Report No. 25665-PK Pakistan Public Expenditure Management Strategic Issues and Reform Agenda (Volume I) January 28, 2004 Poverty Reduction and Economic Management Sector Unit South Asia Region Document of the World Bank CURRENCY EQUIVALENTS Currency Unit = Pakistan Rupee US $1 = PKR 57.8 FISCAL YEAR July 1-June 30 ACRONYMS AND ABBREVIATIONS ADB Asian Development Bank MOF Ministry of Finance ADF Asian Development Fund MTBF Medium-Term Budget Framework AGP Auditor General of Pakistan MTIP Medium-Term Investment Plan AGPR Account General Pakistan Revenues NDP National Drainage Program CBR Central Board of Revenue NEPRA National Electric Power Regulatory Authority CCAs Canal Command Areas NFC National Finance Commission CFAA Country Financial Accountability Assessment NHA National Housing Authority CGA Controller General of Accounts NRB National Reconstruction Bureau CWIQ Core Welfare Indicator Questionnaire NSS National Saving Schemes DFID Department for International Development (UK) O&M Operations and Maintenance DPCO Debt Policy Coordination Office OFWM On Farm Water Management DSCs Defense Savings Certificates OGDC Oil and Gas Development Corporation EAD Economic Affairs Division (in MOF) PAAS Pakistan Audit and Accounts Service ESR Education Sector Reform PACs Public Accounts Committees ESW Economic & Sector Work PFC Provincial Finance Commission FATA Federally Administered Tribal Areas PIHS Pakistan Integrated Household Survey FBS Federal Bureau of Statistics PIA Pakistan International Airlines FIP Financial Improvement Plan PIBs Pakistan Investment Bonds Project for Improvement in Fiscal Reporting and FPSC Federal Public Service Commission PIFRA Auditing GOP Government of Pakistan PR Pakistan Railways GST General Sales Tax PSDP Public Sector Development Program HIPC Highly Indebted Poor Countries PTC Pakistan Telecommunications Corporation ICG Internal Cash Generation PV Present Value IDA International Development Association RICs Regular Income Certificates IMF International Monetary Fund SAP Social Action Program IPPS Independent Power Producers SBP State Bank of Pakistan I-PRSP Interim Poverty Reduction Strategy Paper SMCs School Management Committees IRSA Indus River System Authority SOEs State Owned Enterprises KESC Karachi Electricity Supply Company sscs Special Savings Certificates LBOD Left Bank Outfall Drain WAPDA Water and Power Development Authority MDGs Millennium Development Goals Vice President: Praful Patel, SARVP Country Director: John W. Wall, SACPK Sector Director: Sadiq Ahmed, SASPR Sector Manager: Ijaz Nabi, SASPR Task Managers: Paul Wade, SASPR; Hanid Mukhtar, SASPR Parvez Hasan, Consultant TABLE OF CONTENTS EXECUTIVE SUMMARY ........................................................................................................... i INTRODUCTION...................................................................................................................... xvii CHAPTER 1: EXPLORI N ........... 1 G THE DIMENSIONS OF THE CRISIS IN PUBLIC EXPENDITURES Introduction................................................................................................................................................ 1 Historical Perspective ................................................................................................................................ 2 Restructuring and Refocusing an Extended Public Sector ....................................................................... 11 K e y Gaps in Public Goods ....................................................................................................................... 13 Health ............................................................................................................................................... 13 Education ........................................................................................................................................... 14 Irrigation ........................................................................................................................................... 14 Rural Water Supply ........................................................................................................................... 15 Transport .................................. .................................................................................................... 15 Power ................................................................................................................................................. 16 The Strategic Agenda for Public Expenditure Management .................................................................... 16 CHAPTER 2: ENSURING FINANCIAL SPACE .................... DISCIPLINE AND CREATING FISCAL 18 Introduction.............................................................................................................................................. 18 Tum Around in Debt Position.................................................................................................................. 18 Baseline Scenario ..................................................................................................................................... 23 Slowing the Rate ofBorrowing .......................................................................................................... 24 Increasing Revenues ............................. ............................................................ 25 Public Enterprise Losses and Other Conti ............................................................ 26 Defense Spending ........................................................................................... .............................. 30 The Cost ofBorrowing ................................................................................... .............................. 31 Domestic Debt Costs ......................... ............................................................................................ 32 The Cost of External Borrowing ........ ............................................................................................ 34 Structural Reforms and Growth Revival ........................................................................................... 36 Alternative Scenarios ............................................................................................................................... 39 Low Case Scenario ............................................................................................................................ 40 High Growth Scenario .............. ................................................................................................... 42 Conclusions .............................................................................................................................................. 43 PUBLIC SECTOR P CHAPTER 3: REORIENTING RI O RI T I E S ........................................................ 45 Introduction.............................................................................................................................................. 45 Fundingthe Poverty Reduction Strategy ................................................................................................. 46 Remedying the Neglect o f Education ............................................................ Improving Health and Population Outcomes .................................................. Health Sector Reforms ............................................................................... The Way Fonuard .................................................................... Accelerating Water Resource Development ............................................................................................ 55 Transforming the Power Sector ............................................................................................................... 61 Conclusion ............................................................................................................................................... 64 CHAPTER 4: IMPROVING EFFECTIVENESS ................................... OF PUBLIC EXPENDITURES 66 Introduction.............................................................................................................................................. 66 Responsiveness to Citizens ...................................................................................................................... 66 Reaching the Poor .................................................................................................................................... 66 Improving Budgetary Processes............................................................................................................... 67 Recent Initiatives ............................................................................................................................... 68 Medium-Term Budgetary Framework ............................................................................................... 68 Re-invigorating the Planning Processes ................................................................................................... 73 Strengthening Expenditure and Program Monitoring ....................................................................... 75 Redefining the Role o f t h e State .............................................................................................................. 77 Reducing the Role o f the Government in Producing and Providing Private Goods and Services .... 78 Furthering Devolution ....................................................................................................................... 82 Federal-Provincial Responsibilities ......................................................................................................... 87 Inter-Government Financial Flows .......................................................................................................... 87 Improving Govemance and Fighting Corruption..................................................................................... 88 Strengthening Financial Accountability ................................................................................................... 91 CHAPTER 5: STRENGTHENING THE CIVIL SERVICE ................................................................. 94 K e y Issues ................................................................................................................................................. 94 Structure. not size. i s the problem ............................................................................................................ 95 Bottom heavy ..................................................................................................................................... 97 Rigid .................................................................................................................................................. 97 Mismatch Between Skills and Rewards ................................................................................................... 98 Persistent Patronage ............................................................................................................................... 100 A Pragmatic Approach ........................................................................................................................... 101 Seize the emerging consensus.......................................................................................................... 101 Managing the Transition ........................................................................................................................ 102 Address pay ..................................................................................................................................... 102 Complete devolution ........................................................................................................................ 103 ............................................................................................................... 106 REFERENCES/BIBLIOGRAP Annex A: Debt Reduction and Management Strategy Annex B: Medium-Term Framework Annex C: Simulations Based o n Dynamic Benefit Incidence o f Public Education Expenditures in Pakistan STATISTICAL ANNEXES TABLES Table 1.1: Non-Interest and Non-Defense Expenditure for Selected Countries ......................................... 1 Table 1.2: Investment as a percentage o f GDP ........................................................................................... 3 Table 1.3: Fiscal Indicators as a percentage o f GDP .................................................................................. 5 Table 1.4: Consolidated Federal and Provincial Total Expenditure in 1999/00 Prices .............................. 7 Table 1.5: Consolidated Federal and Provincial Total Expenditure ........................................................... 7 Table 1.6: Planned Privatization o f State Enterprises, November 2002-June 2004 .......................... :......12 Table 1.7: Pakistan in Comparison with Regional Countries in Health and Population Outcomes .........13 Table 1.8: Present Status o f Health & Population Outcomes and MDG Targets for Palustan .................13 Table 1.9: Public Expenditure on Education for Selected Counties (% o f GDP) .................................... 14 Table 1.10: Canal Water Diversion ........................................................................................................... 14 Table 1.11: Additional Coverage Required for Rural Water Supply (% o f population served) .................15 Table 2.1 : Extemal and Public Debt Burden ............................................................................................ 19 Table 2.2: Summary o f K e y Macroeconomic Variables in Baseline Scenario. 1998/99-2006/07 ........... 23 Table 2.3: 3ase Case Scenario: Principal Budget Aggregates .................................................................. 23 Table 2.4: Losses o f State Owned Enterprises (Rs. Million) .................................................................... 30 Table 2.5: ?ahstan: Real Borrowing Cost on Public Debt. FY1996-2002 .............................................. 32 Table 2.6: Summary o f K e y Macroeconomic Variables in L o w Case. 1999/00-2006/07 ........................ 4 1 Table 2.7: Palustan: L o w Case Scenario. Principal Budget Aggregates ................................................... 42 Table 2.8: Summary o f Fiscal Space Alternatives over the Medium Term .............................................. 43 Table 3.1 : 1ESR Targets by Sub.Sector. 2001-2005 .................................................................................. 48 Table 3.2: 1Water Availability, 1960/61 - 1999/00 ................................................................................... 56 Table 3.3: 1Electricity Tariffs (WAPDA System), FY96-2003 ................................................................ 64 Table 3.4: 1 [llustrative Allocation o f Additional Fiscal Space ................................................................... 65 Table 4.1 : 1Functional Responsibilities o f Districts, Tehsils and U n i o n Councils .................................... 84 Table 4.2: 1 Key Governance Indicators in South Asia ............................................................................... 91 Table 5.1: (3 v i l i a n wage bills in Palustan (current) by comparison with selected counties ................... 97 Table 5.2: < Sanctioned L Posts by Province (selected years) ........................................................................ 97 Table 5.3: 1 [mpact o f t h e January 2002 Pay Reforms ................................................................................ 99 Table 5:4: 1 Proportion o f monetary allowances in federal civil service wage bill ................................... 102 FIGURES Figure 1.1: Trends in Selected Expenditures in 1999/00 Prices ............................................................... 6 Figure 1.2: Trends in Selected Expenditures as a percentage o f GDP...................................................... 6 Figure 1.3: General Administration Expenditures in 1999100 Prices ....................................................... 8 Figure 1.4: L a w & Order Expenditures in 1999/00 Prices........................................................................ 8 Figure 1.5: Education Expenditures in 1999/00 Prices ............................................................................. 8 Figure 1.6: Health and Population Expenditures in 1999/00 Prices ......................................................... 8 Figure 1.7: Agriculture Expenditures in 1999100 Prices ........................................................................... 8 Figure 1.8: Irrigation Expenditures in 1999/00 Prices .............................................................................. 8 Figure 1.9: General Administration Expenditures as % o f GDP .............................................................. 9 Figure 1.10: L a w & Order Expenditures as % o f GDP............................................................................... 9 Figure 1.11: Education Expenditures as % o f GDP .................................................................................... 9 Figure 1.12: Health and Population Expenditures as % o f GDP ................................................................ 9 Figure 1.13: Agriculture Expenditures as % o f GDP.................................................................................. 9 Figure 1.14: Irrigation Expenditures as % o f GDP ..................................................................................... 9 Figure 1.15: Spending o n Education (Per Facility) ................................................................................... 10 Figure 3: 1: Budgetary Expenditure on Education................................................................................... 49 Figure 3.2: Per capita Benefit Incidence of Primary and Secondary Education ..................................... 50 Figure 3.3: Distribution o f Public Expenditure for Primary and Secondary Education............................. 50 Figure 3.4: Benefit Incidence o f Public Primary Education (Urban vs Rural) .......................................... 50 Figure 3.5: Benefit Incidence o f Public Secondary School (Urban vs Rural) ............................................ 50 Figure 3.6: Public Sector Expenditures ...................................................................................................... 57 Figure 3.7: Public Sector Expenditures as % o f GDP ................................................................................ 57 Figure 3.8: Public Sector Expenditures as % o f total PSDP ...................................................................... 57 Figure 3.9: Depletion o f Existing Storage and Planned Augmentation ..................................................... 59 Figure 4.1 : Decision Tree for Identifying Core Governmental Tasks ........................................................ 80 Figure 4.2: The 2-Step Logic o f Decentralization ....................................................................................... 82 Figure 5.1 : The M a i n Components o f Public Sector Employment in Palustan.......................................... 96 Figure 5.2: Composition o f Public Employment ......................................................................................... 97 BOXES B o x 1.1: Privatization................................................................................................................................... 4 B o x 2.1: Refining the Debt Strategy .......................................................................................................... 21 B o x 2.2: Determinants o f the Rate o f Public Debt Accumulation ............................................................. 22 B o x 2.3 : Contingent Liabilities .................................................................................................................. 28 B o x 2.4: Beneficiaries o f National Saving Schemes ................................................................................. 33 B o x 2.5: Growth Revival and Poverty Reduction: Private Investment and Public Spending ................... 38 B o x 3.1 : The Rich-Poor Gap and Public Expenditure on Education ......................................................... 50 B o x 4.1 : Budgetary and Financial Management Reforms ......................................................................... 69 B o x 4.2: Steps in Preparation o f an MTBF ............................................................................................... 70 B o x 4.3: DFID Assistance to Strengthen the Medium-Term Budget Framework ..................................... 72 B o x 4.4: Planning Processes: Sources o f Weaknesses ............................................................................... 74 B o x 4.5: Options for Public Financing o f Private Delivery o f Education.................................................. 79 B o x 4.6: Public-Private Partnership in Tuberculosis Control .................................................................... 81 B o x 4.7: Corruption in Palustan ................................................................................................................ 90 B o x 5.1 : Monetary Allowances and in kind Benefits .............................................................................. 100 B o x 5.2: Incomplete Devolution .............................................................................................................. 105 ACKNOWLEDGEMENTS This report was prepared by a core team including Parvez Hasan (Team Leader), Hanid Mukhtar, Paul Wade and Nicholas Manning. Shahid Javed Burlu and Mohsin Khan served as Peer Reviewers. The field work for the report was done in three missions during the period October 2002 t o February 2003. The basic draft o f the report was completed in early M a y 2003 and shared with the government. Discussions with high-level government officials took place in late August 2003. The draft has been revised in light o f the government’s comments and has been updated to reflect information available up to middle September 2003, including macroeconomic data provided to the September IMF mission. Volume I1o n Accelerated Water Development was prepared by Usman Qamar and Walter Garvey. Sector contributions from Tahseen Sayed K h a n and Ambar Narayan (Education), Inaam ul H a q (Health), Rashid Aziz (Power), and Navaid Qureshi (Transport) are gratefully acknowledged. The report also benefited from advice given by Tariq Hussain on the c i v i l service, water and education sections o f the report. Shanza Khan provided valuable research assistance. Shaheen M a l i k helped with data problems and prepared the Statistical Annex. Ms. Fritschel contributed to the editing o f this report. Thelma Rutledge processed the report and was responsible for i t s production. During the preparation, the team worked closely with the government authorities. I n this regard, the team acknowledges the active and constructive participation from the Planning Commission, Ministry o f Finance, especially the Debt Office and the Budget Wing, and the Economic Affairs Division. The Bank team’s discussions with the key counterparts, Secretaries Finance, Planning, and Economic Affairs division, Director General Debt Office, and Additional Secretary Budget, at the various stages o f the work starting from the concept paper finalization were instrumental in clarifying the main issues and developing the analysis and the conclusions presented in the report. The team also benefited from discussions with the heads o f planning and finance departments o f the governments o f Punjab and Sindh. The team leader made a Power Point presentation in February 2003 on Ensuring Fiscal Discipline and Creating Fiscal Space to the government’s economic team, headed by the Federal Finance Minister Mr. Shaukat Aziz and including Dr.Ishrat Hussain (Governor, State Bank o f Pakistan), DrShahid Amjad Chaudhry (Deputy Chairman, Planning Commission) Mr. N a v i d Ahsan (Secretary Finance), Dr. Mutawaklul K a z i (Secretary Planning), Dr. Waqar M. Khan (Secretary Economic Affairs Division), Dr. Ashfaque Hasan Khan (Director General Debt Office), Javed S. M a l i k (Additional Secretary Budget, Ministry o f Finance) and W a j i d Rana (Joint Secretary, Ministry o f Finance). The Bank team benefited greatly from the discussion of the principal issues involved in creation o f additional fiscal space that followed the presentation. The team’s work o n civil service reforms was very much helped by extended discussion in a brainstorming session on the subject at the Establishment Division Those attending from the government side included Mr. Javed Hasan Ali (Secretary Establishment Division), Mr.Tariq Saeed Haroon (former Secretary Establishment Division), Mr. Tariq Sultan (Principal, Pakistan Administrative Staff College) and Messrs Mueen Afzal and Saeed Qureshi (former Secretary Generals Finance). Finally, the team thanks John Wall, Abid Hasan, Ijaz Nabi, and Sadiq Ahmed for continued support and guidance throughout the course o f preparation o f this report. EXECUTIVE SUMMARY 1. Pakistan appears t o be at an important economic turning point and there are good prospects o f sharply reversing the poor economic and social performance witnessed for more than a decade. The comprehensive and forcehl reform agenda, which the government has been pursuing during the last three years, and i s reflected in i t s debt and poverty reduction strategies, has already improved the financial position very considerably and has helped to revive growth. For the first time in many years, there i s n o threat o f a financial crisis hanging over Pakistan’s head. Looking ahead, the main challenges n o w are to increase economic growth from the current annual rate o f about 5 percent t o close t o a sustainable growth rate o f 6 percent or higher over the medium term, substantially reduce poverty incidence, and begin to narrow Palustan’s social gap, while maintaining macroeconomic stability. 