluation through a third-party to assess its impact and effectiveness at increasing enrollments. T h e first evaluation i s plannedin December 2004, by which time the program will have been runningfor one year. 4.45. Competition i s not always an appropriate instrument, particularly in sparsely populated areas (that could sustain only one school, for example) or at primary level where long travel i s not acceptable. More appropriate might be the third way to empower clients: through participation, or, 51 World Bank 2004b, p. 69 52 Many of these low-literacy districts are in Southern Punjab, where private schools have relatively limited coverage, particularly inruralareas and at the middle-schooland secondary level. 53 Followingdisbursement of two quarterly installments, monitoringdata shows encouragingresults:between the first and second quarter, girls' attendancewas found to have gone up by 6 percent. 78 more direct management of facllities. There are many examples around the world in w h c h giving parents more authority-usually the authority to hire and fire teachers-has dramatically improved performance of schools. A celebrated example i s the EDUCO program in El Salvador in which the Ministry o f Education entered into explicit contracts with communities to run their schools. Ths included hiring teachers and undertaking regular inspections. Studies indlcate that each visit by the parent's association increased test scores in both math and reading.54 4.46. Closer to home, a recent study reports that Madhya Pradesh in Indla i s the state that has benefited most and early on from the national District Primary Education Project. Once again, a major difference was parent associations that could hit-e (at below official pay scales) and fue teachers. Absenteeism was found to be much lower and b o t h student enrollments and attendance much higher than in the control group.55 4.47. Prospects for this form of client power in the Punjab are being developed by the provincial government through NGO contracting and strengthening of SCs. In the past, school management committees in the Punjab have shown mixed results. Studies conducted during the nineties under the Social Action Program demonstrated that SCs performed well when there was a committed head teacher along with grants for school improvements; however, a large number of SCs were not very successful due to frequent reconstitution o f membership and changes in government policies. There are exceptions: for instance, a recent study finds that SCs in Faisalabad are active throughout the district. The day-to-day running of schools i s under the responsibllity of the Committees, which are chaired by the School Headmaster and have 10 members in all. They meet monthly and approve expenditures on a monthly basis (e.g. equipment, repair and maintenance etc.). Contributions from the community can also help finance school costs. To help finance the non-salary recurrent costs, contributions are collected from pupils,56and in the girls' primary school visited, private schooling in an evening shift generated funds that contributed to paying bdls such as electricity. This school was in a peri-urban area, and the situation in a more rural, less wealthy, environment is hkely to be far worse. I t i s clear that the relationship between staff and the local community i s crucial.57 4.48. Developing client power through participatory means may take some time to take root in the Punjab. The recent CIET survey indcated that just over 2 percent of households nationwide had participated in any voluntary group. A source of concern i s that relatively well-off people were more than twice as likely as others to have participated in such groups than those identified as "vulnerable". This i s a concern since, should such voluntary groups b e used as a basis for citizen participation, it risks services being less-well suited to the needs of the poor. In the Punjab only 3 percent o f men and 1percent of women (identical to the national average) had even heard of Citizen Community Boards. 4.49. This perspective on improving service delivery depends on government being as concerned, as focused and as knowledgeable about ultimate outcomes of services, as are the people themselves. It requires the government to listen to the needs of the public via `Lvoice'', particularly of the poor. It also requires that providers (as organizations, such as h e ministries and as individuals) be faced with SJJimenez and Sawada, 1999. 5.5Glinskaya andJalan, 2003 56 In general, 1 Rupee p e r child i s collected in classes 1-4 and in classes 5-12,2 Rupees are collected per child. 57 Williamson et al. 2004 79 incentives to achieve better outcomes on behalf of the people either by "compact" via the long route of accountabiltty or by "client-power". All o f these relationships of accountabiltty require more information about what outcomes are and what can be done to improve them via policy, a point to be discussed below. If the goal i s to reach more people-to improve the health of the vast majority of people who do not use public fachties or ,to increase enrollments, necessarily by children not currently in the public system-the focus must be on the population as a whole, not merely on the functioning of the public system. Figure 4.7: Framework for Policy, Planning& Budgetinginthe Punjab Plans & Strategies Allocation Instruments MediumTerm ReductionStrategy Budgetary Framework, Budget White Paper Local Government Local Government Budget and Development Annual Development Plan 4.50. The Government of the Punjab has recently developed a framework for Policy, Planning and Budgeting for the province. I t has gone a long way in reforming its entire public sector, and at the macro level has had significant achievements. T h e approach being implemented i s consistent with modern approaches to public management and delivery-the establishment of aggregate fiscal discipline, the orientation of government towards performance; the devolution of managerial powers; and encouragement of participation in service provision. There are several ongoing initiatives to support improved sector performance. In particular, reforms to the planning and budgeting framework are underway, illustrated by Figure 4.7, to promote m o r e rational resource allocation. So far the PRSP has been finalized, a three-year PESRP i s being implemented, and a Medium T e r m Budgetary Framework i s being developed.58 T h e Provincial Finance Commission has issued an interim award. Sectors and local governments wdl soon be encouraged to develop pro-poor 58 Inaddition to the Education Sector Reform program already mentioned, two other key donor programs in the Punjab supportingcross cutting reforms are the Punjab ResourceManagementProject and the Devolution Support Program. 