2. The achievement o f Palustan’s economic and social goals would require a strong recovery in private sector investment supported by further improvements in general investment climate, deepening and widening o f the financial sector, better incentives for exports and direct foreign private investment to promote both technological change and export orientation. Whereas in the past aggregate investment levels have been constrained by resource availability, in particular l o w national savings, this has recently changed. Higher worker remittances, the sharp reduction in the government revenue deficit (before grants), higher external grants, lower external interest payments, and the reversal o f capital flight have all helped to push the national savings rate t o the average o f more than 19 percent o f GDP in FYs 2002-03, i t s highest level ever. A part o f the improvement--- compared to the national saving rate o f 14 percent o f GDP during FYs 1999-2000 - i s due to transient factors such as reversal o f capital flight and exceptionally high level o f external grants. Still, provided confidence in the currency can be maintained and the improvement in public savings continues, there i s every reason to hope that much o f the improvement in national savings rate will prove to be durable. Thus, at least from the point o f view o f resource availability, the stage i s set for sustained and strong increases in the level o f investment. The more important issues that are likely to influence future private investment are the investor perceptions about political stability, the new government’s commitment t o sustained economic reform and i t s ability t o bring about improvements in governance and delivery o f public goods and services. Non-economic factors such as the regional security situation and the threat o f terrorism will also continue t o have a bearing o n private investment trends. 3. While private investment must provide the bulk o f the growth in investment, it must be accompanied and supported by significant improvement in both the quality and level o f public spending to overcome the key gaps in the delivery o f public goods, including infrastructure. The l o w level o f human resources development in Palustan i s limiting productivity growth and hindering competitiveness. The possibilities o f increasing value added in traditional manufactured exports and developing new service and technology related exports would improve with higher levels o f s h l l development. The high economic costs o f port handling, road and rail transport, and electricity are also sources o f discouragement to the private sector. 4. Better and more effective delivery of public investment and public services in key areas must, however, contend with the reality, at least in the medium term, o f a slow-growing resource envelope for the public sector. Our indicative macroeconomic projections suggest, that even with a moderate but steady growth in government revenues, the level o f public spending would not likely increase from i t s present (and recent) level o f around 22 percent o f GDP during 2004-7 (compared with 25 percent in the 1980s and the first h a l f o f the 1990s), if the necessary further progress towards reduction in public debt burden has also to be ensured. Fortunately, there i s considerable scope for both more effective use o f public resources and restructuring o f public spending. 5. The economic and social reform program has been tackling the various constraints on progress towards a viable and vibrant public sector in a comprehensive and bold fashion for the last four years. There has been notable progress and distinct success in a number o f areas, such as reducing macroeconomic imbalances and public debt burden, accelerating the efforts towards privatization, and re- orienting public expenditures towards social services and poverty groups. A start has also been made in improving governance and strengthening o f institutional capacity, though the reform measures are yet to yield concrete results in terms o f improvement in delivery o f public services. 6. Against the above background, the objective o f this report i s to assist the government in exploring ways to enhance resource mobilization and to further intensify and strengthen public expenditure policies and management. The report focuses o n three aspects o f public expenditure management: the institutional reforms that are needed to make more effective use o f public spending, economic and fiscal policies that would make possible a steady and significant enlargement o f the fiscal space, defined as non-interest, non-defense spending (also excluding public enterprise losses) as a percentage o f GDP, and the issues in re-orientation o f public expenditure priorities to make optimal use o f the fiscal space that could become available. The scope of the work and analyses was shared and discussed with the government counterparts at different stages o f the task. Related work to strengthen public expenditure management and service delivery capacity at the provincial and district levels i s also underway. 7. The main finding o f the report i s that the next few years represent a period o f great opportunity for Pakistan to accelerate i t s economic and social development. Efforts t o strengthen governance, improve delivery o f education and health services, and last but not least, make economic and social policies pro- poor by making public institutions stronger and the civil service better paid and more accountable will be central to the public sector playing i t s role in increasing growth and reducing poverty. The relaxation o f the financial constraints in the public sector can strongly reinforce these efforts. With the maintenance o f the momentum o f economic and fiscal reforms and effective implementation o f the priorities outlined in the govemment’s poverty reduction strategy, there could be a sharp increase not only in the level but also the quality o f social and economic development spending over the next five years, in sharp contrast to the stagnation in outlays and their declining effectiveness during the last decade. The creation o f additional fiscal space, however, would depend critically o n hrther success in raising the level o f government revenues, reducing the cost o f government borrowing, limiting the public enterprise losses, and continued restraint on defense spending in relative terms. With policy slippages in these areas, the additional fiscal space could virtually disappear, making i t difficult to attain and sustain high growth and meet social goals. Even if the fiscal space i s created, it could be wasted either because o f unproductive expenditures or lack o f progress towards more effective use o f otherwise high priority spending. The challenge for the new government i s to build on the improvements in economic management o f the last few years and t o avoid mistakes that often led to very ineffective use o f public resources in the past. Challenges o f Improved Public Sector Management 8. That Pakistan under-performs other countries with similar per capita incomes in delivery o f basic social goods i s well lmown. What i s perhaps less commonly realized i s that the gaps in public goods delivery have been widening during the last decade especially in relation o f delivery o f services to the poor. Furthermore, the inadequacy o f public investment in key areas i s takmg a heavy toll in efficiency losses for the economy and hindering private investment. The depth o f the crisis in public spending in Palustan can also be gauged from the fact that consolidated public expenditures (excluding interest and defense) averaged little over 10 percent o f GDP in the late 1990s compared to the average o f about 15 percent o f GDP in the 1980s. Even though non-interest, non-defense spending rose sharply in FYs 2002- 03, Pakistan’s spending levels compare unfavorably with other developing countries including South Asian counties 11 9. While public development and social spending must recover further from the extremely depressed levels in the last decade and must play a significant part in sustaining high growth and reducing poverty, the biggest economic and social gains must come fi-om improved efficiency in the use o f public resources. Improving the effectiveness o f spending not only through a better allocation o f resources and a more careful choice o f priorities but also through better implementation, more efficient delivery o f services, plugging o f leakages and waste i s a key challenge. This requires, above all, restoring the capability and the authority o f public institutions especially those concerned with economic and financial management and the delivery o f social services. 10. The following will be critical for success in refocusing the public sector and increasing i t s effectiveness : Improved governance and more effective delivery o f law and order, justice and social services through effective devolution not only from the provinces to the districts but also from the Federal government to the provinces. This would need to be accompanied by suitable financial transfer arrangements and strengthening o f the civil service through improved compensation and greater . accountability. Creation o f fiscal space consistent with further debt burden reduction Using the fiscal space to adequately finance the government’s poverty reduction strategy, maintenance o f existing public assets and selective important long-tenn investments especially in water Ensuring FinancialDiscipline and Creating Fiscal Space 11. Creation o f adequate fiscal space i s a necessary but not sufficient condition for better public sector management. But opening up o f the fiscal space, ifproperly utilized, can certainly increase the room for maneuver for improving governance (through more adequate funding o f l a w and order and better c i v i l service salaries) and re-balancing spending (by increasing allocation for operation and maintenance o f infrastructure and more adequate funding o f non-salary expenditures in the social sectors.) Pakistan i s fortunate that it would not be so deeply constrained financially as in the past decade in i t s efforts t o expand the level o f i t s public spending in economic and social fields. With continued reform o n the lines mentioned below, the non-interest, non-defense spending will rise, in the base case scenario, t o 14.3 percent o f GDP by 2006-07 compared to only 10.9 percent in 2001-02 (and an estimated 12.0 percent in 2002-03) consistent with a significant further reduction in public debt burden. In the l o w scenario, it increases only to 11.3 percent o f GDP by 2006-07 and the debt burden reduction i s modest. 12. The major elements that contribute to the increased fiscal space under the base scenario are the reduction in interest payments (2.9 percent o f GDP), higher government revenues (1.3 percent o f GDP), and reduction in enterprise losses (1.4 percent o f GDP. I t i s noteworthy that significant progress towards raising revenues and reducing interest payments as a percentage o f GDP was already made in 2002-03, the f i r s t year o f the scenario. In the l o w scenario, the government revenue declines slightly as a percent o f GDP over 2002-07, interest payments fall more moderately, and public enterprise losses remain high. The interesting point i s that even with higher fiscal deficits (including grants) o f 4.3 percent o f GDP in 2004-7, compared to the baseline scenario deficit o f 3.0 percent o f GDP, the fiscal space under the l o w scenario i s only a fraction o f what could materialize under the continued reform scenario. 13. A scenario with more fiscal space than in the base case i s certainly possible. Achieving this will, however, require three things: (i) even more determined effort to lower the real cost o f domestic borrowing, (ii) total elimination o f public enterprise losses, and (iii) strong improvements in the effectiveness o f government spending and institutional capacity t o optimally utilize the additional fiscal space on poverty and growth related spending. ... 111 Reducing costs o f Government Borrowing 14. With the recent sharp reversal o f the trend towards real depreciation o f the rupee, increased flow o f concessionary assistance, very favorable terms o f Paris Club rescheduling, and the government actions t o retire high cost external debt, the real cost o f external borrowing has come down sharply. Pakistan's case for continued soft blend of external assistance i s strong because o f the high level o f public external debt and the strength of the reform program. The challenge i s to preserve and deepen the reform program to ensure continued flexible assistance from donors at l o w cost. If the momentum o f quick disbursing concessionary assistance can be maintained, and the real exchange rate remains stable, it should be possible to keep the real cost o f external borrowing to 1-2 percent per annum in the near future. a 15. The real cost o f domestic borrowing, however, remains high. The government has taken several strong measures in recent years to address some o f the weaknesses o f N S S instruments. It eliminated the access o f institutional investors to the schemes in early 2000. Recently it took commercial banks out because o f abuses o f the system. I t made, after M a y 1999, a number o f successive reductions in the administratively set interest rates o n new issues since and has during the last year or so linked the N S S rates to the market benchmark Palustan Investment Bonds (PIBs), which have been revived as an important instrument o f debt management. Still real long-term rates o f return o n 10-year certificates for widows and pensioners o f 6 percent (assuming inflation rate o f 4 percent per year) and on Defense Saving Certificates (DSCs) o f 4.5 percent on what are r i s k free assets are high by international standards; the reductions in nominal interest rates have been largely offset by a drop in the rate o f inflation.. Ideally i t would be desirable to phase out the use o f N S S instruments and rather fund government deficits domestically by selling PIBs. However, phasing out the N S S instruments would meet with substantial public opposition and i s not practical partly because the government sees some o f the recently launched schemes under N S S as a safety net to pensioners, widows and small savers. But the Government must continue with progress o n reforming the N S S while educating the public on the true opportunity cost and problems o f the schemes. The successful launch o f the PIBs has not only set a market-based long-term benchmark yield but has also helped the government to finance i t s budgetary requirements with stable long-term debt. In any event, recourse to PIBs should be further enhanced, the on-tap feature o f N S S instruments should be eliminated, and the margins over PIB yields should be phased out. Revenue mobilization 16. Pakistan's revenue mobilization has been low, both by international standards and compared with i t s potential. In FYs 2000 and 2001 total revenue collection was the equivalent o f 16.2 percent o f GDP; tax revenue was only 13 percent of GDP and had actually stagnated for several years.' However, revenue collections during the last two years increased sharply to 17.9 percent o f GDP in 2002-03 reflecting both the success o f tax collection efforts and the revival of the economy. The structure o f taxation has, also improved considerably in recent years, and the heavy dependence on foreign trade taxes has been reduced. Still, revenue mobilization falls short o f that o f many developing countries at comparable income levels in Afnca, Asia, and Latin America.2 In light o f the weaknesses in the current revenue mobilization system, there i s considerable scope for enhancing collection and broadening the income tax base. For instance, while there are only 1-1.5 m i l l i o n income tax filers, there are 3 million cell phone subscribers, over 10 million electricity consumers, 3-4 million gas consumers and 2-3 m i l l i o n car owners. 'Note that Pakistan has undertaken considerable trade liberalization in recent years with an initial import tariff revenue loss. *Countries at comparable income levels with higher revenue/GDP ratios include Bhutan, Botswana, Kenya, Maldives, Namibia, Nicaragua, Sri Lanka, Vietnam, and Zambia. All of these countries had revenue mobilization in excess o f 20 percent o f GDP, with the exception o f S r i Lanka and Zambia. iv 17. In our view, i t should not be difficult, in light o f current efforts, to increase revenue mobilization further to 18.3 percent o f GDP by FY07, the assumption in the base case. But the suggested level o f revenue mobilization will merely regain the level o f the early 1990s. Therefore, to underpin strong public finances and t o adequately fund poverty-oriented expenditures, determined efforts are necessary to complete institutional and policy reform especially the fundamental restructuring o f C B R in the next few years. In the area o f tax policy, reforms should continue to aim at continuing to reduce the number o f tax exemptions (for customs duties, income and withholding tax) and make sales taxation more effective. Reforms are also needed for tax o n income from financial and land assets. In particular the current practice that allows a single person to hold assets under different names (and more easily evade taxation) should be abolished. On the non-tax revenue side it will be necessary to seek greater cost recovery in higher education, health (subject to considerations about access to basic services for the poor), and water usage for irrigation. Sustained increase in revenue mobilization also requires reversal o f the declining trend in provincial tax collection, which i s only 0.5 percent o f GDP. Tax authority powers o f the provincial as well as local governments should be reviewed and incentives for sub-national level resource mobilization need to be strengthened as a part o f larger discretionary resource transfers to the provincial and local governments that appear necessary (see para. 45 below). P u b l i c enterprise losses 18. Though there has been progress during the last two years in eliminating the operating losses o f a number o f public enterprises notably Pakistan Railways, PIA, and most state-owned banks, the losses in the power sector are continuing at a high level. The WAPDA and K E S C losses in 2002-03 were close t o 1 percent of GDP. The continued large losses o f public owned power entities i s the most troubling element of Palustan’s fiscal position and the only one that has not shown improvement in recent years. While the urgency o f need t o privatize KESC and parts o f WAPDA has grown because o f continued losses and inadequate resources for investment as well as the costs to other industries from unreliable power supply, the progress towards privatization i s the slowest in the power sector, though WAPDA has moved to create separate corporate power entities. 19. The power sector reforms are fundamental. WAPDA needs to decisively reduce line losses, enforce bill collection, if needed by cutting o f f non-payers more aggressively, and ensure timely implementation o f tariff adjustments determined by NEPRA. The govemment should help WAPDA in resolving the issue o f nonpayment by FATA o f i t s utility bills since this seems to require a political solution. The cost o f pursuing noncommercial objectives as imposed by government mandate should be fully financed by explicit grants in the fiscal budget. Fortunately, in the years beyond FY 2004, power sector costs will be favorably influenced by the rising share o f hydro- electric generation due to commissioning o f Ghazi Baroda project and some reduction in IPP payments. Defense 20. In the 1990s the real spending on defense increased little and as a percentage o f GDP, it has declined steadily from the peak o f 6.9 percent in the second h a l f o f the 1980s. Despite the spurt in spending in 2001-02 due to regional tensions, defense spending remained at 4.7 percent o f GDP (4.1 percent excluding defense pensions which are shown separately since 2000-0 1). 21. The government feels that in current circumstances assuming further reduction in defense spending as a percentage o f GDP would not be realistic. While easing o f tensions o n the Eastern border i s very likely, the situation on the Western border as well as combating terrorism in the country will continue to put pressure o n the defense expenditures. The report, therefore, assumes that in the base case scenario defense spending (excluding pensions) in FY04, would be as in the federal budget approved by the National Assembly and in coming years would remain at 4 percent o f GDP. V Reorienting public sector priorities 22. Proper use o f the considerable fiscal space that can be created under the continued reform scenario will be critical. The financing o f the poverty reduction strategy should be the first charge on the additional fiscal resources that are likely to become available. But it should also be possible to increase funding for maintenance and undertake selective new public investment in infrastructure. As the past experience suggests, the availability o f the fiscal space alone i s n o guarantee that i t will be utilized for high priority spending. Indeed the danger i s that it could lead to a relaxation o f the financial discipline and undertaking o f l o w priority spending. The only way to guard against this danger i s for the government to exercise self-discipline and to strictly adhere to well laid out processes and procedures for approval o f projects especially large projects. The l o w scenario will largely mean business as usual and will neither enable reversing much o f the decline in the quality o f public goods witnessed during the last decade nor ensure sufficient progress towards Millennium Development Goals (MDGs), though governance improvements could s t i l l ensure better performance than in the past. Funding the poverty strategy 23. The original indicated funding for poverty reduction (4.2 percent o f GDP in 2003-04) in the Interim Poverty Reduction Strategy Paper (I-PRSP) appears to be totally inadequate in relation to ambitious targets especially in education. On the one hand, there i s need to set more realistic but still ambitious social targets consistent with MDGs, o n the other hand, the financing needs o f meeting these targets require more careful assessment. It i s also recommended, that priority poverty spending should be defined more broadly to include l a w and order maintenance, access to justice, and rural electrification. The governance problems resulting from inadequate spending o n police and the justice system impact the poor the most. Similarly, it i s the poor in the rural areas who most lack access to electricity; o f the roughly 40 percent o f Pakistani households that do not have electric connections most live in the rural areas and are poor. 24. Using the broader definition o f poverty related spending suggested here, the actual priority poverty spending in 2001-02 was about 4.5 percent o f GDP. Our very rough estimates suggest that this spending should increase to about 7 percent o f GDP by 2006-07. Larger education and health spending would probably take up h a l f o f this increase in order to achieve goals set; the rest being accounted for largely by higher allocations for l a w and order, provision o f justice and rural and urban infrastructure including rural electrification. I t i s encouraging that the recently completed PRSP (December 2003) provides both for a broader definition of poverty spending suggested above and proposes to increase this spending steadily to 6.8 percent o f GDP by 2007-08. Remedying the neglect of education 25. The lessons learnt from the experience during the last decade in which expansion o f primary education spending resulted in only moderate benefits are already being reflected in approaches t o primary education. Less emphasis i s put o n building new schools and more attention i s being given t o improvements in quality through hiring o f new teachers, greater accountability o f teachers, better physical facilities in schools and higher non-salary spending. Human Development Foundation set up by the Government o f Palustan as an autonomous corporation i s emphasizing enrollment processes, non-formal education especially for recent dropouts, and innovative approaches to capacity building. But it i s n o t at all clear that programs, new initiatives, institutions and funding commitments are in place o n a scale t o achieve the key targets in primary education: 35-40 percent growth in primary enrollments over the next fi v e years, narrowing o f gender and geographical differences, significant improvements in quality, reduction in drop out rates and expansion o f programs o f non-formal education. vi 26. Four sets o f issues relating to the social sector appear especially important. First, devolution o f responsibilities for education and health t o the local government holds the greatest promises o f improved effectiveness o f delivery of social services. But as yet there are a large number o f unresolved issues in this area. Second, though enrollment levels in Pakistan are among the lowest in the developing world and gender disparities and rural-urban gaps are among the widest, there are few incentive programs that have been used around the world to stimulate demand for education and encourage enrollments. Third, the effective public-private sector partnerships that will tap into the strength o f the private sector have yet to be developed. The private sector i s the most dynamic segment of education. I t accounted for over 40 percent o f the growth in enrollments in the primary education in the last decade. The average teacher salary in private school i s actually lower than in public schools but the quality o f teaching appears to be better because o f much greater accountability. Last but not least, more detailed and realistic estimates o f public funding requirements are needed not only for the primary education but also for the entire education sector. The government plans for reversing the decline in higher education deserve high priority and need to be adequately funded. 27. Clearly increasing public funding alone will not help meet the urgent challenges in the education sector. But higher public spending i s a necessary if not sufficient condition for improving educational attainments. Detailed work i s necessary t o firm up the education funding requirements including for the secondary and higher levels. Our tentative estimates suggest the need o f gradually raising public education expenditures from 1.8 percent o f GDP in 2001-02 t o 2.5 percent in 2006-07. 28. Within the recommended overall increase in education expenditures, a more rational allocation o f spending i s needed. This includes increasing resources for: non salary recurring expenditures for primary, middle and secondary school improvements, especially textbooks at the primary level; provision o f missing facilities in existing infrastructure, especially at the elementary and secondary level; provision o f quality services such as teacher training; increasing resources for new infrastructure where there i s a need at i.e., middle and secondary level of education; girls’ incentive programs; and demand side interventions. 29. The fundamental problems that the government has to resolve to get better outcomes in education are institutional. In the absence of fundamental reform, additional funding by itself will achieve little. M a n y o f these problems, accountability, responsibility, service orientation, and incentives that reward people for doing their j o b well, are generic t o the entire public sector, but they manifest themselves most visibly and with greatest impact in the education sector because o f the size and multitude o f actors that constitute the system. Accelerating Water Development 30. Because o f great neglect of water investments in the recent past, a substantial acceleration in overall public spending in the water sector appears justified. WAPDA and Planning Commission must be commended for their initiatives in reviving water sector investments. There are, however, several problems with the thrust of the present program that require careful consideration and review by the Government o f Palustan. M a n y of the issues involved, the tackling o f which can yield a more optimal and balanced expansion o f irrigation water related public spending, are discussed at length in Volume I1 . o f the Report. These can be summarized as under . Need for definition o f a clearer vision which integrates long-term agriculture, irrigation water and hydro-electric development . Excessive emphasis o n extension o f irrigation facilities some o f w h o m may prove marginal and all o f which will yield economic benefits only after a number o f years Inadequate attention to the urgent needs o f rehabilitation o f existing irrigation assets and on- farm development vii . . Inadequate attention t o the political and technical issues which may stand in the way o f an early resolution o f a much needed additional storage on Indus . The need for harmonization o f ambitious water sector plans with overall resource availability and to avoid pre-emption o f urgent social investments Balancing institutional development needs with higher water sector investment including Federal- Provincial co-ordination, and greater investment in planning and investigations. 3 1. A balanced program (including priority activities) covered under the poverty reduction strategy can help initiate the transformation o f the water and agnculture sector even in the medium term. The near trebling o f the real spending in water over 2002-07 will take up 20-25 percent o f the additional fiscal space that i s expected to be created under the base scenario and will sharply reverse the declining trend o f water sector spending. Needless to say, such large spending carries risks and could waste resources if it i s not accompanied by much tighter control on program and project priorities, stronger implementation and institutional arrangements than exist at present. A fundamental strengthening o f the sector planning, research and development, project implementation environment, and monitoring mechanisms including much greater inter-provincial and Federal-Provincial coordination i s also necessary to facilitate decision making o n the more difficult longer-term issues. Transformingthe Power Sector 32. While the Government has taken a number o f steps to promote private investment in the power sector, the desired goal to ensure the development o f the sector without recourse to the GOP budget, except for hydro-electric development and possibly transmission, i s n o where in sight. The continuing financial distress in KESC, notwithstanding substantial injection o f public monies to fund past losses, and intemational conditions make it highly unlikely that K E S C can be privatized in the near future. I t also appears somewhat unlikely that the newly corporatized distribution and generation companies under WAPDA will actually be privatized on a timetable to fund power sector investments during 2003-07. 33. Consequently, there i s a disconnect between large national power sector requirements and the amounts that can realistically be expected from the development budget and own resources o f WAPDA. Whereas the TYPP has notionally allocated Rs.90 billion for WAPDA mainly from external loans, WAPDA’s proposed program including Rural Electrification for 2003-07 i s Rs.26 1 billion. So there will be a large additional requirement, after providing for WAPDA’s own resources, which will have to be mobilized from the budget andor borrowing in the market with government guarantee. KESC investment needs will be over and above this. The extra requirement for the power sector investment could wipe out one-third of the fiscal space that i s supposed to be available under the baseline scenario. 34. While the system o f automatic adjustments in electricity tariffs to reflect costs must be further strengthened and strictly enforced, there may be only limited scope for increasing real, as distinct f r o m nominal, electricity tariffs3. A detailed sector plan for the power sector for the period 2004-08 needs t o be developed on an urgent basis and harmonized with the development o f a rolling three-year plan and the MTBF. I n the absence o f a plan, the power sector could derail government’s economic and social priorities. Some o f the elements o f the power sector review should include the following: . A realistic but pro-active time table for privatization o f corporate entities in the sector . Careful review o f the present plans o f WAPDA especially for generation . Reinforced efforts to reduce technical losses including theft Identification o f key actions t o achieve cost cutting Clear identification o f power sector financial requirements under a reform scenario According to recent World Bank estimates, 10 -15 percent increase in nominal tariffs over the next few years w i l l be inescapable even with aggressive cost reduction and speedy progress towards reduction o f theft. ... Vlll Effectiveness o f Public Spending 35. The effectiveness o f public expenditures has several dimensions; strategic approaches to public interventions; soundness, transparency and accountability o f budgetary and planning processes; careful screening o f projects, speedy and effective implementation o f programs and service delivery; and the existence o f strong monitoring and evaluation systems. The presence o f functioning institutions and a framework o f defined distribution o f responsibilities at various levels o f government are the cross-cutting issues which influence all aspects o f effectiveness o f public spending. While the mechanics o f better and more efficient public service delivery are important, a real test o f effective service delivery i s whether it meets the expectations o f the citizens especially the poor. Pakistan’s record in effectiveness o f public spending has deteriorated for several reasons. The public sector became overextended, poor govemance and declining strength o f public institutions led to large leakages from expenditures, strict discipline in choice o f new large projects was not observed. Many o f these weaknesses are being set right. Privatization i s being pursued strongly and public sector role in commercial activities i s being de- emphasized. Devolution i s being stressed as away to move services closer to the people and to increase accountability o f public officials. Fiduciary controls have been strengthened. But in many cases, notably privatization and devolution, reform implementation i s far from complete. In other areas, such as improving budgetary and planning processes and strengthening o f civil service, a full reform agenda has yet to be developed. Sustained improvements in the management o f public expenditures would require not only successful implementation o f the measures in hand but also a clear vision o f the future changes that are required. Some o f the key elements that would shape the effectiveness o f public spending are discussed below. Accelerating Privatization 36. Successive recent governments since the early 1990s have already opted for a much more focused role o f the state and much greater role for the private sector. A great deal o f progress in furthering privatization has been made in the last three years. The need i s to speedily complete the privatization plans that are underway and to extend the program in power, transportation, and insurance. Public-PrivatePartnerships 37. n Pakistan the possibility of inducting the private and non-govemment sector in service delivery I through public-private partnership has not been fully explored. 38. This has become especially important because o f the emerging possibility o f larger fiscal space but continuing capacity problems within the government. Education and health appear to be the most promising sectors for collaboration partly because national priorities justify a rapid expenditure o f public spending on them. Tertiary water management i s another area where efficiency o f water use can be increased and local resources mobilized if farmer associations can be inducted in the process. Devolution 39. The Government o f Palustan has during the last two years taken a major initiative to devolve powers to local levels o f the government especially the district level. The basic idea i s to move the responsibilities for service delivery closer to the people and to increase the accountability o f the officials. While the devolution program provides the best opportunity in years for improving the structure of governance and improved effectiveness o f public services, the system needs considerable fine-tuning and improvement to achieve these objectives. Some o f the areas that require early attention o f the government are: ix . . Extending administrative and fiscal autonomy: . . Extending facility-level oversight . Reducing institutional complexity Improving capacity Improving funding 40. The funding issues are the most urgent and immediate. A significant portion o f the fiscal space that i s emerging must be made available t o the district governments in part through Provincial Finance Commission Awards and partly through incentive mechanisms which would help the local governments to prepare and implement their policies and plans in accordance with the national and provincial priorities, without compromising the autonomy o f the local governments. 41. As the system beds down, it will be important to harness the potential that it offers for strengthening user control over local policy and over the supply o f local services. This will require that government take a distinctly pragmatic approach to the development o f Citizen Community Boards, amending their structure and their govemance arrangements as practical experience i s gained. More strategically, devolution may require some adjustments in the new political environment. The provinces might wish to pursue somewhat diverse approaches to the precise pattern and speed o f devolution. This should not be discouraged by the Federal government as long as the essential spirit o f the devolutionary processes i s preserved and there i s clarity about the functions o f government at various levels. Federal- Provincial Responsibilities 42. There i s significant scope for devolving additional functions from the federal to the provincial governments. The simplest way to proceed on this front i s to have a serious review o f the actual federal- provincial responsibilities in relation to the letter and spirit o f devolution especially o f those service delivery functions that have been assumed by the Federal Government during the last three decades either because o f financial constraints or lack o f technical and administrative capacity (actual or perceived) at the provincial level. Devolution o f some o f the functions presently carried out by the Federal Government would reduce the provincial government concern o f becoming redundant in the wake o f large devolution to the district governments, will lift the morale o f the provincial c i v i l service and would enable the federal govemment to focus more intently on i t s key strategic functions. In most o f the concurrent l i s t areas, the federal government should remain concemed with providing strategic direction, setting national minimum standards and providing financing to the provinces conditional on their compliance with these standards. 43. Over-centralization o f functions i s also reflected in the planning process, which s t i l l exhibits a strong top-down format. To some extent it i s unavoidable because o f the central role o f foreign assistance. But efforts should be made to eliminate those schemes that are regional in nature from the Federal PSDP. Inter-Government Financial Flows 44. The role o f provinces in developing realistic budgetary frameworks and viable poverty reduction strategies can hardly be over-emphasized. W e have argued for much greater devolution o f responsibility to the provinces in order to underpin and strengthen the devolution t o the district level. But even with the present rather ad hoc distribution o f responsibilities, i t i s clear that it i s the provincial/ district level spending on education, health, l a w and order, provision o f justice, irrigation, rural and urban water supply, and roads that will account for the bulk o f the increase in real public spending in the near future. 45. The share o f provincial/ district level public spending i s likely to rise from around 26 percent o f the total public spending in FY03 to over 33 percent in FY07. I n terms o f GDP, provincial/ district level X spending should increase from 5.5 percent to 7.5 percent over the period, if our assumptions about the fiscal space and priorities materialize. T h i s has major implications for allocation o f national financial resources. At present, the sub-national govemments are facing a severe fiscal squeeze. Whereas the National Finance Commission (NFC) has been seriously considering a significant increase in the share o f provinces in the federal divisible p o l l from the present 37.5%, additional flow o f funds would be required to achieve the targets o f the PRSP. Clearly, mechanisms will need to be developed t o allow a large portion o f the fiscal space created at the federal level to be passed down to the sub-national levels. This will require two things: (1) new NFC and PFC arrangements which assign much greater formula-based transfers to the lower levels o f govemments than at present and (2) putting in place a system o f financial incentives and special transfers to the provincial and local govemments so that national priorities can be protected consistent with decentralization and devolution. Such transfers can be linked to provincial revenue efforts and demonstration o f capacity to effectively utilize resources. Improving the Effectiveness of Planning and Budgetary Procedures 46. The introduction o f W R S P and MTBF provide important opportunities for moving away from the excessive focus o n the development budgets and to a more comprehensive approach to public expenditure management. But several planning and budget process issues will need to be revisited and resolved before the MTBF can become a truly operational document. The relative responsibilities o f a range o f public sector entities in relation to various elements o f a MTBF estimation o f resource envelope, costing o f sector programs, choice o f inter sector priorities, adoption o f specific targets etc. need to be looked at afresh and agreed upon because they are in a state o f flux. The arrangements for putting together a MTBF will also need to be harmonized with the preparation and implementation o f the final PRSP. These entities include the federal and provincial govemments, the sector ministries/ departments, Planning Commissiodprovincial planning departments, and the Ministry o f Finance/ provincial finance departments 47. Given that Pakistan i s a federation, it i s critical for improved budgetary outcomes and enhanced effectiveness o f public expenditure that the MTBF process i s “rolled out” to all the provinces (and ultimately to all the districts). 48. The introduction o f the new processes, i.e. MTBF and PRSP, also provide an important opportunity for strengthening overall planning in Pakistan and improve policy co-ordination among the key economic ministries/departments both at the center and in the provinces and between the federal government and the provinces. This will not, however, happen automatically. The Planning Commission and the planning departments should lead the effort to integrate recurrent and investment expenditures in a coherent medium-term budget fkamework. The preoccupation o f planners with the development budget i s hampering the pursuit o f balanced strategies and speedy social development. Strengthening Civil Service 49. In Pakistan the strength o f public institutions has been greatly eroded over time. M a n y inter- related factors have contributed to this decline: increasing centralization and non-transparent decision making, lack o f accountability, the dilution o f the role o f merit in selection and promotion of c i v i l servants, undue political interventions in normal functioning o f government departments and agencies, declining technical and professional competence o f officials, l o w or declining compensation, and a culture of risk aversion behavior. The broader issues o f accountability and over-centralization are n o w being tackled as a part o f the reform process. Some steps have also been taken towards c i v i l service reform. xi 50. But clearly fundamental problems with the civil service, l o w compensation for the higher levels, l o w level o f professional competence due to l o w investment in training, s t i l l excessive weight to seniority, and perverse incentives and risk averse behavior, have not disappeared. The government recognizes this and i s examining a number of proposals, to further strengthen merit, to streamline cadres and t o increase both opportunities and incentives for greater professional development. These issues have acquired urgency with the implementation o f devolution, which requires re-definition o f the roles and responsibilities o f district, provincial and federal levels o f government. 5 1. This report sets out the need for administrative reform, but argues against an over-ambitious or over-elaborate set o f proposals. History suggests that such proposals will generate many opportunities for lengthy reports, but will lead to little sustained progress. Instead, it argues for three steps. First, it suggests that there i s an opportunity t o move ahead rapidly on the creation o f a National Executive Service that would greatly increase incentives both for improved performance and upgrading o f s h l l s through learning and education. The two critical elements o f this reform, on which there i s seemingly a substantial consensus, are the much greater emphasis on merit and performance for selection to top grades, including a high bar for promotion beyond grade 18, and opportunities for higher education and learning for civil servants especially early in their careers. Second, it suggests sharp pay increases (and monetization o f many o f the large in lund benefits) for higher levels o f civil servants so that the serious decompression o f pay that has been going o n for decades can be reversed. Finally, i t notes that completing administrative devolution, by transferring budgetary responsibility for salaries t o the district governments and by progressively passing appointment, promotion and transfer (APT) authorities to the districts, will address a necessary, if not sufficient, condition for local governments to value professionalism more than patronage. 52. The opening of the fiscal space can provide opportunity for reversing the squeeze in real salaries and de-compressing the compensation differentials and further consolidating monetary allowances and in hnd benefits into base pay. However, before a new compensation policy can be formulated the future pension liability must be assessed and actions taken to avoid a crisis in the future. The initiation o f work in this regard in the Ministry o f Finance i s very welcome. 53. Provided future pension liability can be contained and civilian employment growth can be kept down to say 1 percent per annum, there should be room for increasing real salaries significantly over the next five years. The challenge will be to use the fiscal space, on the one hand, for restoring compensation closer to private sector for the higher grades, and o n the other to adjust pay scales for groups for instance primary school teachers, policemen, lady health workers who are particularly poorly paid. Flat rate increases, that do not decompress salaries or provide motivation to important occupational groups, will not serve the objectives o f the civil service reform. 54. Similarly, expansion o f civilian government employment will be necessary but should be extremely selective and driven by identified needs in light o f the goals set in the poverty reduction strategy papers at various levels o f government. The de-compression in compensation that i s being recommended may be politically difficult but i s urgently needed if high-level managerial and professional s l u l l s are to be attracted to the public sector. A move towards more adequate compensation should be accompanied by and make possible other reforms such as much greater selectivity in promotion beyond a certain level, strengthening of incentives for learning and s l u l l upgrading, and selective induction o f higher level professionals from outside. Fiduciary controls 55. The government has recently embarked o n a comprehensive reform program to restructure i t s financial management system. Although significant steps have been taken in this regard, the government xii must carry the program forward and involve the subnational governments, public corporations, and autonomous bodies more intently in the reform process. The continuation, expansion, and deepening o f the reform program can ensure substantial improvement in the government’s financial management system within the next three to five years. The CFAA-Phase-I1 has outlined (among other things) the following key actions for improving the financial management system. 56. These actions include (1) enhancing the transparency o f the budget by disclosing fixed-cost and contingent liabilities and the transparency o f the accountability system by publishing annual and quarterly financial reports o f all levels o f government; (2) developing provincial budget implementation systems along the lines o f the federal government’s Revised System o f Financial Management and Control and malung the federal and provincial systems fully operational; (3) establishing a strong internal control mechanism for each level o f government; (4) building the capacity o f the auditing and accounting establishments and o f the PAC; and (5) renewing the government’s commitment to implementation o f PIFRA. Conclusion 57. With political stability, reduction in regional tensions, improved governance and a continued forceful pursuit o f the economic and social reform agenda, Palustan has good prospects o f raising the annual GDP growth to around 6 percent in the next three to four years and possibly higher in later years and sustaining it at a high level, consistent with progress in debt reduction and a more self reliant and equitable pattern o f development than in the past. With the decline in the population growth rate expected to continue, sustained per capita GDP growth rates in the later part o f this decade o f about 4 percent per annum certainly appear feasible. T h i s would represent a dramatic contrast to the average annual per capita increases o f little over 1 percent in the latter part o f the 1990s and would contribute strongly t o poverty alleviation and employment creation. 58. A center- piece o f the growth revival must be the recovery in the fixed investment rate from 13- 14 percent o f GDP in recent years t o around 17 percent by 2007 l e d mainly by the private sector. The implied high growth in investment over the next few years would require further efforts to improve investment climate both for the domestic and the foreign investors. Better public resource management can assist the growth in private investment in several ways, by improving governance especially l a w and order and access t o justice, by promoting slulls development, providing needed infrastructure, and last but not least by strengthening bureaucracy and malung public institutions more responsive to citizens. 