80 strategies. In FY04, the provincial government has significantly increased i t s allocation to the education sector, and i s passing this on to the &strict governments as an earmarked grant. 4.51. However the challenge i s translating the good intentions, which are set out on paper, into reality. These are ambitious and involve changing behavior that has been entrenched in the bureaucratic and political establishment. The reform process described in Figure 4.7 i s a good step towards fixing the internal problems of coordmating political decisions with policy choices, but improvement on the ground must also take into account the other relationships of accountabhty. 4.52. First, the reforms must engage rather than bypass the reality that engineering change in social service delivery will be highly political at all levels-MNAs, W A Sand Nazimeen. Reforms mustaim to channel the political agenda constructively towards enhancing services. That is, they must continually be aware of the need for "voice" in improving services. Second, while the intention i s to make the process more performance oriented, concrete steps need to b e taken to measure performance. In both senses, the next challenge i s to make the reform process more outward looking rather than inward. 4.4 THEROLEOFTHE PROVINCIAL GOVERNMENTUNDER DECENTRALIZATION. As noted, service delivery does not necessarily improve with devolution per se. Since the devolution process in Pakistan i s s t i l l in its early stages, it i s much too early to make firm judgments about how well it i s working.What i s clear, however, i s that devolution i s a fait accompli and more decisions d be taken at the local level as devolution proceeds. This fact does not mean that the provincial government i s any less important than it has been, but it does mean that i t s functions wiU change. T h e task i s to make sure that the provincial government reinforces the relationships of accountabhty discussed above as well as to make sure that certain functions that transcend local governments are adequately performed. Early indcations are that there are some ``growing pains" in decentralization, that is, that roles between the levels of government are not yet completely clear. I t can be hoped that this lack of clarity is simply a natural transition period that would work itself out in time. However, there are several considerations that might make the transition period shorter. Figure 4.8 adapts Figure 4.6 to make it applicable to the current decentralized setting in Pakistan. 4.53. The discussion of accountabhty in the previous section i s relevant to decentralization with three modifications. First, there now two "voice" relations between people (particularly poor people) and government-to local leaders and to provincial (and to national, but that w e leave aside). There i s a presumption that local policy makers care more about service quality because they are much more likely to be held responsible for it than are provincial governments-a level that has many more functions to perform than health, education, and water supply. On the other hand, dependmg on the social structure at the local level, whether the voice of poor people is more likely to be heard at the local level rather than other citizens is an empirical matter and the provincial government may take it upon itself to be a check on local elite capture. 81 t Workers, for example, have more than one kmd o f supervisor). Again, this may be simply a matter for time to work out but the transition may be spedup with focused intergovernmental discussion. 4.57. Direct compacts are used by provincial governments in two modes. T h e first, and more problematic, i s the use of vertical programs whose planning, financing and operation are, to a greater or lesser extent, retained by the province. These are usually responsible for lack of clarity of responsibhty on the part of providers and run the risk of subverting the democratic responsibhties and accountabhty of councilors and nazimeen to their constituents. The potential benefits of devolution depend on stronger accountabhty for budget priorities and service performance. Subject to a few, very h t e d areas, vertical programs shouldbe examined with a very skeptical eye. 4.58. X second hnd of compact that devolution has introduced i s the grant mechanism from province to district. Generally speaking transfers need not, and, to the extent local governments protect their poorer citizens, should not be tied to specific services. However, if the province i s interested in improving particular services, per se, part of the grant can be made conditional on outcomes for i t s poorer citizens. So, as a protective mechanism, particularly when there i s doubt that local political leaders are adequately representing the wishes of their constituents, transfer formulae can incorporate performance criteria. In general, if the goal i s to improve living conditions of poor citizens these conditions are likely to involve more attention to communicable disease control, increased enrollments in primary education, and more extensive access to clean water. And all of these need m u c h better measurement if transfers are to be made conditional on them. As noted in B o x 8, the government has already begun to do so in the education sector. 4.59. Enhancing `Went power": Finally, provincial government might experiment with policies that give purchasing power duectly into the hands of clients. This does bypass nazimeen but in a way that increases overall accountabhty to clients. O n e example i s the School Council model under which NGOs are being contracted for school improvements. Another example would be the possible modification of the middle-school g l s ' stipend program based on its evaluation, and in a manner that would allow competition between dfferent hnds of schools as in the comparable program in Bangladesh. Other "voucher"-like mechanisms that allow choice between government, NGO, or other private providers can be experimented with in water and health care as well but, as with any experiment, they shouldalso be subject to rigorousevaluation. 