59. With continued reform, the restructuring o f public expenditures away from interest payments and public enterprise losses will create substantial fiscal space for increasing social and development spending. T h i s will provide a unique opportunity t o fund Pakistan’s poverty reduction strategy and at the same time narrow the gaps in key public goods. Social spending targeted on the poor deserves special priority because poor have seriously lagged behind in access t o basic services. I t should also be possible to selectively increase infrastructure spending and also fund governance and civil service improvements more adequately. However, without addressing the institutional issues that are at the heart o f the effectiveness o f public spending, the opportunity afforded by likely creation o f fiscal space would be wasted. Some o f the elements o f public expenditure management that can make possible more effective public spending are: continued focus o n accelerating privatization plans, making devolution work at the district level, devolving more responsibilities and finances to the provinces, adoption o f medium-term budgetary frameworks at all levels o f government, moving away from excessive focus on development spending, developing public-private partnerships in delivery o f social services, and last but not least strengthening the civil service by improving incentives, expanding opportunities for learning and more transparent accountabilities. ... XI11 60. Even though there i s a clear need for sharp expansion o f both private investment and non- interest and non-defense public spending, i t i s the qualitative aspects o f this expansion that will, in the ultimate analysis, assure the productivity improvements vital for growth and poverty reduction. Even with continued reform momentum relative levels o f total investment and non-interest and non-defense spending do not fully recover to the level of the early 1990s. Higher growth and more positive social outcomes must come mainly from the expected improvements in the quality and effectiveness o f public spending and a private sector that rises t o the challenge o f diversifying export and industrial structures, xiv PUBLIC EXPENDITURE MANAGEMENT: STRATEGIC AGENDA Challenges & Objectives Key Recommendations A. Reduction in public debt t o sustainable levels, 1 Contain the overall fiscal deficit (including external i.e. limiting public debt t o n o more than 400 grants) to 3 percent o f GDP o n average during . percent o f government revenue by 2006107. FY04-07. A I . Increase government revenue (to over 18 Accelerate CBR reforms, including a complete percent o f G D P by 2006107, and t o an average reorganization over a three-year period (FYs 04-06) of 17.7 percent o f G D P per annum during and fkndamental changes in human resource 2003107). management. A2. Reduce financial losses o f State Owned 1 Speedily complete the privatization plans that are . Enterprises. underway and to develop a realistic but proactive timetable for privatization o f WAPDA and KESC. Reinforce efforts to reduce technical losses in the . power sector. . A3. Reduce cost o f public borrowing. Reduce the margin o f NSS instruments over the PIB yield. End the on-tap feature o f NSS instruments. II.Reorienting Public Sector Priorities B. Improve social indicators and narrow the gender 1 Fund t h e poverty reduction strategy adequately by . gap by improving the effectiveness o f social raising priority poverty spending by 2% o f GDP to spending, and allocating more resources t o the about 7% by 2007. sector. Substantially increase allocation for education and health focusing o n b o t h quantitative and qualitative improvements. Devise incentive programs t o stimulate demand for education and encourage enrollments, especially among the poor. Develop public-private partnership in social sector, . 1 especially in remote and underserved areas. C. Enhance agricultural productivity through Develop a clear vision, w h i c h integrates long-term accelerated development o f water sector. development o f agriculture and power sectors with investments in irrigation sectors. 1 Develop a water sector investment strategy which strikes appropriate balance between expansion of irrigation facilities and: (i)development o f additional water storage facilities; (ii) O&M needs o f the system; (iii)investments in the drainage sub- sectors; and (iv) investment in water saving schemes, e.g. lining o f water courses and leveling land under OFWM projects o f the provinces. 1 Implement the institutional plans for raising the effectiveness o f spending o n irrigation, drainage and water storage. xv Challenges & Objectives Key Recommendations I III.Improving Effectiveness of Public Expenditure 1 D. Improve outcomes o f public spending at lower unit 1 Develop and adopt a rolling Medium-Term Budget Framework (MTBF) at the federal level and in provinces, based o n sector strategies and hard budget constraints and integration o f development and recurrent expenditures. 1 Reinvigorating the planning processes by m o v i n g away f r o m excessive focus o n development spending and integrating planning and monitoring with the preparation o f MTBFs. govemance through finther devolution. 1 Complete devolution to district level by extending adrmnistrative and fiscal autonomy, improving capacity, and expanding funding. 1 Undertake devolution f r o m the federal t o provincial governments with appropriate funding arrangements. 1 Establish mechanism, including financial incentives (e.g. matching grants), through which national and provincial priorities could be internalized by the local g o v e m e n t s . I I K Strengthenin, the Civil Service 1 M o v e ahead rapidly o n the creation o f a National Executive Service that would greatly increase incentives b o t h for improved performance and upgrading o f s k i l l s through learning and education. 1 Use the fiscal space, o n the one hand, for restoring compensation closer t o private sector for the higher grades, and o n the other to adjust pay scales for groups, for instance primary school teachers, lady health workers w h o are particularly poorly paid. 1 A v o i d flat rate increases that do not decompress salaries or provide motivation for important occupational groups should b e avoided. 1 Give the FPSC (and PPSCs) an unambiguous responsibility for oversight o f the recruitment process at a l l grades (even if i t i s decided that i t should not itself undertake direct recruitment for positions below BPS16). xvi INTRODUCTION 1. Pakistan has been grappling with the serious problems o f slow growth, heavy debt, social gap and widespread poverty for some time. These problems have their roots in economic and social policy as well as institutional weaknesses. There was extraordinary and persistent reliance on external resources for financing economic development, almost relentless growth in public spending until at least the mid-1980s and the consequent persistent high levels of fiscal deficits leading to unsustainable levels o f public debt. Though the reliance on public sector was very heavy, the strength o f public sector institutions was eroded over time, reflecting both the growing weaknesses in the civil service and undue political interference in such matters as choice of projects and pricing decisions o f public corporations. Even though public spending was large, social sector investments were greatly neglected. Palustan’s social gap i s well known. N o t only did social indicators not improve in line with the growth in GDP but also the gap between Pakistan and comparable low-income countries has tended to widen (see Figure 1). 2. As the debt burden rose and the financial constraints deepened, the investment rate showed a steady decline reflecting not only political instability but also a lack o f confidence in the currency. The impact o f reduced investment o n the economy was aggravated by poor governance and continued slides in the institutional capacity in the government, which adversely affected the quality o f public and private investment as well as the effectiveness o f public spending. GDP growth averaged only 3.3 percent per annum in the second h a l f o f the 1990s. Table 1.1: Growth Trends Annual Average for Indicator (%) 1980s 1990191-1994/95 1995/96-1999/00 2000/01-2002/03 Compound growth rate of real GDP 6.5 4.9 3.3 3.7 Inflation (period average) 7.2 11.5 7.9 3.4 3. The Musharraf Government, which assumed power in October 1999, has been giving high priority to revival of economic growth, reduction o f poverty, and speedy human development. I t has pursued a comprehensive and forceful reform agenda aimed at restoring macroeconomic stability, improving governance, accelerating investments in education and health, improving the climate for private investment, and last but not least malung econqmic and social policies pro-poor. The Govemment o f Palustan’s (GOP) determination to restore financial stability i s reflected in its Debt Reduction and Management Strategy adopted in March 2001.4 The fairly ambitious targets o f public and external debt reduction are intended to be made a part of a Fiscal Responsibility Ordinance (approved by the Federal Cabinet) and are to be monitored by a high level Debt Office. The GOP’s development agenda and reform program were detailed in the I n t e r i d Poverty Reduction Strategy Paper (VPRSP) issued in November 2001. In December 2003, GOP issued the final PRSP, which elaborates and substantially strengthens the poverty reduction strategy. A great deal o f success was already achieved in F Y s 2001 and 2002 in attaining macroeconomic stability, reviving growth, building up foreign exchange reserves, reversing the decline in public spending and increasing pro-poor expenditures. “Debt Reduction and Management Strategy”, Govemment of Pakistan, Ministry of Finance, March 2001 (see Annex). T h i s strategy was approved by the cabinet following the recommendations of the Debt Reduction and Management Committee Report (hereto referred as the Debt Committee report). T h e Committee was set up in January 2000 and submitted i t s report in early 2001. Figure 1: Trends in Social Indicators ‘overtyHeadcountfell in Pakistan during the 1980s but stopped alling in the 1990s, leaving 45 million people below the poverty line. @Pakistan 3Pakistan I El Bangladesh 140 Bangladesh 60 .India India ONepal 120 -- $ 50 h 100 -- 3 ft 40 SO -. 8 E 3 e 30 60 -- G g 2o -. 3 40 4 -. 20 10 s 0 7 0 F’ iin?avSchool Proportion o f girls in Infant Mortality Rate M a t m a l Momlit E n r o l k t rate all Primary student? @er 1,WO) @R 10,000) 1985-1986 1995-1996 1999-2000 loaeent) hremt) Pakistan’soverall spending on education has fallen, while that of Fertility Ratio has begun to fall in Pakistan- but it has fallen comparator countries has risen. much faster, earlier in comparator countries and is among the highest in Asia. Bangladesh 8.0 ElPakistan BBangladesh 7.0 6.0 E 5.0 d 4.0 e 8 3.0 1 2.0 1.o 0.0 1960 1970 1980 1990 2000 1997 1998 1999 2000 2001 4. Building on the analysis in the Debt Committee Report, the W R S P and the World Bank’s recent economic work referred to above, the present report, which i s prepared in close collaboration with the govemment o f Pahstan,’ focuses on the central issues in public expenditure management that will be critical not only for accelerating growth but also for malung growth more pro-poor. There can be little doubt that export led growth by the private sector i s crucial for Pakistan. But private sector efforts must be strongly supported by improved public sector management. The broader issues o f macro-economic imbalances, political instability and poor governance have adversely affected the climate for private sector investment especially in the manufacturing sector. However, it i s equally true that during the last decade or so deeply constrained public sector spending for social and economic development, continuing The Review Team intends and hopes that the same spirit o f cooperation and collaboration between the World Bank and the Govemment will continue during the disseminationof, and follow-up work to, this report. xviii lags in social and human development, poor allocation o f resources, and growing inefficiency in the use o f capital and delivery o f public services have all compounded the trend towards reduction in growth and a stagnation in poverty levels. A sharp turn around in Pakistan’s economic and social performance requires, above all, a strong recovery in private sector investment. However, it all also requires a significant increase in public economic and social development spending, a better management o f scarce public resources, and a resolution o f the problem o f large loss malung public enterprises. 5. This i s the first in a series o f reports dealing with public expenditure management issues planned for 2003-05. I t i s highly selective in coverage.6 N o attempt i s made to cover all major sectors or all fiduciary issues in expenditure control and management. In a number o f public expenditure areas, the base o f knowledge i s already good and critical policy actions have already been taken. The approach here i s (1) t o refine and synthesize the agreed-upon parameters o f public expenditure management as they have emerged from the government’s development policy agenda and considerable World Bank economic and sector work, and (2) t o focus on public expenditure issues o f strategy, planning, and implementation that remain either understudied or unresolved. The World Bank has been involved--albeit somewhat unsuccessfully in terms of i t s impact--in analyzing public expenditure issues in Pakistan. Some recent Public ExpenditureReports include the following: “Pakistan Reforming Punjab’s Public Finances and Institutions,” August 2001, World Bank; and “Pakistan Public Expenditure Review: Reform Issues and Options,” October 1998, World Bank. xix CHAPTER THE DIMENSIONS 1: EXPLORING OF THE CRISIS IN PUBLIC EXPENDITURES Introduction 1.1 This chapter outlines the various dimensions o f the crisis in public expenditures in Palustan: the l o w and stagnant revenue base, the high burden o f public debt, the great squeeze o n public resources available for economic and social spending in the 1990s, the relative neglect o f social sectors, and the ineffective use o f public spending resulting from both declining capacity o f public institutions and governance problems. It traces the historical origins o f present-day problems and gives an overview o f the key gaps in provision o f public goods. Finally, i t presents a strategic agenda o f public sector management, which i s the focus o f this report. 1.2 I t i s well h o w n that Pakistan under-performs other counties with similar per capita incomes in delivery o f basic social goods. What i s perhaps less commonly realized i s that the gaps in public good delivery have been widening during the last decade especially in relation o f delivery o f services to the poor. Furthermore, the inadequacy o f public investment in key areas i s taking a heavy toll in efficiency losses for the economy and hindering private investment. While a higher level o f social and economic development spending i s n o guarantee for growth revival and poverty reduction, i t i s highly unlikely that these will occur without more spending. The depth o f the crisis in public spending in Palustan can also be gauged from the fact that consolidated public expenditures (excluding interest and defense) averaged l i t t l e over 10 percent o f GDP in late 1990s compared to the average o f about 15 percent o f GDP in the 1980s. ’Even though non- interest non-defense spending rose sharply in 2001-02 and 2002-03, Palustan’s spending levels s t i l l compare unfavorably with other developing countries including South Asian countries (see Table 1.1). Table 1.1: Non-Interest and Non-Defense Expenditure for Selected Countries (Percent of GDP) 1990 Argentina 9.0 13.2 12.7 Bangladesh 15.1 12.6 11.8 Chile 16.2 17.4 21.1 Egypt 20.4 24.3 19.4 India 21.0 17.8 19.8 Indonesia 14.7 12.1 13.0 Malaysia 20.5 15.9 17.5 Nepal n.a. 14.3 13.8 Pakistan 14.5 11.7 10.3 Philippines 11.2 12.9 14.3 Sri Lanka 20.9 19.3 15.9 Thailand 9.8 13.2 15.5 Turkey 12.2 17.3 22.2 Source: Staff estimates, Govemment Financial Statistics, Economic Survey and World Development Indicators, Selected Issues. Data for India has been obtained for Govemment Authorities and interest expenditures have been calculated for both the central and the state governments. Note: Data in the 2000 column for Egypt and Malaysia are for the year 1997, while those for Indonesia and Turkey are for the year 1999. ’ T h i s decline i s somewhat exaggerated because in the early 1990s, the definition o f development spending was revised to exclude spending financed from own resources o f public corporations. The latter constitutes about 1-2 percent o f GDP. 1.3 Beside inadequate resources devoted towards social and economic development, weak governance led t o l o w efficiency o f public expenditure, which contributed significantly t o the poor development outcomes. Over time, the role o f the government has become increasingly out o f touch with reality. Rigid internal structures made i t impossible for the government t o change i t s role and organization to meet challenges o f running a modem day government. Rather than focusing on increasingly complex tasks o f efficient and effective management and regulation o f the economy, and providing good governance, the government continued to devote a large part o f i t s efforts and resources to direct provision o f private goods and services, an area where it had neither the advantage nor the need to be involved. 1.4 The transition from the government’s traditional functions to a modern government was badly hampered by a continuous decline in the quality and authority o f the civil service. Falling real income, coupled with the waning authority o f the civil servants in running the affairs o f the government, made it difficult for the government to attract the best and the brightest into the civil service. The system o f recruitment, placement and promotion, which got increasingly distanced from merit, caused further damage to efficiency o f civil service. The problem was compounded by a system, which was unable to provide incentives for improving the individual skills, productivity or performance. A sharp decline in resources for training and office and technological support resulted in poor s k l l development and loss o f productivity o f the civil service (see Chapter 5). Historical Perspective 1.5 The present inadequate levels o f public spending o n economic and social services as well as other key public services such as l a w and order and the lag in providing these services are rooted in historical developments in Pakistan. Pakistan’s public sector became seriously overextended in the 1970s and 1980s. Several aspects o f this overextension should be distinguished. First, at the most basic level the growth in expenditures far outstripped the growth in revenues, thus leading t o large fiscal deficit and unsustainable levels of public debt. Second, the growth was especially rapid in nondevelopment expenditures, including defense, and as a result the general government savings turned negative after the 1980s, a pattern that has only slowly begun to change. Third, the state intervention in commercial and industrial activities served as a strong disincentive for the private sector until at least the early 1990s. M a n y activities in which the private sector could provide an efficient service without draining public resources are in the hands o f government departments or public corporations. Fourth, in the process the government’s ability, both financial and institutional, to deliver basic public services such as basic health and education was greatly undermined. 1.6 The level, financing, pattern, and effectiveness o f public sector expenditures greatly influence the level o f economic activity, employment, and the distribution o f income in any country. The public sector’s role i s especially crucial in the early stages o f development in developing countries in providing not only strategic direction. and macroeconomic stability, but also key social services and infrastructure. In Pakistan the level, pattern, and mode o f financing o f public expenditures initially influenced development outcomes--that is, growth trends, progress toward poverty reduction, and social progress--to an extraordinary extent at least until the 1980s. 2 1.7 There i s a great deal o f nostalgia in Table 1.2: Investment as a Percentage of GDP (%) Palustan about the relatively high average Public Private Total Gross Fixed Investment Investment Investment growth rate o f nearly 6 percent per year during 1949-50 1.6 2.5 4.1 the thirty-year period from 1960 to 1990. 1954-55 3.9 3.2 7.1 There i s not, however, sufficient memory that 1959-60 7.3 4.2 11.5 the high and sustained growth during this 1964-65 11.6 9.2 20.8 period was made possible largely by the 1969-70 7.3 7.0 14.3 dramatic increases in public investment that 1970's 10.3 5.6 15.9 started in the 1950s and continued well into 1980's 9.2 7.8 17.0 the 1980s (see Table 1.2). 1990's 7.5 9.1 16.6 1999100 6.0 8.4 14.4 1.8 Public investment (including that o f 200010 1 5.5 8.4 13.9 public corporations) as a share o f GDP rose 2001102 4.8 8.4 13.1 steadily from the l o w initial level o f 1.6 2002103 4.5 8.6 13.1 percent in 1949-50 to the peak o f 11.6 percent fource: Economic Survey, State Bank of Pakistan and Hasan (1998). in 1964-65. Following the 1965 war with India, both private and public investment fell drastically. Public investment recovered from the setback in the second h a l f o f 1960s and remained on average at the high level o f nearly 10 percent o f GDP during the 1970s and 1980s. These levels were, however, unsustainable. Public investment showed little real growth in the 1990s and had by 2002-03 come down to 4.5 percent o f GDP. 1.9 Although public investment was quite large until the 1980s, not all public investment was well directed. In the 1960s and the first half o f the 1970s the government made huge investments in the water sector (including storage) equal to 4-5 percent o f GDP. N o doubt, these investments, related in part to the Indus Basin Treaty, helped to sustain agricultural output growth and electricity production well into the . ~ even in the water sector, large programs o f public tube wells proved costly and have remained 1 9 8 0 ~But a drain o n the operation and maintenance budgets o f the provincial irrigation departments. The investments in water storage and irrigation system expansion have not been matched by adequate maintenance o f existing assets or by better water management systems to improve the efficiency o f irrigation water use. Moreover, l i t t l e progress has been made toward solving the major drainage problems in the Indus Basin and building up an institutional infrastructure to guide and implement the plans for use of scarce water resources. 1.10 The major power utilities in the public sector, the Water and Power Development Authority (WAPDA) and Karachi Electricity Supply Company (KESC), have become a major drag on public finances. Indeed, as discussed later, the financial losses o f public enterprises in the power sector are at present the principal source o f the government's contingency payments. The historical roots o f WAPDA's financial problems l i e not only in the inefficiencies o f i t s operations and theft o f electricity, but also in the government's reluctance in the 1970s and early 1980s to allow timely adjustment o f electricity tariffs. Following the sharp rise in the intemational price o f o i l in 1973, i t took the government nearly a decade t o complete the necessary energy price adjustments. Consequently, the demand for electricity remained high while WAPDA's ability to generate funds for investment was badly damaged. This lack o f financial capacity in the face of huge load shedding after the mid-1980s led to Palustan's heavy reliance o n independent power producers (IPPs), leading to a new set o f problem^.^ * According to Hasan (1998: 154), total water and power investments (including the Indus Basin Replacement Works) during the 1960s exceeded US$2.5 billion and accounted for more than 50 percent of total public sector development spending. T h e costs of power from IPPs are high, especially in the medium term, and are fixed in foreign exchange. In the rush to involve the private sector in power investments, the government did not carefully scrutinize the demand for power or sufficiently take into account the impact o f needed increases in real electricity tariffs. Also, the slowdown of the economy was to some extent unexpected, and thus lower electricity demand has necessitated big payments to IPPs for underutilized capacity. 1.11 Public sector involvement also extended heavily into industry, banlung, state trading, and education in the 1970s. Between 1969-70 and 1976-77 public investment in industry (including the steel mill project) alone increased tenfold in real terms while private industrial investment actually declined by 50 percent. Even though the private investment climate improved in the 1980s, the public sector presence in industry, trade, and banking remained heavy. By and large the public investments in industry, trade, and banks tumed out to be uneconomic. Losses in the public enterprises have been a major factor responsible fo he push toward privatization and a rethinkingo f the role o f the state in development (see B o x 1.1). Box 1.1: Privatization In 1991 Pakistan launched a broad-based program o f privatization with several objectives: reducing the drain o n government resources caused by losses o f state-owned enterprises, improving the efficiency of resource use, and providing greater incentives and opportunities for private sector investments because the high public debt burden had made public sector resources for development very scarce. Though the urgency of these objectives has grown over time, the actual progress in transferring assets and control to the private sector during the 1990s was slow. Although the sale o f shares (12 percent o f total equity, mostly to foreigners) o f Pakistan Telecommunication Corporation Limited (PTCL) when netted more than US$SOO m i l l i o n in foreign exchange and provided about 50 percent o f privatization receipts, management and control remained in the public sector. Then the government delayed and lost the prime market so that n o w very little interest or demand exists for telecom. The only significant progress in transferring both ownership and control t o the private sector was the sale o f t w o public banks, one power station, and a large number o f cement plants. The preparation and pace o f privatization has picked up greatly as part o f the reform program o f the Musharraf government. A major public bank has been sold, and a considerable number o f large transactions in banking and finance, o i l and gas, and power are ready t o be brought t o the market in the next twelve months during 2003-04. At the moment, however, a large part o f the banking system, as w e l l as almost the entire o i l and gas, telecommunications, and power sectors are in the public sector, and as discussed earlier the past and present losses o f power and the banking sectors remain a particular source of concern. If current plans t o privatize significant parts o f the power and banking sectors are not successfilly executed, public corporations will remain a serious drain o n public finances. This situation will also limit the role the government can and should play in providing services that the private sector cannot easily provide, such as l a w and order, justice, basic education and health, infrastructure, irrigation, roads, and urban and rural water supply. 1.12 An important consequence of the heavy emphasis on physical infrastructure development and public sector industry until relatively recently was the l o w priority accorded to the social sectors. The neglect o f education spending, especially in the 1980s, in face o f a rapidly growing population, i s a root cause o f the social lag with which Palustan i s n o w confronted. The l o w level o f human and s k i l l development i s in turn constraining productivity growth. I t i s also interesting to note that the nationalization o f private educational institutions in March 1972, undertaken ostensibly to accelerate educational development, had perverse effects o n the growth in basic education and the quality o f education. Though primary education was not nationalized, spending on primary education was squeezed because national resources allocated to education did not increase, the management capacity o f the government was overextended, and the urban bias in education was intensified by the takeover o f all high schools and colleges." 