4.60. Beyond improvingaccountabhty, there are four specific areas in which provincial government can make a significant contribution to service delivery by overcoming certain systematic problems in a devolved setting: e ensuring that activities with substantial spillover effects across district boundaries are properly handled e providing services with large economies ofscale that would not be efficiently handled at district level improving the distribution of income across the province by disproportionately augmenting the local revenue collection of poorer districts generating and disseminating a wide variety of types of information. 83 4.61. Districts answering to their own constituents' needs may not adequately consider spdlover effects from one district to another, and services may be sub-optimally provided. Many of these lie outside the health, education and drinking water services. Watershed management, of particular important to the province of the "five rivers", i s the most obvious example. However in health also, there are aspects of the control of infectious disease, particularly at border areas of districts, which do not respect political boundaries. Some o f these activities remain vertical programs. Ingeneral, vertical programs disrupt the budgeting processes of district government. In the specific case of infectious disease, however, there are good reasons for provincial and district officers to work closely together and for the province to be able to support activities that benefit more than the district alone. 4.62. Interestingly, some of the vertical programs with the greatest potential for dealing with such spillovers are not as completely devolved, that is, more o f the expenditure i s delegated, than others with somewhat less of a claim for addressing such "external" effects. For example, while the Federal Government supplies materials of the malaria control program, employees are on the payroll of the districts. Similarly, tuberculosis, now handled through the "directly observed treatment system" has procurement of materials a provincial responsibility, prioritization and supervision of BHUs are a &strict responsibhty. The provincial govemment should certainly monitor the effectiveness of each of these (withthe Extended Program of Immunization being somewhat more of a "gray" area), even ifit does not drectly administer them.60 4.63. In contrast, the Lady Health Worker (LHW program remains a national program-the only one fully funded by the federal government includmg salaries paid through the provincial coor&nator. Provincial and district coordinators undertake parallel supervision. The LHW program certainly has high potential for improving health. Recent experience in Iran seems to be very positive.61 However, the types of services it provides (family planning, nutrition promotion, immunization (again, a "gray" area), and treatment of diseases like respiratory infections and diarrhea) have fewer cross-district effects than other activities. Such anomalies as to who i s assigned responsibhty for what program shouldbe worked out over time. 4.64. Scale economies for the set of services we are focusing on refer mostly to referral hospital based treatment at levels that district hospitals should n o t have to duplicate. Provincial health authorities should be apprised, at least, of equipment purchases by district hospitals that may be more efficiently handled at a tertiary fachty. The tertiary facihties, themselves, are most likely to be either provincial fachties or private facilities funded by insurance that would have to be regulated, if not provided directly at province (or even federal) level. Health insurance i s an area of sufficient complexity as to go well beyond the scope of this report. 4.65. A second area in which scale may favor an active provincial government i s in the choice of curriculum for primary and secondary schools. As argued elsewhere in this report, the growth of high quality, high paying jobs province-wide has been hampered by shortages of workers with particular slulls. N o t only do large firms report poor human capital and skills as an important constraint, but small and medium enterprises do as well. T h e provincial government may be better placed than &strict staff to confer province-wide with potential employers, or even internationally, in order to identify areas inwhich curricula of schools may be made more relevant. 60World Bank 2004c. 61Islamic Republic of Iran: Primary I l e a l t h Care and the Rural Poor. Case study presented at the 2004 Shanghai conference. 84 4.66. Analogous to the problem o f tertiary hospitals, the province dneed to be involved with regulations relevant to university education. W e are not recommendmg public expenditure for university education even at provincial level. Such spending i s universally regressive (benefits go to students from f a d e s who are relatively well-off compared to average taxpayers) and the benefits are largely private with the possible exception of basic research-very loosely, if at all, related to tuition. However, the market for loans for higher education i s underdeveloped due to the absence of collateral for human capital investments. Therefore the guarantee and the collection of student loans may be a government, likely the provincial government, responsibility. Slrmlarly, h e c t support for research outputs as well as general certification of institutions have claims on public money, again, much more appropriately financed at levels above the district. 4.67. Redistribution of income from rich to poor has a very h t e d scope within districts. In the Punjab, average incomes clearly fall moving north to south. A balance must b e struck between use of own revenues (which WLLl hkely be more closely watched by local officials) and support of poorer districts by provincial government. As discussed, transfers from provincial to local governments can be made conditional on local policy. 4.68. Better information i s needed at every level of government. H e l p in' collecting and disseminating information might be the most effective way that provincial governments can improve service delivery. Information provides an example of spdlovers (experience in one district can help planning in others) and of scale economies (certain kmds of lessons can be obtained only by comparing experience across districts and of redistribution), as poorer districts may be less inclined to take the longer-term perspective that acquisition of information requires. 