1.13 Another long-term consequence o f public expenditure management during the high-growth decades was the steep rise in public debt. The high overall level o f public spending, at close to 25 percent o f GDP in the 1970s and the 1980s, was unsustainable. As revenue growth lagged, fiscal deficits remained loSee Hasan (1998: 224-25). 4 large, averaging 7.5 percent o f GDP for the period 1980-2000 and showing n o strong downward trend. Consequently, public debt grew from 66.3 percent o f GDP in 1980 to 93.7 percent in 1990 and to more than 100 percent in FY2001. As a percentage o f government revenues, public debt grew from about 400 t o 670 percent over the period. This level i s extraordinarily high by international standards and also substantially higher than commonly judged sustainable levels (60 percent o f GDP and 300 percent o f fiscal revenue, respectively)." Initially, a large part o f the public sector deficit was financed with external borrowing at relatively concessionary terms. Over time, however, the government's reliance on domestic borrowing grew, with the result that domestic debt and borrowing costs became increasingly more important. The problem was compounded by a somewhat premature liberalization o f interest rates in early 1989 and moving all government borrowing to market-based instruments when the fiscal deficits remained high. Nominal interest rate payments were the main driving force behind the continued high deficits during the 1990s, as the primary deficit first declined and then disappeared in the late 1990s. But since a large part o f the deficits in the first half o f the 1990s was financed through money creation, the resulting inflation offset the gains in nominal interest rates. The real interest rate o n domestic public debt was actually slightly negative during 1990-96.12 Paradoxical though it may seem, the liberalization o f interest rates did not, in the ultimate analysis, lead to a deepening o f the public debt burden. It did, however, lead t o much larger fiscal deficits and much higher rates o f inflation. 1.14 The high and rising interest payments have made fiscal adjustment difficult. In the past three fiscal years (2001-03) there was an average primary surplus (excluding grants) o f 0.8 percent o f GDP, in contrast to a primary deficit o f nearly 4 percent o f GDP in the 1980s. Still the fiscal deficit remained high (see Table 1.3). Realizing the severity o f the problem, the government has recently set up a Debt Policy Coordination Office (DPCO) in Ministry o f Finance, with the main purpose o f establishing an improved debt management system. The DPCO i s formalized through the Fiscal Responsibility and Debt Reduction Ordinance (2003), which has been approved by the Federal Cabinet and i s likely to be tabled before the Parliament in near future. This Ordinance aims at enforcing and preserving the fiscal discipline through time bound restrictions o n revenue deficit and pubic debt. Table 1.3: Fiscal Indicators as a % o f GDP Total Total Fiscal Interest Primary Revenue Expenditure Dedcit/Surplus Expenditure Deficit'Surplus 1981-85 16.5 23.6 -7.1 3.1 -4.1 1985-90 17.9 26.3 -8.4 4.7 -3.7 1990-95 17.3 24.9 -1.7 5.4 -2.3 1995-00 16.3 23.1 -6.8 7.1 0.4 1999-00 16.3 22.5 -6.6 7.8 1.2 2000-01 16.2 21.0 -5.2 6.8 1.6 200 1-02 16.6 22.0 -5.2 6.8 1.5 Source: Economic Survey, Govemment Authorities, JMF and Staff estimates. Notes: 1) Fiscal deficit excludes foreign grants 2 ) Total Revenues do not include surplus of autonomousbodies and grants. 3) IMF total expenditure data has been used from 1999-00 onwards. Statistical Discrepancy has been taken into account to compute fiscal balance. 4) Total Expenditure for 2001/02 excludes one-time expenditures o f Rs. 52 billion on KESC recapitalizationand CBRbonds. If included, fiscal deficit would be 6.6% and primary surplus 0.1%. 1.15 Overall, public spending adjusted for inflation increased by 19 percent over 1991-2001. All o f this increase, however, was accounted for by the rise in interest payments; excluding interest, real See the Debt Committee Report, page 12. Debt Committee Report Table 17, page 53. 5 budgetary spending declined slightly. Both defense and non-defense spending (excluding interest) changed little over 1991-2001, though both fell sharply as a percentage o f GDP (see Figures 1.1 and 1.2). Figure 1.1: Trends in Selected Expenditures in 1999/00 Prices .............................................................................................................................. 1 +Total Expenditure +Total Non-Interest Expenditure +Total Non-Interest Non-Defense Expenditure I I Figure 1.2: Trends in Selected Expenditures as a Percentage o f GDP 1 30 I I 25 s .......... .% I 1 +Total Expenditure +Total Expenditure excl. Interest +Total Expenditure excl. Interest & Defence I 1.16 The burden o f stagnant public spending (excluding interest) has fallen especially heavily o n development spending. Development spending in real terms dropped 40 percent over 199 1-200 1, and i t s share in total public spending dropped fi-om 25 percent t o 12.8 percent over the period.13 During 2001-02 development spending recovered by 34 percent as public resource availability improved thanks t o large external grants. Still, public investment by general government (excluding investment by public enterprises) was only over 2 percent o f GDP in gross terms.14 If one allows for depreciation o f infrastructure like roads, irrigation works, and buildings, the public capital stock has probably grown little in recent years. 1.17 The consequences of growing interest payments, stagnation in revenues, and relatively inflexible defense spending over the past decade are brought out in Tables 1.4 and 1.5, which also show the trends in the level and pattern o f major categories o f non-interest, non-defense public expenditures. These figures, which include both current expenditures and development spending, give a better idea o f the real decline i s somewhat exaggeratedbecause o f a definitional change in development spending. l3T h i s l4T h ePublic Sector Development Program, which rose from 2.6 in FYOl to an average o f 3.3 percent o f GDP in FYs 02-03, includes a lot o f current spending covering the new category of pro-poor expenditures. 6 trends in spending and the sectoral shifts than the development spending figures alone. (See Figures 1.3- 1.14). Table 1.4: ConsolidatedFederal and Provincial Total Expenditure in 1999/00 Prices 1991192 1992193 1993194 1994195 1995196 1996197 1997198 1998199 1999100 2000101 Total Budgetary Expenditure 645,448 644,014 625,132 636,203 717,987 654,854 684,455 660,132 709,118 703,740 Interest 125,287 145,565 153,853 138,529 173,684 184,466 213,426 219,079 245,078 221,781 Defense 152,091 161,554 153,763 162,772 144,225 142,443 129,911 126,725 150,390 148,418 General Administration 38,963 40,963 40,548 50,958 65,318 54,786 42,002 46,606 46,584 55,758 Law & order 26,091 25,360 25,551 28,429 27,510 24,348 25,230 25,018 27,624 28,790 Education 50,420 50,365 51,673 54,883 55,944 49,820 53,380 50,755 54,002 53,965 Health and Population 15,450 15,487 16,526 18,714 19,242 16,966 17,390 17,519 20,026 18,373 Agriculture 13,244 11,939 11,446 11,194 10,997 8,717 8,676 8,608 8,520 8,528 Irrigation 21,833 24,529 23,458 27,764 28,441 18,247 19,339 17,320 17,611 16,481 Fuel and power 38,874 26,037 30,950 30,480 25,363 21,153 22,775 15,226 15,875 18,631 Transport and Communication 24,951 35,283 32,670 25,524 22,700 18,863 19,829 21,849 23,912 20,316 Others 138,243 106,933 84,694 86,958 144,562 115,046 132,498 111,427 99,497 112,699 Memorandum items: Total Non-Interest Expenditure 520,161 498,450 471,279 491,674 544,303 470,389 471,029 441,052 464,040 481,958 Total Non-interestNon-defense Expenditure 368,069 336,896 317,516 334,902 400,078 327,945 341,117 314,328 313,650 333,540 bf which: Development 183,418 140,745 116,915 118,317 124,940 99,990 113,205 103,242 90,681 87,4941 Sources: IMF, "Selected Issues and Statistical Appendix" November 9,2002; Planning Commission and Finance Accounts of the federal and provincial govemments. Note: Expenditures for Irrigation, Fuel & Power and Transport & Communication sectors include expenditures financed by the corporations (WAPDA, OGDC, PTCL and NHA) from the budgetary loans of the federal govemment. (Percent of GDP) 1991192 1992193 1993194 1994195 1995196 1996197 1997198 1998199 1999100 2000101 Total Budgetary Expenditure 26.7 26.2 24.5 23.7 25.5 23.1 23.5 21.9 22.5 21.8 Interest 5.2 5.9 6.0 5.2 6.2 6.5 7.3 1.3 7.8 6.9 Defense 6.3 6.6 6.0 6.1 5.1 5.0 4.5 4.2 4.8 4.6 General Administration 1.6 1.7 1.6 1.9 2.3 1.9 1.4 1.5 1.5 1.7 Law & order 1.1 1.o 1.0 1.1 1.0 0.9 0.9 0.8 0.9 0.9 Education 2.1 2.0 2.0 2.0 2.0 1.8 1.8 1.7 1.7 1.7 Health and Population 0.6 0.6 0.6 0.7 0.7 0.6 0.6 0.6 0.6 0.6 Agriculture 0.5 0.5 0.4 0.4 0.4 0.3 0.3 0.3 0.3 0.3 Irrigation 0.9 1.o 0.9 1.0 1.0 0.6 0.7 0.6 0.6 0.5 Fuel and power 1.6 1.1 1.2 1.1 0.9 0.7 0.8 0.5 0.5 0.6 Transport and Communication 1.o 1.4 1.3 1.0 0.8 0.7 0.7 0.7 0.8 0.6 Others 5.7 4.3 3.3 3.2 5.1 4.1 4.6 3.7 3.2 3.5 memo items: Total Non-InterestExpenditure 21.5 20.2 18.5 18.6 19.4 16.6 16.2 14.6 14.7 14.9 Total Non-interest Non-defense Expenditure 15.2 13.7 12.4 12.5 14.2 11.5 11.7 10.4 10.0 10.3 o f which: Development 7.6 5.7 4.6 4.4 4.4 3.5 3.9 3.4 2.9 2.7 Sources: IMF, "Selected Issues and Statistical Appendix" November 9,2002; Planning Commission and Finance Accounts of the federal and 1.18 Though the overall level o f non-interest, non-defense spending was essentially stagnant over m u c h o f the past decade, the pattern o f spending shifted somewhat. Public sector p o w e r spending 7 declined as financing fiom WAE'DA's own resources was excluded and more reliance was placed o n the IPPs for thermal power generation. In contrast, there was real growth in spending o n primary education under the impetus o f the Social Action Program. But even in primary and secondary education spending did not increase as a percentage o f GDP in the 199Os, and their shares remained at the relatively l o w levels o f 0.8 and 0.5 percent, respectively, in 2000/01. Spending o n college, university, and professional education declined by nearly 20 percent in real terms and even more sharply as a percentage o f GDP: only 0.3 percent o f GDP was being spent o n higher education, including colleges, in 2000/01 compared with 0.5 percent in 1990/9 1. ~~ Figure 1.3: General Administration Figure 1.4: Law & Order Expenditures in 1999/00 Prices I 31 65 ~ 29 60 'i;: I - 55 27 - 0 50 z45 240 12 21 35 1 19 ' 30 17 I 25 I 2 0 4 I Figure 1.5: Education Expenditures in 1999/00 1 I Figure 1.6: Health and Population Prices Expenditures in 1999/00 Prices I 60 1- ~ 1 I lI = o 50 1 545 40 1999/00 Prices Prices 14 31 13 12 26 - .-g 11 2 10 I - - f ,021 .- I a $ 9 I 2 16 ~ I 8 11 7 ~ 6 6 I iource: Economic Survey 2001/02 and Finance Accounts o f thefederal and provincial governments. 8 I I Figure 1.9: General Administration Expenditures Figure 1.10: Law & Order Expenditures as a as a YOof GDP Yo o f GDP 2.4 ~ 1.2 2.2 2.0 1.8 IS S 09 1.6 08 1.4 I 1.0 J 11 0'7 0.6 i Figure 1.11: Education Expenditures as a YOof ~ Figure 1.12: Health and Population GDP Expenditures as a YO o f GDP 2.2 f I-- 0 75 0 70 0 65 s. 0 60 1 0 55 1.6 1.5 0 50 - ~ p $ ; ; ~ ~ m o ~ q q b m w ~ m m o - 2 2 m o t E x g ~ g ~ 2 2 ~ 2 z : ~ Figure 1.14: Irrigation Expenditures as a % o f ' Figure 1.13: Agriculture Expenditures as a O h o f GDP GDP I 0.60 , i 1.3 0.55 1.1 - A 0.50 0.9 \ 0.45 - S 0.40 S 0.7 4' 0.5 25 030 0 1 0.3 1 1 ource: Economic Survey 2001/02 and Finance Accounts o f thefederal and provincial governments. 1.19 Agriculture and irrigation spending went down sharply in real terms, especially in the second h a l f o f the 1990s. In 2000-01, agriculture and irrigation spending was 30 percent lower than in 1990-91 and constituted only 0.8 percent o f GDP, compared with 1.6 percent a decade earlier. Spending o n transport, communication, highways, and buildings dropped by 40 percent or 3 percent o f GDP over the period. A large part o f spending o n the Lahore-Islamabad Motonvay was outside the budget as it was financed directly from the supplier's credit. Though the Highway Authority was in n o position to service any debt, 9 i t s borrowing was outside the budget and i t s spending and projects were excluded from normal business processes. Nevertheless, the Motorway-related debt did increase the total public debt. 1.20 Spending o n general administration and law and order maintenance fared better in the 1990s than public expenditures o n social and economic services. Indeed, the real expenditure on general administration nearly doubled between 1990-91 and 1995-96, far outstripping growth in either GDP or revenues. Clearly, financial discipline was lax in these years and low-priority spending increased. With the financial constraints deepening, the expenditure on general administration was reduced sharply in late 1990s in real terms. This was achieved mainly by delaying the civil service salary adjustments, which were finally made in 2001 and were fully reflected in the 2001-02 budget. The net result was that spending on l a w and order and general administration grew broadly in line with GDP, and the relative reduction in the non-interest, non-defense spending o f nearly 5 percentage points o f GDP (from 14.5 percent to 9.5 percent over 1991-2001) fell entirely on economic and social spending. 1.21 The negative impact o f the squeeze in social and development spending during the 1990s on growth and poverty reduction was greatly aggravated by declining effectiveness in the use o f public resources, both borrowed and domestic. First, many low-priority projects were funded; the most egregious example was the large uneconomic investment in the Lahore- Islamabad Motorway to the neglect o f the iational, provincial, and rural network. The Saindak copper project and the Tamir-i-Watan program are 3ther examples o f misallocated resources. Second, many large and important projects and programs, iotably the Left Bank Outfall Drain (LBOD), the National Drainage Program (NDP), the Chashma Right Bank Canal Project, and the Social Action Program, did not have the desired impact because o f poor lesign, poor feasibility, poor coordination, delays in implementation, or leakages in the system. Figure 1.15: Spending on Education ” Secondary Tertiary 90 f; 7,000 - 7 I 270 . ... ... ... ... ... ... ... ... ... ... . 6,000 go & 5,000 . ... ... ... ... ... ... ... ... ... .. 6 e 4,000 ............................. I . ... ... ... ... ... ... ... ... ... 3,000 ~ 2,000 . ... ... ... ... ... ... ... ... ... ... . . ..... ..... ..... ..... ..... ..... .... .- ... ..... ..... ..... ..... ..... ..... .. v . ... ... ... ... ... ... ... ... ... ... ... .. x 1,000 . ... ... ... ... ... ... ... ... ... ... . n g o 2 150 ~ ” Annual spending per facility 1.22 The attempt t o improve the delivery o f social services under the much-heralded Social Action Program was a particular disappointment both because there was n o sustained increase in the level of social spending and because poor governance l e d to tremendous leakage and waste o f resources. The problems o f governance, related to personnel and resource management, were the critical constraint t o achieving the government’s development goals. These problems include absenteeism, hiring o f staff on grounds other than merit, frequent transfers, and mal-distribution o f staff to the disadvantage o f rural areas. 1.23 In addition, some perennial public expenditure issues were to some extent intensified in the 1990s because o f stagnant or declining overall levels o f economic and social spending. Because the 10 development budget cuts were particularly severe, there was inadequate spending in relation to the optimum phasing o f projects, and resources were spread thinly, delaying the benefits. 1.24 Notwithstanding the slowdown of development spending, the imbalance between maintenance o f existing assets and creation of new ones intensified. As recurrent expenditure budgets stagnated or actually declined, a greater portion of the budget was preempted by establishment costs. The nonsalary expenditures crucial for running education and health facilities and maintaining physical infrastructure were badly squeezed. 1.25 For example, official data show that during the 1990s the real per facility non-development expenditures declined for each level o f education, most drastically for the tertiary level. Accounting for the periodic increases in the salaries o f government employees during this period, i t i s possible to infer an even steeper decline in the non-salary component o f recurrent expenditures. This decline in quality-related operation and maintenance expenditures has led to a marked deterioration in school infrastructure, and a large number o f schools across the country are in a state o f disrepair and lack basic amenities. A survey o f rural facilities in 200 1 showed that only one-third o f rural schools had an adequate building, only about half had h i t u r e , less than one-quarter had textbooks, 52 percent had dnnkmg water, 48 percent had toilets, less than one-third had electricity, and only one-sixth had a separate toilet for girls. 1.26 Service delivery and sectoral reforms have also suffered owing to a lack o f continuity at key leadership positions. This i s especially true for important service delivery departments, like education and health, in the provinces. Lack o f control o f line departments over appointments and transfers can result in frequent transfers o f civil servants. Anecdotal evidence from Sindh suggests that department heads have an average tenure o f just over a year-barely enough time to understand the issues o f the sector and far too little to develop a policy to tackle these issues. Restructuring and Refocusing an Extended Public Sector 1.27 The preceding discussion brings out two key points. Palustan's public sector became seriously overextended in the 1970s and 1980s. Until the 1960s the massive expansion o f public sector activities did not seriously hinder the development o f the private sector. In addition to providing normal public services such as l a w and order, justice, health, and education, the public sector concentrated o n mobilizing large foreign aid and loan resources and expanding investments in physical infrastructure, especially water and power. The state role in industry was confined t o a few strategic projects. This situation changed in the 1970s, when very large-scale expansion o f public investments in industry and trade actually led to a. sharp decline in private investment in manufacturing. Even though private investment recovered after the late 1970s, the inexorable growth o f public expenditures, notably defense, continued in the 1980s: the level o f public expenditures as proportion o f GDP rose further from 23.5 percent in 1979-80 to a peak o f 26.7 percent in 1987-88. The efforts in the 1990s to restructure and refocus the public sector toward i t s traditional functions-maintaining macroeconomic stability, provision o f l a w and order, basic social services, and essential infrastructure--were not very successful owing to a number o f factors, including a lack o f strong political will resulting in part from political instability, limited institutional capacity, poor governance, and limited fiscal space. 1.28 The current economic and social reform program attempts to deal with the various constraints o n progress toward a viable and vibrant public sector in a comprehensive fashion. Visible progress has been made in a number o f areas, such as reducing macroeconomic imbalances and containing the debt burden, accelerating the efforts toward privatization, and reorienting public expenditures toward social services and poverty groups. But in other areas, notably governance and strengthening o f institutional capacity, the reform measures have yet to yield concrete improvements in the delivery o f public services. M u c h hinges on the success o f the ambitious program o f devolution o f authority to elected representatives at the district 11 level. The following chapters of this report discuss the various challenges the government faces and some hard choices it will have to make in dealing with difficult issues. Here i t i s necessary only t o highlight four broad areas, which will be critical for success in refocusing the public sector and increasing i t s . effectiveness: . Creation o f fiscal space for spending excluding interest, defense and public enterprise losses. The successful completion o f the present privatization program, including early progress . . especially on the relatively slow-moving power privatization plans. Improved delivery o f social services through devolution and private-public partnerships. Improved govemance and more effective delivery o f law and order and justice. 1.29 The creation o f fiscal space for public spending in support o f growth revival and poverty will b”e central t o the better management o f the public sector. As discussed in the next chapter, if broad-based reform efforts continue, prospects are good for expanding real spending in the critical public service areas. 1.30 The preparation and pace o f Table 1.6: Planned Privatization of State Enterprises privatization has picked up greatly as a crucial November 2002-June 2004 part o f the Reform program o f the Musharraf Sale Proceeds Government. A considerable number o f large (Rs Million) transactions in banlung and finance, o i l and November 2002 -June 2003 21.813 gas, and power are ready to be brought t o the NBP Shares Lot 2 767 Pak Saudi Fertilizers (lOYo Shares) 815 market by the end o f FY04 (see Table 1.6). GOP Working Interest in 6 Oil Fields 540 Still at the moment, large part o f the banlung UBL 11,749 system, almost the entire o i l and gas, Ban1AI-Falah 620 telecommunications and power sectors are in MCB Zares through CDC 66 1 the public sector and as discussed earlier the Pakistan Oil Fields shares through CDC 4,699 past and present losses in the power sector Attock Oil Refinery shares through CDC 1,011 Mutual Right - ICP Lot A Fund 175 remain a particular source o f concern. Mutual Right - ICP Lot B Fund 303 AI-Haroon Building 83 1.3 1 The political will to narrow Pakistan’s Falaties Hotel 450 social gap appears strong. Nonetheless, the Pak Arab Fertilizers 6,000 sheer magnitude o f the task means that large July 2003 -June 2004 93,992 financial resources and exceptional PSO (51% shares) 20,300 institutional arrangements will be needed t o PTCL (26% o f GOP shares) 34,800 OGDCL (5% shares through IPO) 4,640 achieve even such basic and longstanding NPCC (100% o f GOP shares) 667 goals as universal elementary education. FESCO (56% o f GOP shares) 2,900 JPC (1 % o f GOP shares) 2,030 1.32 Finally, improved govemance remains HBL 23,200 a prerequisite not only for reviving private National ConstructionCompany Limited 100 investment, but also for improving the Lasbella Textile Mills 140 Lyllpur Chemicals 150 effectiveness of public expenditures, Bolan Textile Mills 140 especially in the delivery o f services to the Pak American Fertilizer 4,500 poor. Republic Motors 45 AC Rohri 280 Others 100 TOTAL 121,865 12 Key Gaps in Public Goods 1.33 A b r i e f review o f the present position with respect to the social sectors and key physical infi-astructure i s helpful in explaining the magnitude o f the challenge the public sector faces in providing more and better public services. 1.34 I t i s well known that Pakistan underperforms other countries with similar per capita incomes in delivery o f basic social goods. What i s perhaps less commonly realized i s that the gaps in public goods delivery have been widening during the past decade, especially delivery o f services to the poor. Furthermore, the inadequacy o f public investment in key areas i s taking a heavy toll in efficiency losses for the economy. Health 1.35 Palustan's population o f about 140 million i s growing by 2.4 percent per year. The high fertility rate o f 4.8 children per woman puts an enormous burden o n the health o f mothers, resulting in a high maternal mortality rate o f 200 per 100,000 live births. With 80 percent o f women s t i l l delivering without assistance from trained personnel, a l o w contraceptive rate o f 28 percent, poor nutritional status, and only 48 percent o f tetanus toxoid immunization o f pregnant women, one in about three hundred women die in childbirth. 1.36 N o change in malnutrition Table 1.7: Pakistan in comparison with Regional Countries in Health and Population Outcomes levels in children has taken place in the 1990s: the percentage o f Infant Under 5 Annual Total Life Mortality Mortality Population Fertility children under five who are Country Expectancy Rate rate Growth rate underweight ranges fi-om 33 to 45 Pakistan 63 85 110 2.4 4.8 percent; micronutrient deficiencies Bangladesh 61 60 82 1.7 3.1 Bhutan 62 62 n.a. 2.9 5.4 o f vitamin A and iron persist; and India 62 69 87 1.8 3.1 the immunization coverage rate i s Nepal 58 73 104 2.4 4.3 only 52 percent, with a large urban- Sri Lanka 73 15 17 1.6 2.1 rural differential and a smaller Source: Pakistan Country Update; World Bank, 2003. gender differential. Moreover, the " infant mortality rate o f 85 per 1,000 live births i s 10 percentage points higher than the average for low- income countries and 16 percent higher than the average for South Asia (see Table 1.7). 1.37 Public dissatisfaction with the Table 1.8: Present Status of Health & PopulationOutcomes government's outpatient health services i s and MDG Targets for Pakistan reflected in the large numbers o f people seelung Millennium the help o f private practitioners. Eighty percent o f Indicators Latest Development Targets outpatient contacts are with private practitioners. 110 47 Govemment regulation o f private practitioners i s virtually nonexistent. Poor access to basic services Infant Mortality Rate 85 40 Proportion o f fully immunized such as safe water and sanitation, along with poor children 12-23 months 50 80 nutritional status, play a significant role in the Matemal Mortalitv Ratio 300 120 20 100 l o w compared with those o f other countries in the region and amounted to 0.6 percent o f GDP in 2000-01. The l o w level o f resources contributes significantly to the poor quality o f health care, as do various inefficiencies in the allocation o f resources and waste attributable to weak governance. The performance o f the health sector must be improved drastically if Palustan i s to reach the MillenniumDevelopment Goals (MDGs) by 2015 (see Table 1.8). 