4.69. Further, individual districts may be perfectly happy to take credit for progress but less happy to admit problems. So, information needs to be collected by relatively disinterested thud parties. The province can take this role. This leads duectly to the last topic. 4.5 BETTERINFORMATION IS CRITICAL 4.70. As has been raised repeatedly in this chapter, at the core of all attempts to improve services i s the need to generate and disseminate information. All of the relationships of accountability that d help lead to sustained improvements in service delivery require better information-particularly o n the ultimate outcomes of services. If government i s to provide services that people want and need, it has to be as focused on the ultimate goals of these services-better health, better literacy-as people are themselves. 4.71. People themselves need more information on what services they are entitled to, what services they are getting, what services others in their &strict or in other &stricts are getting, and who i s deciding these issues. They also need to know how the health and education outcomes in their districts compare to other districts. These are all necessary to help them be aware of and insist on the most effective services. For instance, a start has been made in the education sector, with a major education awareness campaign launched in the first phase using the print and electronic media to publicize the various dimensions of the PESRP. In the next phase, specific information campaigns wdl be launched at the &strict level and below. 85 4.72. Policy makers need to know what the state of health and educational status i s at a level o f government (district or below) for which someone can be held responsible. They also need to know which policies work-they need better analysis to help them figure out which of the policies they try actually translate into improved health and education outcomes. This means more regular and systematic collection of data on both inputs affected by policy and outcomes so that such analyses can be conducted. 4.73. T h e first sections of this chapter examined a few questions that available data might address. But the truth is that these data are too aggregated in most cases, incomplete, outdated, and inadequate for answering most policy questions. The conclusions in those sections are only illustrative. Recent developments, such as the adoption of the Multiple Indicator Cluster (MICS) and the Core Welfare Indicators Questionnaire (CWQ surveys, are encouraging. It i s important that such surveys be done regularly and with large enough samples that they can be used at the district level. Repeated surveys are necessary in order to find causal relationships. Over time, they can also be used as benchmarks for assessing improvements. Findmg out that health is worse and enrollments lower in the southern, poorer, districts i s helpful as a first step. But no policy maker i s really responsible for that. What policy makers can be responsible for i s changes in the health and education status o f their districts, whether poor or not, if the impact of policy can, in fact, be discovered along with other determinants. 4.74. In addition, for the sake of both policy makers and f r o n t h e providers, more rigorous experimentation and evaluation of policies in terms of their ultimate impact on outcomes should also be a priority. Monitoring and evaluation units in all service oriented ministries need to be created and improved to be able to answer the fundamental question: `"hat works?".W h e n something i s found to work, this information can be transmitted to all district policy makers, who can then replicate it. When something i s found not to work-it should either be analyzed to fmd out why it did not work, or it should be dropped. It is often dfficult for governments to adrmt mistakes but, as a senior policy maker in the Punjab very aptly noted, correcting mistakes (`is not admitting failure-it i s just changing strategy". 4.75. Further, information must be "outward loohng" as well as "inward loohng". T h e latter can be found by improvingManagement Information Systems. T h e former requires data on the population as a whole. In sectors such as health and education where the private sector i s large and growing and where (as in the case of education) many may not be served at all, having an accurate picture of the internal workings of ministries i s not sufficient-what i s needed i s an accurate picture of how all people are living their lives: are their children in school? A r e they suffering from avoidable health problems? 4.76. It must be noted that the Department of Education in the Punjab has made very large and very rapid strides in improving the information base within the department with regards to the physical status of government fachties, enrollment levels, the attendance of staff and students and the distribution of needed supplies throughout the system. This has markedly improved the functioning o f the department, and in this regard the education sector in the Punjab provides a very good model for other provinces. What remains, however, i s population-based information that covers those who are n o t in school at all or who are being served by the private sector (as i s presently being considered by the provincial government). 86 4.77. Finally, information collection i s an area in which the provincial government must play a leading role. As m o r e responsibilities of government are devolved to bstricts, tehsils, and unions, there are opportunities to learn from experience. There is also a need to make sure that such information i s objectively collected and analyzed. Both require a third party to collect and analyze the information-a very large, very important role for the provincial government. 