13 Education 1.38 Pakistan’s education sector has performed poorly in the past, with enrollment levels among the lowest in the developing world, and gender disparities and rural-urban gaps among the widest. 1.39 Nationwide, net enrollment rates at the primary level in 2000-01 were about 53 percent, masking considerable differences between rural and urban populations (67 percent versus 45 percent) and males and females (57 percent versus 44 percent).” Although the gender gap narrowed slightly in the 1990s, this improvement was in part due to the decline in male enrollment levels. Quality problems are widespread and often arise from institutional deficiencies such as nonfunctional schools. Were i t not for the expansion o f private education, the current situation would be much worse. From 1981 t o 1998, private school enrollment grew by 937 percent, compared with population growth o f 57 percent. 1.40 According to the Higher Education Task Force report, Pakistan’s university and college system i s in an unreformed state. The country has a population o f 18 million in the 17- to 23-year age bracket, o f which only 2.6 percent are enrolled in higher education institutions (1996 data), compared with 6.2 percent in India (1990 data) and 12.7 percent in Iran (1994 data). Public universities, which enroll about 85 percent o f university students, are not only resource deficient, but also plagued by inefficient governance and politicization. 1.41 Literacy levels, reflecting historical Table 1.9: Public Expenditure on Education investments in education, are similarly poor (see Table for Selected Countries (% ofGDP) 1.9). In 2000-01 literacy was estimated at 48 percent, Expenditure on Education with male literacy at 62 percent and female literacy at 34 percent. Although adult literacy rates are similar 2.3 across South Asian countries today (with the exception 3.7 o f Sri Lanka, where they are considerably higher), srihnka 3.1 Palustan’s lower school enrollment rates imply that it Nepal 2.9 will lag behind i t s neighbors in improving literacy in \Average for South Asia Average for low-income I 3.1 I the future. 3.3 countries Source: Pakistan Country Update, World Bank, 2003. Irrigation Table 1.10: Canal Water Diversion 1.42 h g a t i o n and drainage are crucial to Pakistan’s Canal Head agriculture sector, with about 80 percent o f arable lands and Year Diversions MAF Remarks 90 percent o f agricultural output (25 percent o f GDP) 1949-50 67.0 Just after independence entirely dependent on irrigation. Despite the country’s Before Indus Basin impressive irrigation infrastructure, overuse, aging, and poor 1959-60 85.0 Treaty maintenance have resulted in extremely l o w efficiency o f 1967-68 95.0 Post Mangla water delivery. Average delivery efficiency i s about 35-40 After Tarbela percent from the canal head to the root zone (primarily 1975-76 101.0 Commissioning because o f seepage). Other than the Chashma Right Bank 1977-82 105.0 Post Tarbela (average) Canal, expansion o f irrigation in the past fifteen years has Just before Water 1990-91 109.8 been limited to small isolated schemes outside the Indus Accord Basin and increased use o f tubewells (see Table 1.10). The 199 1-01 102.6 Post Accord Average improvements made in the irrigation system through 2000-01 81.1 Inflow 99.9 MAF (drought) investments over the years have been offset by deterioration Source: Accelerated Development o f Water Resources and Irrigated Agriculture, World Bank, 1993. The net enrollment rate i s from the Pakistan Integrated Household Survey (PIHS). The govemment states net enrollment as 66 percent. In either case, Pakistan compares poorly with other low-income countries. T h e rural-urban and male-female differential data pertain to 1998-99. 14 o f the system because o f inadequate operation and maintenance (O&M budgets fall short o f requirements by more than 40 percent, whereas salaries o f the overstaffed irrigation department account for more than 50 percent o f the total O&M budget) and poor performance o f the public institutions handling the irrigation and drainage system. 1.43 Unlike the Indus Basin Irrigation System (IBIS), the drainage network i s not continuous and integrated. L o w water charges, particularly compared with the productivity o f irrigated agriculture, created incentives for farmers to increase water flow through illegal pumping o f water, resulting in inequitable distribution o f water and inflicting physical damage o n the canal system. Inadequate drainage facilities have led to resource degradation, threatening the sustainability o f irrigated agriculture. Pakistan’s irrigation system, which was designed as run-of-the-river to maximize crop area, has created serious problems o f waterlogging and salinity in the command area owing to lack o f drainage facilities. Waterlogging problems claim about a third o f gross cultivable land, with another 16 percent o f cultivable area suffering from salinity. Depending on the degree o f waterlogging and salinity, crop yields are reduced by 25-60 percent from their potential levels. Table 1.11). Rural water supply coverage i s particularly Water Supply Water Supply Water Supply Province Coverage Objective Objective l o w in the large provinces o f Punjab and Sindh. (2000) (2011) (2025) Punjab 27 40 75 Transport Sindh 10 40 75 NWFP 45 65 75 1.45 Transportation currently accounts for about 10 Baluchistan 55 65 75 percent o f GDP and 17.3 percent o f gross capital FATA 63 75 75 FANA 16 75 75 formation. Palustan’s transport sector i s inefficient and CDA 55 100 100 costly and imposes an annual cost o f Rs. 180 billion, or 5 AJK 32 75 75 1.46 The road sector dominates the transport sector, accounting for 90 percent o f passenger traffic and 95 percent o f freight. The network i s comparable in density (0.3 M s q u a r e lan) to other developing countries in the region. The road network, however, faces severe limitations in the form o f poor quality, a serious shortfall in tertiary roads needed for basic access in poor areas, investments in economically unviable megaprojects, a huge unfinished portfolio o f investment works that has increased budget rigidities, and one o f the worst safety records. 1.47 The road sector has been the main recipient o f public sector funding. Total public expenditures o n roads are about Rs. 20 billion per year, o f which 65 percent are spent on national roads. Road maintenance expenditures have been about 20 percent o f the required amounts for the national and provincial networks. I n 2001, whereas Rs. 21 billion were spent o n roads, assets worth about Rs. 16 billion were lost as a result o f deferred maintenance. 15 1.48 The rapid growth o f the road sector has in part been at the expense o f railways. Currently, Palustan Railways (PR) accounts for only 5 percent o f freight traffic, compared with 73 percent in 1955- 60. I t s share o f the passenger market has dropped from 42 to 9 percent in the same period. This shift i s primarily due to mismanagement and resulting poor service. 1.49 P R i s one o f the largest public money-losing operations in the country, with recent (1997-2000) annual deficits regularly hitting the Rs. 6-7 billion mark. Since 2000 the new management has succeeded in slowing the loss-malung trend. The reduction in the railway deficit i s not sustainable, however, in the face of conflicting fare structures, decaying rolling stock and infrastructure, and huge fixed costs coupled with l o w employee productivity. Power 1.50 Palustan’s power sector faces a number o f challenges in contributing more effectively to the country’s development. Only about 60 percent of the population, mainly in urban and periurban areas close to the transmission corridors, currently has access to electricity, whereas the bulk o f the rural population relies mainly on traditional fuels and kerosene for meeting i t s cooking and lighting needs. The Strategic Agenda for Public ExpenditureManagement 1.5 1 This report focuses principally on three key dimensions o f better public expenditure management in Palustan. 1.52 First, it i s paramount to continue financial discipline and reduce the overall size o f the public sector deficit, including the sizable losses o f public enterprises. The modest progress made in reducing the government’s fiscal deficit during the past few years has been undermined by the persistence o f high level o f losses o f public enterprises, especially WAPDA and KESC. To reduce the unsustainable burden o f public debt, the fiscal deficit, which has averaged 5.5 percent o f GDP (excluding grants) and 3.4 percent (including grants) during the past three years, must be brought down further. Provision needs to be made for the large and continuing public enterprise losses and unfunded contingent liabilities o f the public sector.16 A strong and successful government revenue mobilization effort, which will gradually raise the ratio o f revenues from 17 percent o f GDP (FY02) t o say 20 percent over the next decade, remains central to restoring Palustan’s fiscal health. But as the experience o f the past few years shows, the structural weakness in the taxation structure (relatively heavy dependence o n trade taxes) and the institutional weaknesses in the tax collection machinery (especially o n the income tax side) will continue to dampen revenue growth for some time. Thus it will be prudent to assume, at best, only moderate growth in the ratio of government revenues to GDP over the next five years. 1.53 Even on the assumption of a steady increase in the ratio o f government revenue to GDP, the growth in overall public spending in real terms will be modest over the next few years because o f the need to reduce the deficit further and to fund public enterprise losses and contingent liabilities. Indeed, in the medium t e r m overall public spending as a proportion o f GDP i s unlikely to increase f i o m the level o f 22 percent witnessed in recent years, even if grant assistance remains at a relatively high level. 1.54 Yet the revival o f growth and the reduction in poverty, the goals articulated in the Interim Poverty Reduction Strategy Paper (WRSP), will require substantially higher infrastructure investments as w e l l as higher social spending. The dilemma i s this. With the relative size o f overall public expenditures l6 These figures exclude the one-time expenditures o f Rs. 52 billion on K E S C recapitalization and Central Board o f Revenue bonds in FY 2002. If these are included, the average deficit for the past three years excluding grants i s 6.1 percent and including grants i s 4.6 percent. 16 remaining stagnant, more adequate support for the government’s economic and social goals can be achieved only through a restructuring o f expenditures away from interest payments, defense, and nonessential expenditures toward social and development outlays, adequate maintenance o f infrastructure, and support for improved governance. What are the possibilities o f restructuring public expenditures away fi-om the inflexible burden o f defense and interest payments, as well as public enterprise losses? What are the policy and institutional actions that can facilitate a better allocation o f public resources within the major economic and social sectors? This then i s the second major theme o f the analysis in this report. 1.55 While public development and social spending must sharply recover from the extremely depressed levels o f the past decade and must play a significant part in reviving growth and reducing poverty, the biggest economic and social gains must come from improved efficiency in the use o f public resources. Improving the effectiveness o f spending not only through a better allocation o f resources and a more careful choice o f priorities, but also through better implementation, more efficient delivery o f services, and plugging o f leakages and waste i s the third strategic issue discussed in this report. This goal requires, above all, restoring the capability and the authority o f public institutions, especially those concerned with economic and financial management and the delivery o f social services. 1.56 Public institutions in Pakistan have suffered a long-term decline because o f poor pay for civil service, increasing corruption, and widespread political interference in administrative m a t t e r ~ . ’The ~ institutional issues have a special urgency because o f the major devolution o f responsibilities to the district level. The turnaround in efficiency o f education and health spending that i s needed n o w hinges on the success o f the new district governments. But the adequacy o f planning, budgeting, expenditure control, and monitoring and evaluation processes at all levels o f government need to be examined in relation to both the distribution o f responsibilities among various levels o f government and the capability o f the civil service to respond and contribute to the desired productivity gains. In looking at the selected sector allocation issues, the analysis focuses more on the institutional strengthening needed to determine priorities and improve implementation and less o n the evaluation o f specific government plans that, in any case, become quickly out o f date. The effectiveness o f expenditure analysis gives attention to both the process issues and the institutional capacity problems. “Pakistan: A Framework for Civil Service Reform in Pakistan,” December 1998, World Bank; “Pakistan Development Policy Review: A New Dawn?” April 2002, World Bank; and “Pakistan Country Financial Accountability Assessment” (Phase I), World Bank, 2001. 17 CHAPTER 2: ENSURING FINANCIAL DISCIPLINE SPACE AND CREATINGFISCAL Introduction 2.1. T h i s chapter examines the key elements o f macroeconomic and fiscal policies that will ensure steady and simultaneous progress toward the goals o f debt burden reduction, acceleration in economic growth, and poverty alleviation with the help o f a medium-term macroeconomic and fiscal fkamework that integrates growth, debt reduction, and poverty alleviation goals. 2.2. The focus, however, i s on public expenditures in the context o f a consolidated medium-term budgetary framework18 and on the critical issue o f h o w to create fiscal space for growth-enhancing and poverty-reducing expenditures while maintaining financial discipline and malung progress in reducing public debt burden to sustainable levels 2.3. This focus involves the examination o f the scope o f expanding social and development spending, while maintaining the current progress on public debt burden reduction (and keeping aggregate spending at about the same level in % o f GDP). The illustrative baseline scenario explores the scope for mobilizing more government revenues, cutting public enterprise losses, reducing o f interest payments, containing defense spending and mobilizing privatization proceeds. This scenario representing macroeconomic stability and continued structural reforms resulting in productivity improvements i s contrasted with a l o w case in which the structural reform sputters; there i s n o significant progress in tax revenue mobilization; public enterprise losses and defense spending increases in relative terms; there i s little further progress in reducing the fiscal deficit; and consequently there i s less external support in the form o f grants and concessionary loans. Quite obviously in this l o w scenario there will be only limited progress towards the goals of debt reduction, revival o f growth and poverty alleviation. Turn Around in Debt Position 2.4. The debt problems, which have been a major source o f the sharp slowdown in Palustan’s economic growth, and the consequent stagnation in poverty incidence, over the last decade have been brought under control. While the debt burden, especially the public debt burden, i s s t i l l at an unsustainable level and needs to be brought down steadily, the risk that economic and other shocks will lead to a worsening o f the public debt burden in the medium term appears small. 2.5. For nearly two decades, the rising interest payments limited fiscal space and squeezed public spending on development and social sectors. At their peak in 1999-2000, interest payments accounted for nearly h a l f the govemment revenue. The high interest payments reduced pro-growth public spending directly while indirectly they discouraged private investment because the large government borrowing kept domestic real interest rates high and tended to crowd out the private sector. Private investment also suffered due to the shortages o f complimentary infrastructure, limited access to international capital markets, and declining confidence in the domestic currency, which contributed to capital flight. The medium-term budgetary framework being presented here consolidates the budgetary accounts o f both federal govemment and the provinces. The purpose i s merely to illustrate the public expenditure level and pattem for the country as a whole and to highlight future expenditure options. Since Pakistan i s a federation, the operational budgetary frameworks must be prepared separately for the central govemment and for the provinces o f Punjab, Sindh, North West Frontier, and Balochistan. In addition, AJK and the Northem Areas have independent budgets. 2.6. In sharp contrast to the almost inexorable rise in the debt burden in the 1980s and 199Os, there was a significant decline o f 2 1 percent in the present value o f public debt from 600 percent to 472 percent o f government revenues, between mid 1999 and mid 2003 (Table 2.1). The ratio o f interest payments to government revenues dropped to about 30 percent during FY 2003. The drop in the external public debt burden in present value terms and as a proportion o f exports o f goods and non-factor services has been even more pronounced -- 39 percent between mid 1999 and mid 2003. The dramatic increase in foreign exchange reserves, to $9.5 billion by mid 2003 from $0.9 billion in mid 2000, has transformed the liquidity position. The improvement in debt indicators i s attributable to the impact o f policy reforms as well as a number o f external factors that have resulted in a narrowing o f fiscal and external current account deficits (excluding private and public transfers), reduction in cost o f government borrowing, sharp rise in private remittances, and increase in concessional and grant external assistance. The very favorable terms o f rescheduling o f Paris Club debt have also contributed significantly t o the reduction in the burden o f debt. Still the external debt burden i s too high. By the criteria applied under HIPC, the present value o f public and publicly guaranteed external debt in relation to exports o f goods and non- factor services should not exceed 150 percent. In Palustan, this ratio i s currently about 200 percent. Table 2.1: External and Public Debt Burden 1989/90 1995/96 1998/99 1999/00 2000/01 2001/02 2002/031’ (US$ Billion) Total Extemal Debt (excl. foreign exchange liabilities) 19.4 31.0 33.5 32.4 32.2 33.5 33.4 Total Extemal Debt (excl. foreign exchange liabilities) 48.5 48.7 58.8 59.5 54.8 56.5 48.7 as a % of GDP Total External Debt (excl. foreign exchange liabilities) 234.7 330.6 341.0 324.9 327.0 287.3 287.2 as a % of Foreign Exchange Eamings’ Total External Debt (incl. foreign exchange liabilities) 22.0 31.3 34.7 38.1 37.2 36.6 35.5 Total External Debt (incl. foreign exchange liabilities) 55.0 49.1 61.0 69.9 63.4 61.8 51.8 as a % o f GDP Total External Debt (incl. foreign exchange liabilities) 266.2 333.3 353.9 381.7 377.9 314.2 305.4 as a % of Foreign Exchange Eamings (Rs. Billion) Total Public Debt 802.1 1,891.0 3,003.7 3,175.8 3,707.5 3,633.7 3,689.9 Total Public Debt as a % of GDP 93.7 89.2 102.2 100.9 108.3 100.1 91.8 Total Public Debt as a % of Revenues 505.1 510.3 632.5 619.6 670.5 588.9 512.0 (US$ Billion) Extemal Public & Publicly Guaranteed Debt 19.1 27.9 30.2 29.6 29.8 31.2 31.3 Memorandum items: Present Value of Total Public Debt as a % of Revenues” 599.8 528.1 472.3 Present Value of External Public and Publicly Guaranteed Debt as a % of Exports of Goods and Non-Factor 320.1 248.7 196.1 I-...- +mice9-- 3141 Source: “Debt Burden Reduction and Management Strategy”, GoP, 2001; Government Authorities; and Staff Estimates. ” Data for 2002/03 is provisional. ” Foreign exchange eamings include goods, services, income, and workers’ remittances (World Bank definition as used in the Global Development Finance reports). ” Estimates assume a 10% grant element in 1998199, increasing t o an expected 20% in 2002103. ” Estimates assume a 10% grant element in 1998/99, increasing to an expected 20% in 2002/03. 2.7. The government’s basic approach to debt management and reduction i s sound. The government attaches high importance to steady reduction in the burden o f public debt but has at the same time opted for a ten-year time frame for attaining debt sustainability, defined as 60 percent o f GDP, which will not require harsh cuts in the already squeezed levels of public spending. The government recognizes that debt 19 reduction, vital for long-term sustainable growth, cannot be the only economic goal even in the short t e r m and must be balanced with the need to accelerate growth, create jobs, and reduce poverty in the short term. W e endorse this target and the basic strategy. As long as the public debt i s moving on a favorable downward trajectory, most o f the negative effects o f high burden o f debt on growth and poverty, already noted, can be minimized. 2.8. The Government has approved a Draft Fiscal Sustainability and Debt Limitation Law. I t i s a good step towards operationalizing the government’s debt reduction strategy. It specifies fiscal r u l e s t o constrain discretion on fiscal policies and enforce fiscal discipline consistent with the target o f reducing gross public debt to 60 percent of GDP by FY11/12. It limits annual issuance o f explicit guarantees to less than 2 percent o f GDP, including rolling over existing guarantees and payments o f called guarantees. The Draft L a w provides for establishing a Debt Policy Coordination Office (DPCO) to serve as a secretariat for the Fiscal Sustainability and Debt Limitation Law. I t also creates an enforcement mechanism through public scrutiny by requiring fiscal transparency and open fiscal reporting. Notwithstanding this progress it would be useful to further refine the Debt Strategy with the following improvements: expand the coverage o f guarantees subject to the limit in the Draft L a w to other explicit and implicit contingent liabilities; apply the reporting requirements to provincial governments as well; and supplement the requirement for a mid-year economic report with a requirement for corrective fiscal measures when there are deviations f r o m mid-year targets. The Government could also increase the overall debt strategy’s credibility and i t s operational relevance with the additional suggested refinements presented in B o x 2.1. 2.9. The ten-year goal should be to reduce public debt as a share o f revenues to 300 percent (and n o higher than 275 percent in present-value terms). For the medium term, a goal o f reducing public debt as a share o f revenues from the 512 percent in mid 2003 to 400 percent by mid 2007 appears desirable. This step will reduce the public debt burden in terms o f GDP t o 70-75 percent by the end o f FY07, which will be o n track to reach the 60 percent goal by FY12. 2.10. What would the target o f public debt reduction to 400 percent o f government revenues by F Y 0 7 mean for the level o f fiscal deficits during FYs 2003-07? M o r e important, h o w can the medium-term economic and fiscal policies ensure achievement o f the debt reduction goal consistent with progress in reaching goals o f growth acceleration and poverty alleviation? B o x 2.2 presents an analytical framework for examining these issues. Broadly speakmg, as the analytical framework suggests, three sets o f economic policies will be key to the success o f the government’s strategy o f growth revival and debt reduction: Policies that slow down the future rate o f government borrowing by keeping the primary balance positive (normal fiscal balance before interest payments) and reduce public enterprise losses and payments o f these and other contingent liabilities. Privatization can help limit public enterprise losses and generate revenues, which reduce the need for new public borrowing. I t can also facilitate growth through productivity improvements and new investments. Needless to add, the mobilization o f extemal grants can reduce the overall need for borrowing. Reductions in the overall cost of government borrowing, both domestic and external, through better debt management. Structural economic reforms that promote growth, exports, and government revenues through increases in investment and productivity as well as governance improvements. 20 Box 2.1: Refining the Debt Strategy The government could increase the credibility and operational relevance o f its debt strategy with the following refinements: 1. Because the overall time frame for working out the excessive debt burden i s long, Pakistan should define medium-term debt burden goals that can become an integral part o f a budgetary framework. 2. Public debt burden goals should b e defined in terms o f b o t h G D P and govemment revenues. In a sense revenues are a m u c h better indicator o f government’s ability t o service debt than GDP. In any case it i s the public debt-to-revenue ratio that i s exceptionally high in Pakistan and needs t o be brought down sharply. 3. The government should set explicit medium-term targets for levels o f external debt in relation t o exports/foreign exchange earnings so that it can achieve the H I P C initiative criterion o f reducing the present value o f external debt to exports o f goods and nonfactor services to 150 percent (from the present level o f about 200 percent, assuming a grant element in external debt o f about 20 percent). This step will strengthen the case for continued large concessionary assistance and external grants. 4. Related t o the previous recommendation, the external borrowing strategy for the next few years needs t o be made explicit and linked to the fiscal deficit financing requirements so that a clear domestic borrowing strategy can also emerge. Optimal mobilization o f concessionary assistance will greatly alleviate the burden o f domestic financing and, combined with financial discipline, could lead t o declining real levels o f outstanding public domestic debt in the not-too-distant future. 