4.78. Very little real progress will occur-in fact there will b e no way to determine if there has been progress at all-until better information i s brought to bear on improving services and their outcomes. So, while "information" i s the last section of this chapter, it is the first step that must b e taken by the Government of the Punjab. Box 12: International Lessons for Health Policy This chapter has emphasized the need for developing Punjab-specific data to monitor progress and, eventually, determine appropriatepolicy based on a clear understandingof the causal factors of health status. This is a medium to long term perspective and the value of data collected will become more easily discernable over time. Collection should start as soon as possible.In the meantime, some directionfor health policy are likely to be more successful than others based on internationalexperience and the insightsafforded by economic theory. These can serve as hypothesesto be tested as data accumulates or, alternatively, as a starting point for re-orientingpolicy. As far as allocating resources within the health ministry i s concerned, we can look at three broad categories of intervention. First is the set of core, populationbased, public health interventionssuch as sanitation, vector (pest) control, basic healtheducationconcerninghygiene or nutrition and immunizationor other specific infectious disease interventions. Second is the provision of primary health care - simple medical procedures.Third is the provision of more expensive, hospital based care, preferablyaccessed via referral from cheaper-to-runfacilities. From the perspectiveof the field of public economics, there are three main criteria by which to judge these types of interventions. The first is to increaseefficiency- that is, to use governmentto do things that privatemarketscan't do adequatelyeveninprinciple. This means identifyingthe most important "market failures" in the sector and making sure that public efforts - particularlypublic expenditure - address these problems. The second criteria for public policy is to improve equity Government can be a means to improvesocial justice and transfer benefits to poor people.Differentpolicies have differingeffects on the relativelywell-offand the poor and greater emphasis and resources can be given to the latter.The third criterion, not quite as standard in textbooks but an area emphasized in this chapter, i s the relative ease the public sector has in implementingdifferent policies. It i s important for government to be able to "make good" on the promises it makes to the people both for ensuring that public resources are not wasted andalso to bolster confidence and credibilityfor governmentaction. The question for settingprioritiesbecomes: how do the three main types of policies measureup on these three criteria? Corepublic heafth:The two characteristicmarket failures in healthare the large externalities associatedwith infectious disease and the exposureto financiallycatastrophic risk due to the absence of insurance markets. Core public healthinterventions almost always directly attack infectious diseases. The health sector gives some of the only examples we have of "pure public goods", that is, activitiesthat cannotbe sustainedina privatemarket becausetheir benefits cannot be limited to those who can pay for them. While the private sector can be contracted to perform such activities by governments, they will not perform them without government intervention.Some forms of pest control- sprayingfor mosquitoes, control of the rat or snail populations that spread disease - are good examples. Health education for basic hygiene and nutrition are also examples. Other activities that have large externalities includetuberculosis controlthat can limit the spread of a dangerousinfectious disease and of resistanceto standard drugtreatments, On efficiencygrounds, therefore, core public healthinterventionsrank highly. On equity grounds as well, infectious disease control looks good. While poor people suffer from almost everything to a greater extent than do the rich, the differentialis greatest for infectious disease. For example, in India the incidence of tuberculosisis seven times greater for the poorest decile than for the richest. For malaria it i s four times greater while for cataracts (as an example of a chronic, non-infectious, illness) the differential is only 30%. Any shift of focus away from infectious disease to other health problems hurts the poor. Finally,many of the activities that fall under the category of "core public health" are, arguably, relatively easy to implement.Some of them are "one shot" interventions such as swamp drainage or only require periodicvisits (say for immunizationcampaigns) rather than sustained presence of professionals in rural areas. On all three grounds, then, making sure that traditional public health activities are adequately funded is a hlgh priority While there has been much talk of late of the "health transition" in which the disease patternhas shifted from infectiousto chronicillnesses,it is importantto emphasize that for public expenditure,the priority remains-and will continueto remainfor some time in Pakistan-with core publichealth. 87 Box 12:InternationalLessons for Health Policy (continued) Primary health care; In contrast to core public health, maintaininga widespread network of simple curative care centers is more problematic.The "market failure" often associatedwith routine curativecare is "asymmetric information'', or, the fact that medical professionals know more about your illness than you do allows them to exploit this knowledge to their own advantage. The importance of this "supplier induced demand" is contested in the literature. Its main manifestationwould be over-use of medical care - not often raised as an issue for rural areas (though in urban areas in South Asia generally, this may be something of a problem).It should be rememberedthat there is a very large private sector in health care - almost all of it in primary care (or for treatments that couldbe done at primary level). Public provisionruns the risk of displacingwhatever private capacity does exist. In any case, the "efficiency" argument for primaryhealthcare is not particularlystrong. On equity grounds, primary care may do better but this is entirely an empirical question. Internationalexperience on who benefits from subsidies to primary health varies widely - sometimes with poor people benefitingmore than others, sometimes with the benefitscapturedlargely by the middle and upper classes. It is for this reason that more data needs to be collectedin Punjab- there are no consistent lessons from other countries. Fortunately, this can be measuredin one survey and one does not have to wait for the answer. It is on implementation capacity that primary healthcare is most vulnerable to criticism.Medical professionals doctors especially - but not hited to them - have better earningopportunities inurbanareas. Further, they are themselves usually urbanbornandbred and want the amenities (and educationalopportunities for their children) afforded