5. The D P C O should give a high priority to developing estimates o f the present value o f both external and public debt. The calculation o f present value for public debt is n o t common. However, because more than 50 percent o f Pakistan’s public debt i s external and a large part o f it i s concessionary, present-value calculations assume special importance. Estimates o f present value will help Pakistan set more appropriate debt burden reduction goals and better enable it t o monitor progress toward them. 6. The government should survey the exposure to exchange-rate and interest-rate risk o n i t s external loan portfolio and, if necessary, restructure i t s currency composition and possibly use market-based hedging instruments to manage these risks. The currency composition o f the external debt should reflect as far as possible the currency composition o f external trade and official international reserves. In addition, market-based risk management such as entering into currency swaps and interest-rate swaps at the time o f borrowing can hedge against exchange and interest-rate fluctuations at a relatively l o w cost. Furthermore, the government could minimize the risks o f commodity price fluctuations by employing commodity swaps, put options, or commodity- indexed loans. 7. Each year the D P C O should reconcile the final fiscal deficit numbers with the public debt stock changes n order to ensure accuracy and transparency.’ In the past, debt growth has been consistently larger than reported i fiscal deficits. ’/ For instance, whereas the cumulative recorded domestic borrowing b y the consolidated govemment to finance fiscal deficits from FY99 to FY02 was Rs. 827 billion, the increase in the govemment’s domestic debt (excluding SBP) in the same period was Rs. 1,054 billion. The difference arises from the govemment’s taking on liabilities o f other public sector entities or other operations outside the budget. 21 I Box 2.2: Determinants o f the Rate o f Public Debt Accumulation The analytical framework derives the following conclusions about the evolution o f the public debt burden. It shows (equation (2)) that the public debt burden (measured as a percentage of GDP) will decline if: 1 The borrowing need i s lower. The lower the expression in the f i r s t bracket (the sum o f the primary fiscal deficit and contingent liabilities p a i d by the government minus the fiiancing f r o m privatization receipts and foreign grants), the lower i s the borrowing need and the accumulation o f new debt. In particular, when the sum o f the primary fiscal deficit and payments o f contingent liabilities are lower than privatization receipts and foreign grants, the debt burden (debt/GDP ratio) declines, other things being equal. 1 Domestic and foreign real interest rates o n public debt are lower. 1 Pakistan's r e a l GDP growth rate i s higher. 1 The Pakistani rupee appreciates in real terms. When the nominal exchange rate measured in local currency units per foreign currency unit appreciates, the local currency value o f foreign currency debt and the local currency cost of servicing this debt declines. In the case o f nominal depreciation, the exchange losses on the external debt can be substantial, since the local currency value o f the whole stock o f foreign debt i s revalued. If the nominal depreciation goes along with a real depreciation, the rise in the G D P deflator does not fully offset the valuation loss f r o m the nominal depreciation and overall the debt/GDP ratio rises. In a stable real exchange rate environment thls last factor in the equation i s zero. (Of course, the exchange rate should only appreciate in real terms when this i s the long-run equilibrium real exchange rate matched by domestic productivity gains. If not, the loss o f competitiveness f r o m an overvalued exchange rate would lead to M e r accumulation o f extemal debt.) The time path for the evolution o f public debt i s given by the relationship: Dt+l = DDOMt+i+ DFOR$t+IERt+i DDOMt + DFOR$tERt + (PDt+i - PRIVZNtii + CL+i - GRANTSt+i) + DDOMt(P+ nD)+ DF0R$,ERt(rF + nF)+ DFOR$tERte'rt+l + DFOR$tERte'rt+l(rF + nF) (1) where Dt = total public debt at the end o f period t, DDOM = domestic public debt, DFOR$ =foreign public debt measured in units o f foreign currency, ER = the nominal exchange rate, local currency per unit o f foreign currency, PD= primary fiscal deficit PRIVzN= fiscal revenue f r o m privatization, CL= contingent liabilities being assumed by the government, G R A N T S = foreign grants, rD = domestic real interest rate, rF:foreign real interest rate, nD = domestic inflation rate, nF:foreign inflation rate, e*rt= percent change in nominal exchange rate f r o m one period t o next (+ shows depreciation), re'rt = percent change in real exchange rate f i o m one period to next (+ shows depreciation) Dividing throughout by GDPt+l the following relationship i s derived for the change in the debt/GDP ratio (lower case shows a l l variables in percent o f GDP): ddt + clt+l = (pd,+l - pri~m,,~ - - ddonq(g+nDY(l+g+nD) - dfort(g+nDY(l+g+n? + ddonq(P+n%l+g+nD) + dfort(?+nFy(l+g+nD) + dfor,e*r,,{( l+g+nD) + dfor,e'rt+l(?+nF y(l+g+n? =( p d,, - privznt+l+ c1,+] - grantqtl) + ddom,[- g - nD+ P+ny/(l+g+n? + dfort)[- g - nD+ ?+nF+ e'r, + e'rt+l(?+nF)] /(l+g+nD) = (pd,+, - privzn,+l + cl,+] - + ddom,(P - g)/(l+g+nD) + dfor,[(?- g) + (nF- nD+ e*rttl)+ e'rt+l(?+nF)]/(l+g+n? = (pdt+l - privzn,+l + - grants,+l) + ddom,(? - g)/(l+gd) + dfor,[(p- g) + e * r t + l ( p d ) + re*r,+l ]/(l+g-he) (2) *This result (2) is described in detail in the bullet points above. 22 Baseline Scenario 2.1 1. The government’s reform program stresses all three policy elements mentioned. The question i s whether these policies will be implemented in a timely fashion to create fiscal space that will allow the public sector to play i t s role in pursuit of the goals o f growth revival, poverty alleviation, and social development. Tables 2.2 and 2.3 summarize the assumptions and findings o f a baseline scenario that explores this question within the context of a medium-term macroeconomic framework that integrates real sector, fiscal accounts, and balance of payments projections. This section provides a brief overview o f the key points. A more detailed description and technical material are provided in Annex B. Table 2.2: Summary of key macroeconomic variables in baseline scenario, 1999/00-2006/07 Est. Prov. Proj. 1999/00 2000/01 2001/02 2002/03 2003/04 2004105 2005/06 2006/07 Real GDP at market prices 4.2 2.6 2.8 5.8 5.3 5.5 5.7 5.9 Consumer prices (p.a.) 3.6 4.4 2.7 3.1 4.0 4.0 4.0 4.0 Savings and investment Gross national savings 14.1 13.6 17.4 21.3 16.7 18.0 18.3 18.5 Gross fixed capital formation 14.4 13.9 13.1 13.1 14.5 15.5 16.1 16.7 Change in stocks (nongovt) 1.6 1.6 1.6 2.4 2.0 1.6 1.6 1.6 Public finances Revenue (including grants) 17.3 17.3 19.3 20.8 18.5 19.3 19.6 19.8 Grants 1.1 1.2 2.3 2.8 1.1 1.7 1.6 1.5 Revenue 16.3 16.2 17.0 17.9 17.3 17.6 18.0 18.3 Expenditure 22.8 21.4 23.6 22.3 21.8 22.4 22.4 22.6 Interest payments 7.8 6.8 6.8 5.2 4.8 4.6 4.2 3.9 Defense 4.8 3.8 4.1 4.0 3.6 4.0 4.0 4.0 Payment of contingent liabilities 0.5 0.6 1.9 1.2 1.9 1.3 0.8 0.5 Non-interest, non-defense, non-SOE spending 9.8 10.2 10.9 12.0 11.6 12.6 13.4 14.3 Overall balance (including grants) -5.5 -4.1 -4.4 -1.6 -3.4 -3.1 -2.8 -2.8 Primary balance (including grants) 2.3 2.8 3.8 3.6 1.4 1.5 1.4 1.o Gross public debt 100.9 108.3 100.1 91.8 83.3 80.2 74.2 70.9 Gross public debt in % o f revenue 619.6 670.5 588.9 512.0 480.6 456.3 411.40 387.3 Extemal sector Current account incl. off. transfers -2.0 -1.9 2.7 5.9 0.2 0.9 0.6 0.2 Ext. public and publicly-guar. debt 48.6 55.7 52.7 45.8 38.6 35.8 32.8 30.4 Gross reserves in months o f next year imports of 3.8 1.7 3.7 7.0 6.3 6.0 6.2 6.2 goods and services Table 2.3: Base Case Scenario: Principal Budget Aggregates Change FY2002 - FY2007 (Percent o f GDP) Resources Expenditures Govemment revenues + 1.3 Defense expenditure - 0.1 Extemal grants - 0.8 Interest payments - 2.9 Fiscal deficits (including grants) - 1.5 Public enterprise losses 1/ - 1.4 Non-interest non-defense non-SOE + 3.4 expenditures o/w Development spending 1.5 Total resources - 1.O Total expenditures - 1.0 1/ Public enterprise losses are approximated by subsidies and payment of contingent liabilities. 23 2.12. The analysis suggests that to achieve the desired public debt reduction goal by FY07, the government should bring down the overall fiscal deficit (including external grants, provision for public enterprise losses, and other contingency payments) to an average o f 3 percent o f GDP during FYs 2004- 07, compared with the average o f 4.3 percent during FYs 2001 - 02 (see Table 2.2). In light o f the strong fiscal performance in FY03 when the fiscal deficit (excluding grants) declined by 2 percent o f GDP the fiscal adjustment required over the medium t e r m i s not large. But the scenario suggests n o room for expanding the overall level o f public expenditures, which i s projected at 22.3 percent o f GDP on average, the same level as in FY03. T h e projected modest growth in revenues i s partly offset by the reduction in external grants (which were at an exceptionally high level in FYs 02 and 03) and partly absorbed by the need to reduce the deficit. 2.13. The projected level o f external grants assumed for FYs 2004-07 (1.5 percent o f GDP) i s lower than in the last two years but somewhat higher than in earlier years. Provided reform efforts continue, there would be a strong case for the donor community to give a high level o f budgetary grant support to Pakistan because o f i t s heavy public debt burden and less favorable social indicators.” But donors will also expect a restructuring o f expenditure toward economic and social sectors and generally greater effectiveness in public spending. 2.14. Our projections show that if defense spending can be contained and losses o f public enterprises and the interest payments can be brought down, non-interest, non-defense and non-SOE spending will rise to 14.3 percent in FY07, compared with 10.9 percent in FY02 (and an estimated 12.0 percent in FY03; see Table 2.2). T h e policy issues and options involved in bringing about a shift in public spending toward development and social spending are discussed at some length below. Slowing the Rate o f Borrowing 2.15. The first step in slowing down government borrowing i s t o continue to run significant primary surpluses. T h i s means that government’s own resources and not fresh borrowing should cover at least a part of the interest payments. Until relatively recently, Palustan actually ran large primary deficits, which meant that all o f the borrowing costs (and off-budget borrowing) were financed f r o m fresh borrowing. During the past few years Pakistan has made clear progress; the primary surplus (excluding grants) averaged over 1 percent o f GDP during 1998/99 - 2002/03. Under the Debt Committee projections and the PRGF program with the IMF, the primary surplus was expected t o rise to well over 2 percent o f GDP by FY 2004. This level n o longer appears either realistic or necessary because our projected fiscal path shows sufficient progress toward debt sustainability with a somewhat less restrictive fiscal stance. Pressures for expanding non-interest public spending from the very l o w level reached in F Y OO have grown following the finalization o f the I-PRSP, the continued sluggish private sector, and hence slow j o b creation. 2.16. Nonetheless, the government will need to weigh some choices. Clearly, it will be desirable to maintain a significant primary surplus (including grants) in FYs 2004-07 in order to underpin continued progress toward debt reduction. Yet genuine political and economic pressures for development and social spending are mounting. Further growth in tax revenues as a percentage o f GDP, reduction in public enterprise losses, and acceleration o f privatization receipts can help t o resolve the government’s dilemmas and minimize the short-term conflict among key economic and social goals. T h e external grants projections also reflect the recent announcement by the U S government that i t will provide support amounting to U S $ 3 billion over the next 5 years. 24 Increasing Revenues 2.17. Pakistan's revenue mobilization has been low, both by international standards and compared with i t s potential. In FYs 2000 and 2001 total revenue collection was the equivalent o f 16.2 percent o f GDP; tax revenue was only 13 percent o f GDP and had actually stagnated for several years.20However, revenue collections during the last two years increased sharply to 17.9 percent o f GDP in 2002-03 reflecting both the success o f tax collection efforts and the revival o f the economy. The structure o f taxation has also improved considerably in recent years, and the heavy dependence o n foreign trade taxes has been reduced. Still, revenue mobilization falls short o f that o f many developing countries at comparable income levels in Afi-ica, Asia, and Latin America.*l In light o f the weaknesses in the current revenue mobilization system, there i s considerable scope for enhancing collection and broadening the income tax base. For instance, while there are only 1-1.5 million income tax filers, there are 3 million cell phone subscribers, over 10 million electricity consumers, 3-4 million gas consumers and 2-3 million car owners. 2.18. The main factors behind the l o w revenue mobilization are governance problems in tax administration, a narrow and inflexible tax base, a large informal economy that escapes the tax net, and pervasive smuggling with associated revenue losses. Governance problems have been reflected in widespread collusion (corruption) between taxpayers and tax officials, a situation that leads to tax evasion and lack o f tax compliance. These problems stem f r o m the discretionary powers o f tax officials, complex tax rules, and weak supervision o f staff. 2.19. T o address these problems, the government took a number o f actions beginning in FY02. (i) n the area o f tax administration, the government has begun a fundamental restructuring o f the I Central Board o f Revenue (CBR). This restructuring includes a complete reorganization over a three-year period (FYs 02-04) and fundamental changes in human resource management. A Large Taxpayers Unit in Karachi has started operations, and a pilot income tax office for medium-sized taxpayers was recently opened in Lahore. Both units operate with greater functional specialization in tax administration. Universal self-assessment for income tax and audits to reduce contact with tax officials has also been introduced. The Government and the W o r l d Bank are preparing a major technical assistance program for tax administration reforms. (ii) Tax policies have also been strengthened: the coverage o f the general sales tax has been expanded to include agricultural inputs and n o w requires compliance o f small traders. A new income tax policy came into effect in FY 03 with a number o f reform measures (minimal tax exemptions, simplified and more equitable tax rates). The provinces have started collections o f the agricultural income tax for high-income farmers that were instituted in 2000. 2.20. In our view, it should not be difficult, in light o f current efforts, to increase revenue mobilization further to 18.3 percent o f GDP by FY07, the assumption in the base case. But the suggested level o f revenue mobilization will merely regain the level o f the early 1990s. 2.21. T o consolidate recent revenue gains and t o underpin future growth o f the economy, determined efforts are necessary to accelerate institutional and policy reform, especially the fundamental restructuring of C B R in the next few years. Pervasive smuggling must be addressed resolutely. I t i s a major cause o f loss o f revenue and foreign exchange. I n the area o f tax policy, reforms should aim at continuing t o ~~ 2o Note that Pakistan has undertaken considerable trade liberalization in recent years with an initial import tariff revenue loss. *'Countries at comparable income levels with higher revenue/GDP ratios include Bhutan, Botswana, Kenya, Maldives, Namibia, Nicaragua, S r i Lanka, Vietnam, and Zambia. All o f these countries had revenue mobilization in excess o f 20 percent o f GDP, with the exception o f Sri Lanka and Zambia. 25 reduce the number of tax exemptions (such as for customs duties and income and withholding tax) and make sales taxation more effective. Reforms are also needed for tax on income f r o m financial and land n particular the current practice that allows a single person to hold assets under different names assets. I should be changed, since it weakens tax collection on income from such assets. On the nontax revenue side i t will be necessary to seek greater cost recovery in higher education, health (subject to considerations about access to basic services for the very poor), and water usage for irrigation. 2.22. I n addition, the government should undertake administrative and governance reforms to ensure better documentation o f economic transactions, ownership o f assets, and sources o f income, by, among other things, expanding efforts to ensure auditing o f more businesses. Better documentation will enable the government to implement existing tax laws more effectively as well as to broaden the tax base in the future. 2.23. Sustained increase in fiscal revenue mobilization nationwide requires stronger revenue effort not only by the federal government but also by provincial governments. Provincial resource mobilization has dropped steadily in recent years and i s currently only 0.5 percent o f GDP. Provincial o w n revenue collection would need to increase in line with the expansion in sub-national spending foreseen in this report (in particular on education, health and certain types o f infrastructure) to reach the Government’s goals for poverty reduction and t o be consistent with the ongoing devolution process. However, provincial (and local) governments only have tax authority over taxes which for the most part are characterized by narrow, inelastic bases and l o w buoyancy. In addition there has been poor utilization o f provincial and local taxes and user charges, reflecting both lack o f political will to collect key provincial taxes and deficient revenue administration marred by governance problems. 2.24. Increased provincial revenue mobilization hence requires reforms both in revenue policy and revenue administration. Revenue policy reforms are needed to expand the provincial revenue base. This would require a review o f tax authority powers o f the provincial as well as local governments. Reforms could possibly include allowing the provinces t o share selected federal taxes through levying “add-ons” o n the federal taxes with federal collection; determinedly expand and enforce collection o f the agricultural income tax; and increase reliance o n user charges wherever appropriate and feasible. Furthermore, incentives for sub-national level resource mobilization need to be strengthened as part o f larger [discretionary] resource transfers t o the provincial and local governments that appear necessary. Such transfers can be linked to provincial revenue efforts and demonstration o f capacity to effectively utilize resources. Significant revenue gains could also be realized by strengthening the provincial tax and non tax administrations and plug leakages of revenue; by merging the two revenue collecting departments into one; by better liaising with the federal Central Board o f Revenue; through improved automation and computerization in key areas of provincial taxes (e.g., land and property records); and through enhanced training o f and incentives for tax officials. Public Enterprise Losses and Other Contingent Liabilities 2.25. There i s not a sufficiently widespread appreciation by the public and Pakistani policymakers o f the extent to which very large public enterprise losses have contributed directly or indirectly t o the buildup o f public debt andor contingent liabilities. According to our estimates, cumulative public enterprise losses and the cost of government guarantees through the end o f FY03 amounted to close to Rs. 400 billion, or about 10 percent o f GDP22 (see Table 2.4 and B o x 2.3). These losses initially remained partly unfunded and took the form of arrears buildup or borrowing from state-owned banks (that cannot be serviced), but the losses must eventually be reflected in the fiscal budget and hence add to public debt. This estimate i s based on the data contained in the Debt Committee Report up to FYOO and actual losses o f PIA in FYOl and 22 WAPDA and KESC in FYs 01 - 03. 26 n some cases, such as the equity injections by the government into WAPDA and KESC, the losses do not I enter the budget directly. Nonetheless, they affect the level o f debt by either reducing govemment receipts or requiring special off-budget government borrowing, such as policy assistance from ADB to facilitate clearance o f K E S C losses. 2.26. Though Palustan has made progress during the past three years (FYs 01 - 03) in eliminating the operating losses o f a number of public enterprises, notably Palustan Railways, Palustan International Airlines, and most state-owned banks, the losses in the power sector continue at a high level. Indeed, WAPDA and K E S C losses in FY03 (accrual balance) were close to 1 percent o f GDP (Rs 33 billion) whereas total budgetary support to these companies was Rs 73 billion or 1.8 percent of GDP. The F Y 0 4 budget foresees budgetary support to WAPDA and K E S C o f 1.3 percent o f GDP. The continued large losses o f public-owned power entities i s the most troubling element o f Pakistan’s fiscal position and the only one that has not shown improvement in recent years. 2.27. Many factors account for the continued large losses o f WAPDA and KESC: operational inefficiencies, weak bill collection, very high levels o f transmission and distribution losses (about 25 percent o f units generated or purchased in the case o f WAPDA and 40 percent in the case o f KESC) due to theft and leakage, inadequate tariff adjustments to increased fuel prices, the relatively high cost o f purchases from IPPs, including payments for large unused generation capacity, and the politically mandated practice o f subsidizing household and farm consumption at the expense o f industry and commerce. Corruption, concerns about law and order, and lack o f development funds have also contributed to the decline in KESC’s financial performance. In addition, WAPDA’s finances have been suffering from the Federally Administered Tribal Area’s (FATA) failure to pay utility bills amounting to Rs. 12 billion annually, part o f which may be overbilling by WAPDA. The coming on stream o f the large Ghazi Baroda Hydel Project (1,450 MW) and the substantial decline in payments to PPs after F Y 0 4 will, however, help to reduce the average generation cost. A pick-up in economic activity by increasing power demand will also reduce the burden o f large, relatively fixed payments to PPs. 27 Box 2.3: Contingent Liabilities The federal and provincial governments could face serious fiscal costs as a result o f their contingent liabilities. These are fiscal obligations contingent on the occurrence o f particular events and are not adequately budgeted and accounted for in the fiscal budget. Pakistan has made some progress in recent years on contingent explicit liabilities, at least on monitoring. These liabilities are specific obligations, created by law or contract that the government must settle if particular events occur. They include government guarantees to autonomous bodies, public-sector banks, and the private sector. The govemment i s tracking the issuance and stock of explicit govemment guarantees, and the Draft Fiscal Sustainability and Debt Limitation Law limits the annual issuance of such guarantees to 2 percent o f GDP. Over the past decade, however, annual issuance of such guarantees averaged only 0.6-0.7 percent of GDP. In FY 2002 the impact o f contingent explicit liabilities on the budget was Rs. 24.4 billion, or 0.7 percent o f GDP, and in FY03 it was Rs. 16.2 billion, or 0.4 percent of GDP. Contingent implicit liabilities have remained substantial. These are moral obligations or burdens that, although not legally binding, are likely to be bome by the govemment because of public expectations or political pressure. In recent years public enterprises have run substantial operating losses. The impact o f contingent implicit liabilities on the federal budget (and on public debt) stemming from public enterprises was Rs. 126.5 billion in FY02 (3.4 percent of GDP, including a major equity injection to KESC) and Rs. 84.5 billion in FY03 (2.1 percent of GDP). Whereas other public enterprises have improved financially in recent years, the power sector enterprises-WAPDA and KESC-remain major loss makers. In FY03 the drain on the budget from these two enterprises amounted to 1.8 percent of GDP’. Without fundamental reform they will continue to absorb massive fiscal resources. Vigilance i s also needed on other contingent liabilities, however, including nonperfoming loans of the public-sectorbanks. Direct implicit liabilities are expected to increase over the medium term. These include the unfunded part of public pension liabilities, which are expected to rise considerably. Pension liabilities must be estimated and made transparent so that proper measures can be taken in time to avoid a pension time bomb. The treatment o f interest costs on N S S instruments leads to the buildup of a direct implicit liability to be paid when NSS instruments are redeemed. In the medium term a certain lumpiness in redemption and interest payments i s expected, materializing as a sudden large claim on the fiscal budget. The government can limit the impact of contingent and direct implicit liabilities by reducing the potential source of the problem. Privatization and restructuring of the remaining public enterprises to reduce losses could greatly reduce the pressure on the govemment to assume contingent liabilities. In addition, guarantees must be given much more restrictively, subject to a well-fimctioning tracking, monitoring, and valuation system. ’ To put this in perspective, this i s the same as the total education budget (also 1.8 percent o f GDP) and more than double the total health budget (0.7 percent o f GDP). 