by cities. In many countries it is difficult to fill posts in the more remote rural areas (for example, vacancy rates in Indonesiarange from nearly zero in the desirable province of Bati to 60% in the remote area of West Papua).Evenwhen positions are accepted, absentee rates from primary health centers are often very high.A recent study in Bangladeshfound that the absentee rate among doctors as measuredby surprise visits to facilities were as high as 74% in ruralareas (as opposed to a nation-wideaverage of 40%). This result has been replicatedin many countries and has found to be very highin SouthAsia generally.Runningprimary healthcenters is simply not an easy task. On all three grounds, then, the argument for large amounts of public spending on primary health is ambiguous and depends critically on country circumstances.Judged on the basis of neighboringcountries, the case for Pakistanas a whole i s still ambiguous - the performancein India and Bangladesh would lead to pessimism. On the other hand, Iran has been very successfulin implementingprimary care through the use of female village healthworkers. Why Iran has been successfulwhere India has not is unclear andgeneralizations are hazardous.Punjab-specificanalysis i s essential. Nospitabbased care: In contrast to primary health care which is murky on all three criteria, and to core public healthin which all argumentsgo the same way, hospitalbased care is clear on each criteria but they argue in conflictingways. Without health insurance, a market that is inadequate in many ways, and given the difficulty of runninga public health insurance program (which tend to be difficult to administer for the same reasons that private programs fail), subsidizedpublic hospitals may be the only recourse most people would have for protectionagainst catastrophic (financially) illness.The efficiency argument for public hospitals can be very strong. On the other hand, while the data on the distributionof benefits for primary health care varies from country to country - hospitalcare is almost everywhere utilized more by the well-off than the poor. To a large extent this is due to the necessity of hospitals being in urban (or semi-urban) areas where the very poorest do not live in conjunction with procedures that do not keep people from using public hospitals (such as policies of free care). Therefore, public subsidies go disproportionatelyto the well-off. I t is likely that runninga public hospital is easier than runninga diffuse network of primary care centers but this is a matter of speculationor indirectinferences. Evidence from the state of Andhra Pradeshin India suggests that a major source of job satisfaction for medical professionals is the ability to use their knowledge, confer with colleagues and have access to modern equipment. All of these are much more likely to be satisfied in hospitals and, in general, absentee studies confum that staff show up for work more in hospitalsthan in smaller facilities. 88 CHAPTER 5: IMPROVING PROVINCIALFINANCES FOR DEVELOPMENT 5.1. This report has provided a number of recommendations to help accelerate growth and employment generation in the Punjab. Some of the recommended measures simply entail stroke-of- the-pen changes in existing rules and regulations, and therefore do not have any significant budgetary implications. However, others-for instance, the development of city clusters-are likely to require substantial fiscal outlays over the medmm term. When developing further the recommended program of work in such areas, the government should also strive to conduct a comprehensive costing exercise to ascertain the addtional financing that may be required for these purposes. While substantial additional public investment wdl likely be needed over the medmm term, the provincial government should also actively explore the possibility of pursuing innovative public-private initiatives (Box 13). Table 5.1: Punjab FiscalAccounts: 1994/95 -2004/05 (Expresssed aspercent of provincial GDP) lYY4/95 199996 1YY6/Y7 1997/Y8 1998/99 199Y/00 2000/01 2001/02 2002/03 2003/04 2 0 0 4 / 6 Actuals Actuals Actuals Actuals Actuals Actuals Actuals P.Actua1 P.Actual Revised Budget Total Kevenue (including Grants) 6.9 7.5 7.I 5.7 5.6 6.1 6.I 5.6 6.4 6.9 6.8 Federal tax assignments 5.4 6.0 5.9 4.4 4.1 4.3 4.6 4.3 4.5 4.3 4.2 Provincial revenue 1.5 1.5 1.2 1.3 1.5 1.8 1.6 1.3 1.8 2.6 2.7 Provincialtaxes 0.6 0.6 0.7 0.6 0.6 0.7 0.7 0.5 0.6 0.8 0.7 Direct tuxes 0 2 0 2 0 2 0 2 0.2 0.2 0.2 0.2 0.2 0.3 0 2 Urban immovable property tax 0 0 0 0 0 1 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 Agriculture income tax 0 0 0 0 0 0 0 0 0.1 0.1 0.0 0.0 0.0 0.0 0 0 Registrationfee 0 0 0 0 0 0 0 0 0.0 0.0 0.0 0.0 0.0 0.1 0 1 Land revenue(tax) 0 1 0 1 0 1 0 1 0.1 0.1 0.1 0.1 0.1 0.1 0 1 Taxes on profes,trades and callings 0 0 0 0 0 0 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 Others 0 0 0 0 0 0 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 lndirecr luxes 0 5 0 5 0 5 0 4 0.4 0.4 0.5 0.4 0.4 0.6 0 5 Motor vehicle tax 0 1 0 1 0 1 0 1 0.1 0.1 0.1 0.1 0.1 0.1 0 1 GIST on Services 0 0 0 0 0 0 0 0 0.0 0.0 0.1 0.1 0.1 0.1 0 1 Stamp duties 0 2 0 2 0 2 0 2 0.2 0.3 0.2 0.2 0.2 0.3 0 2 Other 0 1 0 2 0 2 0 1 0.1 0.1 0.1 0.1 0.1 0.I 0 1 Non-Tax Revenue &Grants 0.9 0.9 0.6 0.7 0.9 1.1 0.9 0.7 1.2 1.6 1.7 Local Government Revenue 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.2 Total Expenditure 7.7 7.Y 6.7 6.5 6.3 6.8 6.3 5.Y 6.2 7.1 6.7 Current Expenditure 6.1 6.5 5.7 5.1 5.0 5.5 5.3 5.0 5.1 5.5 5.1 General Administration 0 9 1 4 1 0 0 7 0.8 0.8 0.9 0.7 0.7 0.6 0 5 Law & order 0 6 0 6 0 5 0 5 0.5 0.5 0.5 0.5 0.6 0.8 0 7 Social services 2 3 2 2 2 0 2 0 1.8 1.8 1.8 1.8 1.9 0.4 0 5 Education 1 8 1 8 1 6 1 6 1.4 1.4 1.4 1.4 1.5 0.1 0 2 Health 0 4 0 4 0 3 0 4 0.3 0.4 0.4 0.4 0.4 0.2 0 3 Other SocialServices 0 1 0 1 0 1 0 1 0.1 0.0 0.0 0.0 0.0 0.0 0 0 Economic Services 0 6 0 7 0 7 0 5 0.5 0.5 0.5 0.5 0.5 0.4 0 3 Irrigation 0 3 0 4 0 4 0 2 0.2 0.2 0.2 0.2 0.2 0.0 0 0 Other Economic Services 0 4 0 3 0 3 0 3 0.3 0.3 0.3 0.3 0.3 0.0 0 0 Community services 0 3 0 3 0 2 0 2 0.2 0.2 0.2 0.2 0.2 0.1 0 1 Subsidies 0 1 0 1 0 1 0 1 0.2 0.2 0.2 0.1 0.1 0.1 0 0 Interestpayments 1 3 1 2 1 1 1 0 0.9 0.9 0.9 0.8 0.7 0.6 0 5 Grants & Investments 0 0 0 0 0 0 0 0 0.0 0.5 0.4 0.3 0.3 2.6 2 5 Unallocable 0 0 0 0 0 0 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 Development Expenditure 1.6 1.5 0.9 1.4 1.3 1.4 1.0 1.0 1.1 1.7 1.6 Fiscal Deficit -0.8 -0.4 0.4 .0.8 -0.7 -0.8 -0.2 -0.4 0.1 -0.3 0.2 Financing 0.8 0.4 -0.4 0.8 0.7 0.8 0.2 0.4 -0.1 0.3 4.2 Foreign loans 0 3 0 3 0 2 0 4 0.5 0.5 0.3 0.3 0.3 0.9 0 7 Loans from the federal government 0 2 0 1 0 0 0 0 0.1 0.1 0.0 0.0 -0.1 -0 7 -0 5 Domestic borrowing &cash balances 0 3 0 0 -0 6 0 4 0 2 0 2 - 0 1 0 1 - 0 4 0 0 -04 Source: Provincial Finance Accounts (varioub issues) and Budget documents (Various issues), and staff estunates 89 Box 13: Innovative Public-Private Partnerships: The Bangalore Agenda Task Force Example Bangalore, the capital of I wl 6 U E .-5 m a L a2 0 g P x i3 0 c) V c E rr 0 L P Q) z x- 2.. In a2 P $ W P `010 `or- w P w m P ""N 2 " " 8 8- 0 3 " c w Z" 2" O N N - Y - Z 2 0- m" W P W r. w w a w l - U w w l - VI10 vim w 1 0 w O - 2 2 N I? w-m r 0 3 ;. R 2" N x 0 N P W l- m x s w" z2- z m 10" F N a $1 ae 4 ad - f j e 3 z W 0 vi 3 N m vi m 0 m 0 N m 0 00 vi d 8 s W m i- 2 W 0, - d W m i- % If: m 2 2 Q' t-" N m 4 3 If: 3 N vi i- W i- W 3 W 3 3 W 3 9 3 W m W N T - i- m W iz .s 0 vi W N d ic LT m - 0 i- W m m It m" s0 e 2 0N 2 2 r P 2 P -.-I 5 D -0 ! g -.-E n cm P L L a! l/i .- Y Y Bh n Q - - i Q r 4 - - _ e - - L. 0 - '0"f m - c " W U i m 9, c or- u w N cU a m m m ' 0" b : 0 m W m C c -- m o c m*m m 2 - m oD - r N P v .-m 2 0 - m m m b r u b l n m P - - m w l n w m o o w m w W N W zm3 -" e e e " b e e e c a m 0 m m m m m m P - c a m 2 2 2 2 2 2 H m W 2 $1 2 m m - N 3 m t- 2 m m 2 51 e m e m - - - * m a .-P d m r i b w w w o o m - N b t - " N " N m m m q m a d m m m m c r m m m m P o o w v m m m m - r i m m w m v w m - r . -mr . w m m m m b 2 m r i N r i m ~ b m bm w m m w o e e e ~ m - N m eb e e m w r - m m m w m w e e e - r i m m m m 0 - N 2 2 2 2 2 2 0 0 0 0 0 0 " N N - N " N t i l o t i " N N N N C~ P M Z M u m t i " N m m r N N C "" c e w m t i w w t i m r - 0 m0m m 3 m m t i w m r t i t i v w m v , - - 0 w m z w o 1 u u t i N w utiv, m m k - m m i c m - w c e w c e w m v " E - m u . 9223 2325 r - m o 5 5 : t i m c r P e e m w r - " 0 - 0 0 0 ~m t i m - e e e m m c h a : h c h a m r - w c h 2 s 2 2 2 2 0 0 0 0 0 0 N " S - 4 ' 9 0 9 o o v i b ? ? 9 2 2 2 I cB -, - 4 6 vi Q\ i I $ 6 6 m w w ` " 2 i o-0 0 - d w cb? ? ? rl b b FB ` 9 ? c ? y o e m m w w "I?? W z z z I c v i w q : vi 1: ? ? ; 3 22: 00-b "- 0 w v i w " N " N f I c m m - 2 2 % : ?- 009 ?- " N p'b- m m b w D N 0 b b b - 8 P~Q m 80 Z0f0i "N 2 2 2 " ? - 2 2 2 tt"! " ' 9 1 2 2 2 '?To9 " 2 " i D w P - 0 9 - 3 a ? a 1 1 a a > > i ! ! I !! I .- c W c L B VI 4h B .-mW c L I L - U C m 5.- c m 0 Y U Ua a2 W x 0 m 0 N 0 - 0 0 0 7 8 N s 0 m 2 I 0 0 6 a m m c 0 w - p 1 z z z U S m x.. m m m - m m W "" m N m W ~ vN i r - - f m 0 W -" N N r . Z QmI I ' . d Nvi 09 m N h 2. t 2 h v 8 2 2 N - N sr oN 0N - o m 5 2 3 w o w w w w w w r - b m - e e e m w r - b m w " o \ 2 2 2 w z me m w " m w u - vi" P *-i a E a 3 E Q) W E 0 .-a E 2 Y v1 u ei2 Y aE Y 0L Y E" LI 0 .-0 C Y V W a a E 51 E m $a2 m : .. \o 5 Q) c" L s CI 0 M 3 31 m 2 .. 0 0 0 ` 0 -t m a . 0 - N d M 0 0 "1 64 :: 0 .. n m c h a E W n h -i? U e"3 2 W E mW m d N 00 m - 0W mW r. m m m N 0 0 m 0 - m W m W P N m m r. W IT m 52 dm W m 0' r.W m e" L d L 0 0 0 0 0 0 0 0 0 0 0 0 L " v l v l v l m v l v l v l v l v l v l v l q - 0 w i m m m m m m m m m m m m x m - b m * * w a r 4 d d m m w m 2 " o ` 8 6 " ""PI m m o o m " N " N N " N N ~ % S T ; -"r"N o w 0 i m e b m m r . " N m lbn m m " N l n l n w m" oo" m" w- w" w" oo" m" m" m" m*00" r- 0 m W vi r- r- C m 09 - N m" N Z" 3 C W m C m 2 3 Q" N m vi 0 d d W 0 N N W vi% m iz m vi z r- w 3 N Q ) N N N hlz hl r- 00 r- m m N m 3 N m m. W d m m m d d N W C N N - m m W vi m m vi W KiN N N N W d d r- W W r - r - r - r - r - r - r - r - r - N N N N r - r - r- r- N " " N N"" N P- e e e e m d vi W m m m N d mvi 2 2 2 - - - w m - I - m o v i - 0 = * I - "I- "I- -* "* "" 2 2 1 8 Z 2 m m w IvTi 2 2 v i 0 ~ O W " 0 ' 0 0 " " m' 0 ' 0 " 2 0 2 - a m o" 2 m - 8 m m m w w w 0 m 09 m m m 04 2 2 2 2. Y 9 U M E L m z 5 0 c 0 W 0 m E e E 2 2 2 2 2 2 " 2 " 2 2.. 2 L E W v i m 0 N - 0 m m b L? L". Y w w m N " - b e m m m ' 0 ' 0 0 w a m m a w r - w m = 2 2 w a r - r - w 2 2 2 m w w w m m w a w 2 2 2 o 0 = r - m w r - o m 2 ' 0 ' 0 - - - m o o r -b o o m ""m N r - m v l w w b r - 0 "N' O NO 2 2 2 r-r-r. r.r-r- r-r- r- r - r - r . r - r - r . 3 8 eE 2 E 3 .s 5 e3 w w W w w w W W w w w W W awl10 rm- m r - m m a c c i i o 0 0 0 "" v? '"" " N c a r - r - e e e "* a a a - m m 2 2 2 0 0 0 2" 8 0 0 0 Ir 0 0 0 0 m m m m m m b d d b d d co L m m m R 3 3 m m m 3 3 . 3 m m m a" fvj e e e - m m 0 - N 0 0 0 0 0 0 " N w h o * * w r - m w w w r - m m w ' 2 2 w w w VI" 09 09 w w w W" m m m " N e e e w m o m m m r - w m 2 2 2 * - - - 0 2 0 2 0 0 = = - w w 2 2 2 m o m O b W e e L " 00m m m m m m - 0 C O N b - b b - 0 w w w 0" w w \ o c o c a 0 0 m m m w c a m 0 0 W W m w m 01" m m m m c o o o c 0 b b b b b b b b b b b b x.. 0 ! \ C i I F I? i F i r F c t F J 010 3 2 NN P- N" 2 a : r r vi 8 a". In .-0 e r -m - 2 % P L -~m 3 P - 2 P W' P E. P" W m u . m l n z. % - -" m c Z" zm u.- 8 % a ?" N W N m m r - r - - ' m - I. z P - ' - N N v i P ' w m 0 L In e 9 I m - m m m N W 4: 6 3 P 6P 5 e m w - 2 " " - R m v- i w z. g 5" Q v i N - ItC ,m 2 m m P x m 2 0 1" m P P " - t P N Q F .* IN. V P n w 'wo m Q m N m 2 c c B 5 c anm 1 3 5 3 E n -m W ' n 1 P-r- e r- N m 0 * c 00 m t - t- * w - -0 a r- N m m m ;v;zz 3 2 0 ' c? m o m N m m m m ffiffi 2 o m r- m u N w m m c a m 0 1 r- m m u m b Q p ' vi x 8 lL w m o m P - m o m 2 l o wx m p I B L W v, W 0 P- m N W O00 vi vi 0W m "0 dm N v, N v, N 2 O N d m d I.- 0," : 2 G. N 0' 0' N m W R N a m. m ? m : r. m T N m 2 2m x :: oa z m r j a '0 00 m m '0 N m vi N m d 00 a N G. N m m 0 a N - 01 0 2 i d IA r. r. W00 s 4 0 :N s0m v = E.- .-=' c v) L .-C O b s w W b b O N b 0 - m W O W P - 0 0 O u O w " u m w w N .- N 2 % - W m l c i b w m " 120. -w mo mb ww N m I * b P I N O l n v w h -.- h s E .- h Tii m - 0 w u N P I W 5 E- .-m m m w w b W I N N m l n l c i m N u m N s 2 zE a .-C .-C 3 XI r4 0 I. u00 m P - w w % % w w 2 P - P - NP- o " 0 c .-B L .-C .-C .-C .-C n a --.. m 2E a h 3 3 i P ' P ' 3 - N "- .. P *w m C e a P h 2 .. w w d m ~ ~ t - - m - N c N c O m 2 NNZ l -wJm m q , d m w e - N vr m " O O m m m o m m i i N- 3 e m - .-C a .. c rh U E m - O ~ N t . m m w 0 . N m ~ - O m w d d m.o.m.o.m.m.0.m N O u m t . N ~ 0 . P.IIc-4" P1woOc II m d m w m m N " N " -- -ozP4t.o " _ 0 e < 0 a D a 8 m m O N l n a m 3 o m w m a m o m- 0- N N N N N N N v, Ici N o\ 3 t- a t- a N 0 a wm a m t- t U 4 rr) %9 c E 9 a .-.- 10 m o o w - W 0 W" g o o 0 -m L 0 m m c c o aa8 N Y 2 w o o 0 C = Y i 2 .-R Y d - c o o v l 0 04 m W O O b 2 - N d -- z o o m m N 2 .. O C O d 2" d OD CI E .- M v1 5 v) EXPLANATORY NOTES ON MANUFACTURINGINDUSTRIES SECTION This Section includesstatistics on the following topics:. i) Monthly Survey of Industrial Production and Employment ii) Annual Census of Manufacturing Industries(CMI). iii) RegisteredFactoriesandtheirEmploymentbyDistrictandIndustry, iv) IndustrialCredit. MONTHLY SURVEY OF INDUSTRIALPRODUCTION AND EMPLOYMENT. Statistical tables includedin this section containdata on productionand employmentfor the years 1991-92to 2000-01 and monthly data for 2000-01 as well in respect of 34 items manufactured by 21 selectedlarge scale industries in the Punjab. This informationhas beencollectedthrough Bureau's Monthly Surveyof IndustrialProductionand Employmentwhich was started in October, 1971.Industry- wise proformae are used for the purpose. Datafrom non-respondingfactories is collected by Bureau's 0 Coverage is complete in respect of Cotton Cloth (mill sector), Woollen and Worsted Cloth, Jute Textile, Sugar, Cigarettes, Cement, Caustic Soda, Sulphuric Acid, Fertilizers, Bicycles, Soft Drinks and Juices.Coverage for the remaining manufacturedproduct 1 Cotton Yarn, Vegetable Ghee and Paper and Paper Board . It excludessome units. 2 Woollen Yarn, Knitting Wool and Blankets:It excludes some units. 3 LeatherTanning: All large Tanneries are covered. Footwear: 13 large and medium size factories including Bata, Service and Raja are being covered. Production of these factories was about 98% (of Footwear 4 Production) of registeredunits of the Province. Steel Re-rolling:- According to Bureau`s Establishment Survey o f Steel Re-rolling Industry conducted during 1975-76, about 90% of the Production is being 5 covered in the current survey. 6 Diesel Engines:-23 Factories whose production i s about 75% of the total productionof Diesel Engines are being covered. 7 Electric Fans:-All large and medium size factories are covered. Stem for 1morovement:-To comolete coverape in resoect of Woollen Textile, Leather Tannine. Footwear, Electric Fans, SteelRe-rolline Industries Efforts are made to keep the list of factories up-to date in consultationwith Central Board of Revenue, PICIC, ADBP and Directorate of Labour Welfare, Punjab. Scope of the Monthly Survey of Industrial Productionand Employmentis being extendedto some more Industries. ANNUAL CENSUS OF MANUFCTURING INDUSTRIES(CMI) This section contains tables on Summary Statistics (Fixed Assets, Employment, EmploymentCost, Industrial Cost, Value of Production, Gross Value Added of registeredfactories reporingfor the annual Census of Manufacturing Industries for the years 1988-89 Federal Bureauof Statistics:- a) Planning of Census in collaborationwith Provincial Bureaus of Statistics and Directorates of Industries. b) Preparationof Questionnaire, Tabulation Plan and Editing Instructionsin consultation with all concernedFederalMinistries, Provincial Departmentsand c) Preparationand Publicationof CMI tables at national level. d) Maintenance o f an up-to-date list of manufacturingestablishments. e) Printing and Mailing out of Questionnaires Directorate of Industriesand Mineral Develooment,Puniab:- a) Notificationof the questionnaires. b) Collection of CMI returns. c) Preliminary editing of CMI returns. Bureau of Statistics,Puniab:- a) Statistical examination o f CMI retums b) Computer processingo f CMI retums. c) Conducting ofNon-responseSurvey for estimation o f data for the establishmentsnot reporting for the Census Scope & Coveraee: All factories, carryingon manufacturing or repairing activities and registeredor eligible for registration under Section2 (J) or 5 (i)under the FactoriesAct 1934(Amended in 1973),are covered under the Census. It, however, excludes Defence and other Definitions: Importantterms used in the section are defined as under:. i) REGISTERED FACTORIES: An establishmentis said to be registeredunder Section 2 (J) of FactoriesAct 1934, (Amended in 1973)iften or more workers are working on any day of the precedingtwelve months and in any part o fwhich manufacturingprocess is being carried on with or ii) GROSS VALUE OF FIXED ASSETS: It includes the grossvalue of land, building, plant, machinery, transport and other fixed assets i.e.equipment, furniture etc., as at the end of the year iii) EMPLOYMENT: It is the total averagedaily employment duringthe year computedfrom the monthly employment datareportedin CMI returnsadjusted for contract labour iv) EMPLOYMENT COST: It includeswages and salaries, cash and non-cashbenefits and the amount paidto contract labour. v) INDUSTRIAL COST: It includes cost of raw materials, fuel and electricity consumed and paymentsmade to others for repair, maintenanceof building and machinery and processing of establishments materials. vi) GROSS VALUE OF INDUSTRIAL PRODUCTION: It includesvalue o f products and by-products, sale-proceeds from industrial waste, receipts for industrial work done and services renderedto others, sale of electricity and value o f ' net work in process'.The value of products and by-productsincludes vii) GROSS VALUE ADDED This is worked out by subtractingindustrial cost from the gross value o f industrial production. Limitationsof CMI The mailing list used for the census may not be up-to date. The reporteddatacontained reporting errors and inconsistencies, which were correctedon the basis of available supplementary information or/andpast data. Census results become available with atime-lag of 2-3 years due to the time taken for collection, editing and tabulation processes. Size and nature of non-responsevaries from year to year. Since non-responsesurveys could not be conductedfor all the years, the Censusresults were not adjusted for non-response,they may not be comparable from year to year. The Bureau of Statistics has, during recentyears, taken the followingsteps to improvethe Census Results:- List of registeredfactories is beingkept up-todate. Reporteddata is subjectedto manual as well as computer statisticalediting, before final tabulation.Missing information estimated The cases, where reportingerrors cannot be corrected without changing the reported information significantly, managementso f the concernedunits are asked to reconcile the discrepanciesobserved intheir data. Correctionsare made in the light of their Efforts are beingmade to complete editing and tabulation as early as possible, after receipt of returns from Directorateof Industriesand Mineral Development, Punjab.