2.28. The impact o f the financial problems in the power sector companies extends beyond the fiscal budget. These companies have under invested in power generation capacity for some time, and the country’s investment needs in the power sector cannot be met without private sector participation. In addition to the direct consequences for federal and provincial govemment finances, the power sector problems have a major impact on private sector competitiveness and hence economic growth, exports, and employment (See the forthcoming W o r l d Bank study “Investment Climate Assessment in Pakistan”). n fact, the costs associated with the unreliable power supply are one o f the manufacturing sector’s top I competitiveness concerns. 2.29. The govemment has adopted a strategy to restructure public enterprises to improve their operational efficiency and privatize most o f them. In the power sector-the source o f the largest contingent liabilities and enterprise losses-the government has started implementing a strategy that aims at improving the efficiency and the financial viability o f the enterprises through competition, accountability, and managerial autonomy, while also establishing a multiyear tariff framework and improving the regulatory framework. Efficiency gains will be achieved by unbundling and corporatizing WAPDA’s power wing into independent distribution and thermal generation companies as well as a 28 national transmission company. The goal i s to privatize the distribution and generation companies and keep the transmission company in government ownership. 2.30. As will be discussed later, although the need to privatize KESC and parts o f WAPDA has become more urgent because o f continued losses and inadequate resources for investment as well as the costs to other industries, the progress toward privatization i s the slowest in the power sector. In part, this slow progress i s due to factors beyond the government’s control, such as the lack o f foreign interest in KESC. 2.3 1, With regard to the pricing framework, NEPRA has recently followed-albeit with some delay- the policy o f adjusting tariffs for changes in fuel prices to achieve financial viability. O n occasion WAPDA has also delayed applying for and implementing tariff increases. 2.32. I t i s n o w clear that in FY03 public enterprise losses were higher than in FY02 because o f the lack o f financial improvement by the power companies and will remain so in FY04. These higher losses compared to previous projections will limit the projected expansion in fiscal space over the next two years. Beyond FY 2004, i t should be possible to reduce public enterprise losses through a combination o f reforms and structural improvements in the power sector, lower IPP payments and increasing share o f l o w cost hydro- power. In our view, with reasonable strong policy actions, i t should be possible to reduce overall public enterprise losses to 0.5 percent o f GDP by FY07 from the estimated peak o f 1.9 percent o f GDP in FY 2004. 2.33. Looking ahead at the remaining reform agenda, WAPDA needs to decisively reduce line losses, enforce bill collection, if necessary by cutting o f f nonpayers more aggressively, and ensure timely implementation o f tariff adjustments determined by NEPRA. The government should help WAPDA in resolving the issue o f FATA’s nonpayment o f i t s utility bills since this problem seems to require a political solution. The transmission company, remaining in government ownership, i s likely to run modest operating losses even at optimum technical/operational efficiency, because o f the govemment’s pursuit o f noncommercial objectives (for instance, rural electrification, which has a development and social objective). The losses should be minimized and limited to the cost o f pursuing noncommercial objectives as imposed by government mandate. Over the medium term the losses should be fully financed by grants from the government to cover the cost o f pursuing the social objectives, and the cost should be made explicit in the fiscal budget. 2.34. With privatization, gains in operating efficiency, reduction in theft, and expected lower generation costs because o f reduced payments to IPPs, the large Ghazi Barota project coming on-stream and hence an increase in the share o f hydro power, it should realistically be possible to reduce the budgetary support to WAPDA and K E S C from about Rs. 73 billion, or 1.8 percent o f GDP in FY03, t o Rs. 13 billion over the medium t e r m (by FY07). The government needs to ensure, however, that other public enterprises such as Pakistan Railways, Palustan Steel, PIA, and state-owned banks do not fall back into deficit and that guarantees provided by government do not have a net cost to the budget (see B o x 2.3). Meeting this goal will require the government to continue privatizing most commercial and industrial public enterprises, to restructure remaining public enterprises and subject them to hard budget constraints, and to restrain tendencies to set prices administratively in the public sector to protect the consumer. 29 Table 2.4: Losses of State Owned Enterprises (Rs Million) Losses Carry Losses During Cumulative Forward upto Losses upto 1998199 1999100 2000/01 2001/02 2001/02 KESC 24,399 12,787 16,203 17,159 70,548 WAPDA 11 36,400 8,527 15,086 33,128 93,141 Pakistan Steel Mills 8,123 1,403 1,065 1,280 11,871 Pakistan Railways 25,721 2,197 727 -1,819 26,826 PIA 7,046 5,242 1,982 -2,156 12,114 Banks 11 30,700 8,000 0 7.900 46,600 Others 21 60,439 NA NA NA 60,439 Total SOEs loss (Debt Report) 192,828 38,156 35,063 54,492 321,539 1/ Carry forward losses are approximated by GOP equity injection into the enterprise and may therefore overstate the accumulated losses o f SOEs. 21 Include: Pakistan National Shipping Corporation (PNSC), Karachi Steel and Engineering Works (KS&EW), Utility Stores Corporation (USC), Heavy Mechanical Complex (HMC), State Engineering Corporation (SEC), Heavy Electrical Complex (HEC), Pakistan Broadcasting Corporation (PBC) and Pakistan Tourism Development Corporation (PTDC). Also included are the losses (of about Rs 50 billion) of Pakistan (TCP), Ghee Corporation o f Pakistan (GCP), Cotton Export Corporation (CEC), Rice Export Corporation of Pakistan (RECP) and the Saindak Copper and Gold Mining Unit, which were covered through issuance o f govemment guaranteed bonds (as almost all of these entities have been closed down, this debt i s serviced by the federal govemment). Sources: The figures for carry forward losses of KESC, Pakistan Steel Mills, PIA and Others SOEs are taken from the Govemment o f Pakistan's report A Debt Burden Reduction and Management Strategy, Finance Division, March 2001. Losses o f SOEs (other than Banks) during 1999/2000-2001/2002 are takedcompiled from the IMF's report Selected Issues and Statistical Appendix, October 21, 2002. 2.35. According to our estimates, even with continued forceful reform o f the public enterprise sector, the losses o f public enterprises (mainly power entities) are likely to total Rs.150 billion during FYs 04- 07, higher than the estimate in the Debt Committee Report (Rs. 100 billion during FYs 2000-04). The Debt Committee also assumed that the enterprise losses and contingent liabilities would be offset by privatization revenues and thus that there would be n o government borrowing beyond the normal budget deficit. This assumption appears optimistic; w e foresee net privatization revenues to the fiscal budget o f only Rs. 54 billion over the four years, FYs 04-07. Even after taking into account privatization revenues, enterprise losses are likely to be a drain o n the budget. This projection further underlines the need to vigorously pursue the overlapping agendas o f privatization and public enterprise reform. 2.36. With regard to contingent liabilities, the government should assess the full impact on public finances beyond the budget and fiscal debt to include these extra-budgetary items. I t should calculate the potential budget cost in expected value terms23 and make the potential fiscal cost o f off-budget items visible beforehand to ensure a proper cost-benefit assessment o f government support. I t should also limit moral hazard by announcing the limit o f i t s financial support and then sticking to that limit. Defense Spending 2.37. Historically, the high level o f defense spending has contributed to the buildup o f public debt in two ways. It has preempted public resources for development, and it has been a major factor in the government's inability to cover current expenditures from current revenues. Government savings (defined narrowly as the excess o f government revenues over current expenditure) deteriorated from a positive 0.9 percent o f GDP in 1980-85 to a negative 3.1 percent o f GDP in 1996-99. As the Debt Committee has *'T h e financial cost o f the event if i t occurs weighted by its probability. 30 stressed, there was “a serious upsetting of the balance between defense and development” after the early 1980~.I *n ~ the 1990s, however, real spending on defense increased little, and as a percentage o f GDP, it declined steadily from the peak of 6.9 percent in the second h a l f o f the 1980s to less than 5 percent during 1998-2000. Despite the spurt in spending in FY02 due to regional tensions, defense spending remained at 4.7 percent o f GDP (4.0 percent excluding defense pensions, which have been shown separately since FYO1). The steady relative reduction in defense spending in recent years has helped to contain the fiscal deficit, reduce the negative government saving (to less than 2 percent o f GDP in FY02), and t o a degree correct the imbalance between defense and development. 2.38. The government feels that in current circumstances assuming further reduction in defense spending as a percentage o f GDP would not be realistic. While easing o f tensions on the Eastern border i s very likely, the situation o n the Western border as well as combating terrorism in the country will continue to put pressure on the defense expenditures The report, therefore, assumes that in the base case scenario defense spending (excluding pensions) in F Y 0 4 would be as in the federal budget approved by National Assembly and in coming years would remain at 4 per cent o f GDP. T o the extent that the effectiveness o f defense spending can be improved without compromising national security, the government’s policy options will be expanded and the level o f defense expenditures will not be entirely hostage to regional geopolitical realities. The Cost of Borrowing 2.39. As the analytical framework in Box 2.2 indicates, the real cost o f borrowing i s a critical variable in determining the growth of the debt burden. I n Pakistan, because the primary balance (including payment o f contingent liabilities) has been in surplus for some time, the growth o f public debt in recent years has been driven almost entirely by the cost o f borrowing. Unfortunately, the real cost o f public debt rose to 4.5 percent annually during 1996-99, compared with the historical level and intemational n o r m o f 2.5 percent. Surprising though it may seem, real borrowing costs during 2000-02 were even higher, at 5.2 percent per year (see Table 2.5). The high real cost o f public debt since the mid-1990s reflects a jump in the cost o f both external and domestic borrowing.25 On the domestic side, the sharp deceleration in the rate o f inflation from more than 11 percent per year in the mid-1990s to less than 4 percent during the past three years, combined with the locked-in high interest rates on National Saving Schemes (NSS) instruments, kept the real annual cost of domestic debt at 7.4 percent (11.5 percent in nominal terms) during 1999/2000-2001/02. On the extemal side, the nominal interest rate o n external public debt i s about 2%-3 percent. The real net depreciation in the value o f the rupee over FYs 1999-2001 however, added to the real cost o f external debt, raising i t to 7.6 percent per year. Including the recent period o f real appreciation in FY 2002, the real cost o f extemal borrowing averaged 3.0 during FYs 1999-2002 and was substantially below the 4.7 percent annual figure for 1996-99.26 2.40. A significant reduction in the cost o f public debt must be a part o f the strategy to reduce the burden o f debt. The objective o f policy should be to sharply reduce the overall real cost o f public borrowing over the medium term. Unless the cost o f public debt i s brought down, there will be limited scope for sharply reducing the share of interest payments in total government spending, at present 30 percent. 24 Debt Committee Report, pp. 14-19. 25 The real cost o f domestic borrowing is the average nominal interest rate minus the rate o f inflation. In estimating the real cost of foreign debt, the capital loss on foreign exchange because o f the depreciation o f the rupee in relation to the U.S. dollar must be added to the nominal interest payments, and then this nominal cost must be adjusted for the rate o f domestic inflation. For details, see Debt Committee Report, pp. 18-19. 26 T h e estimate reflects the high rate o f real depreciation o f the rupee during the period. 31 Table 2.5. Pakistan: Real borrowing cost on public debt, FY1996 - FY2002 Real borrowing cost (%) ” FY96-99 FY00-02 Domestic public debt 4.3 7.4 External public debt 4.7 3.0 Weighted average 4.5 5.2 1/ Compound rates Domestic Debt Costs 2.41. The most immediate problem i s the high real interest rate on domestic debt. The average real interest rate peaked in FY OO at 10.8 percent and declined to 5.6 percent in FY03. In the past, the high level o f domestic interest rates reflected the upward pressure placed on the country’s narrow financial markets by the considerable borrowing needs o f the government. N o w that the n e t government borrowings are projected to be modest (see Table A.2 in Annex B), the government has much greater degrees o f freedom in the management o f i t s domestic debt. 2.42. A central problem i s the s t i l l high cost o f borrowing under the National Saving Schemes (NSS), notwithstanding recent rate reductions. Interest rates on the N S S instruments such as the ten-year Defence Savings Certificates (DSCs), the five-year Regular Income Certificates (RICs), and the three-year Special Savings Certificates (SSCs) are fixed administratively. Holders o f NSS instruments can redeem them at any time, encashing both the face value o f the bond and the accumulated interest earnings. On the initial assumption that NSS serves a population o f small, relatively unsophisticated savers (such as widows, orphans, and pensioners), who are unlikely to place their funds with banks or other financial institutions, rates o f return on these instruments tended to be generous and tax free. I n the 1980s, as budget deficits grew, reliance on NSS instruments increased, and these certificates were made extremely attractive through highly positive tax-free real returns. Naturally this attracted major institutional investors. During the period from late 1996 to mid-1999, compound rates o n DSCs o f ten years’ maturity peaked at 18.04 percent when held to maturity. The yields on all NSS instruments were tax free except that Zakat was payable. Since the interest i s paid only o n encashment, the large accumulated interest liability i s not reflected in the reported stock o f NSS debt. This situation implies two things: an understatement o f government debt obligations and the prospect that a sizable part o f the govemment’s interest bill will continue to be driven by the high-cost certificates issued during 1997-99. 2.43. The government has taken several strong measures in recent years to address some o f the weaknesses o f NSS instruments. It eliminated the access o f institutional investors to the schemes in early 2000. Recently it took commercial banks out because o f abuses o f the system. It has made a number o f successive reductions in the administratively set interest rates o n new issues since M a y 1999 and has linked the NSS rates to the market benchmark Pakistan Investment Bonds (PIBs), which have been revived as an important instrument o f debt management. Furthermore, the income from NSS on deposits larger than Rs. 150,000 was made taxable effective July 1, 2001, though this change applies only to n e w issues. In the most recent rate reduction in July 2003, the rate o n DSCs with ten-year maturity was reduced to 8.5 percent per year. The rates on 3- and 5-year instruments remain at 7.7 percent. 2.44. However, two fundamental problems remain. First, despite the actions taken in recent years, NSS i s s t i l l a source o f high-cost borrowing for the government. T h i s i s because the sharp drop in the rate o f inflation has greatly offset the impact o f nominal interest rate reductions for NSS o n the real rates. Second, the on-tap nature o f the schemes limits the govemment’s options to explore optimal mix o f 32 borrowing. Also, the way the N S S instruments are administered makes i t difficult for the government to accurately plan the interest payments in the coming budget years. 2.45. Despite the sharp reductions made in nominal rates on N S S instruments during FYs 2000-03 real long-term rates o f retum on DSCs o f 4 ?4percent (assuming an annual inflation rate o f 4 percent) and on the recently introduced 10-year certificates for widows and pensioners o f 6 percent o n what are risk-free assets are s t i l l high by international standards. T h e spread between the yield on PIBs and comparable NSS instruments should be eliminated. This spread i s currently 2.9 - 4.5 percent in the case o f 10-year instruments, 3.5 percent for 5-year instruments and 4.6 percent for 3-year instruments. Also, the rates on SSCs and RICs were not adjusted downward as much as the D S C rates during the past two years, and thus there was internal shift away from DSCs. The continued attractiveness o f N S S instruments can be judged from the fact that their share in total domestic debt continues to rise, increasing from 37 to 45 percent from FYOl to FY03. Indeed, over the past two years the expansion in the stock o f NSS instruments was more than double the increase in overall public domestic debt. Box 2.4: Beneficiaries o f National Saving Schemes The after-tax return o n N S S instruments has persistently been generous compared with market rates, and hence this instrument has been an expensive source o f funding o f the government deficit. The stated rationale for paying high interest rates o n these instruments has been that they serve a population of small, relatively unsophisticated savers (such as widows, orphans, and pensioners), w h o are unlikely t o place their funds with banks or other financial institutions and w h o have l o w incomes f r o m other sources. Certainly, a large number o f pensioners, f r o m b o t h the government and the private sector, widows, and people belonging t o low-income groups depend o n profits f r o m these schemes. The bulk o f the funds invested in N S S instruments, however, come from a relatively smaller group with a relatively large average holding o f N S S instruments. Holdings o f this size are not congruent with the mentioned target group; in other words, most o f the holdings o f N S S instruments are with relatively well-off groups in society, and these groups are the ones that receive the largest share o f the indirect income support (subsidy) that the government provides through the NSS. By providing t h i s subsidy through the NSS instruments, the government i s diverting scarce public h d s f r o m high-priority programs with a m u c h more substantial development impact, such as poverty reduction programs, basic health or education services, o r improved infrastructure. I 2.46. Ideally it would be desirable to phase out the use o f NSS instruments and rather find government deficits domestically by selling PIBs in sufficiently small holdings to households through the existing network o f N S S outlets. However, phasing out the NSS instruments would meet with substantial public opposition. Also, the government feels strongly that i t needs t o balance i t s social obligations with economic imperatives. As mentioned above, some o f the recently launched schemes act as safety net to pensioners, widows and the small savers. Still, the Government must continue with progress o n reforming the N S S while educating the public o n the true opportunity cost and problems o f the instruments. A key objective s i s to provide the public with a safe alternative savings instrument with a good yield. In that regard the Government could make the PIBs more accessible to individuals. The Government i s also considering CDNS entering into mutual funds to facilitate developing a secondary market for such assets. 2.47. The successful launch o f the PIBs has not only set a market-based long-term benchmark yield, but also helped the government to finance i t s budgetary requirements with stable long-term debt.27 In any event, the government should make further use o f PIBs, end the on-tap feature o f the NSS instruments 27 State Bank ofPakistan Annual Report 2001-2002, p. 136. 33 and reduce the margins over PIB yields. The yield curve in Palustan has become very steep. The spread between the yields o f three-month treasury bills and ten-year bonds was 3.9 percent in July 2003 (currently the most recent observation o f the ten-year bonds yield). The over reliance on NSS clearly limits the government’s options for relying on the cheaper short end o f the market. In light o f the current l o w yields on treasury bills (less than 2 percent per year), the govemment could drive down the average borrowing cost by issuing treasury bills and retiring expensive foreign and-to the extent possible- domestic debt. 2.48. The government must do a better j o b o f educating the public on the high cost o f N S S and i t s implications for the national debt problem. I t can be made clear that the bulk o f public debt interest payments go to well-to-do Pakistanis and that unless these payments can be reduced, the domestic debt problem cannot be solved and poverty programs will suffer. Direct income support programs shown explicitly in the fiscal budget can address the policy objective o f providing income support to widows, orphans, and pensioners, if i t i s to be pursued. 2.49. The discussion here suggests that significant reductions in the real costs o f government domestic borrowing will be possible only with a lag. The cost o f domestic debt i s currently underestimated in the fiscal budget. The true cost i s the accrued interest liability. Since the stock o f N S S instruments has been rising steadily and the rates were also increasing until 1997, the cash-based figure for interest cost as reported in the fiscal accounts i s substantially lower than the accruals. In other words, every year the government has been building up an unfunded liability that will burden the budget in the future over and above the accrual cost. This situation reinforces the need to limit fi-esh government borrowing and to carry out further fiscal adjustment. 2.50. W e also recommend that the DPCO estimate the accrual cost o f the interest liability for the N S S instruments with a view to reflecting the true cost in the fiscal accounts. The accrued but not paid interest could be added to debt obligations. The calculations will also improve the estimates o f future interest payments likely to be actually paid. The Cost of External Borrowing 2.5 1. From a fiscal point o f view, the total real cost o f foreign borrowing i s the sum o f the nominal interest payments in domestic currency plus the capital loss or gain in the domestic value o f external debt stemming fi-om depreciation or appreciation o f the currency due to nominal exchange rate changes--all adjusted for the changes in the domestic price As already mentioned, the total real cost o f foreign borrowing rose during FYs 1996-99 to 4.7 percent annually, reflecting, among other things, substantial capital loss due to real depreciation o f the Palustani rupee in this period. The real depreciation continued during FYs 2000 and 2001 but has been sharply reversed since then because o f a remarkable tumaround in the current account balance o f payments and improved external resource availability, which i s reflected in a strong buildup in foreign exchange reserves. For the period FY 1999-2002 as a whole, the average real cost o f external borrowing for the budget came down to 3.9 percent, reflecting real appreciation in the last year and the effect o f debt rescheduling. In FY 2002 and the first h a l f o f FY 2003, the total real cost o f foreign debt was actually a negative 5.7 percent per year. This shift reflects an average annual real appreciation against the U.S. dollar o f 7.9 percent and the effect o f debt relief. 2.52. What assumptions should be made about public external debt costs in the medium term? H o w i s the exchange rate likely to behave in the near future? What should be Palustan’s external borrowing *’ If there i s a real depreciation the valuation element i s a capital loss, and if there i s a real appreciation the valuation element i s a capital gain. 34 strategy, keeping in view the high public debt burden on the one hand and substantial strength in the balance o f payments on the other hand? 2.53. I t i s difficult to predict the course o f exchange rate movements, but the sharp real appreciation in FY03 i s not something one can expect to continue over the medium and long term.” Still, in view o f the very strong foreign exchange reserve position, it seems unlikely that the rupee will depreciate in real terms in the near future. More precisely, we assume that the nominal changes in the exchange rate in this (FY04) and subsequent fiscal years will only reflect the differences between the domestic and international rates o f inflation. This situation suggests that the real cost o f external borrowing can be kept down to about 1 percent per year if the mix o f borrowing remains heavily tilted toward concessionary assistance, as i t has been in the past few years. 2.54. The case for a continued very soft blend o f external assistance i s linked to the high level o f external indebtedness. As mentioned, despite the sharp slowdown in the growth o f public and publicly guaranteed debt and the very favorable terms o f Paris Club re~cheduling,~’ the present value o f public and publicly guaranteed external debt to exports i s s t i l l well above the desirable cutoff suggested for H P C countries. Pakistan’s case for reducing relative dependence on market loans from multilateral banks such as the World Bank and the Asian Development Bank and expanding i t s access to their concessionary facilities-the International Development Association (IDA) and the Asian Development Fund (ADF)- remains strong. In fact, the W o r l d Bank has increased i t s IDA exposure in recent years, from US$3.9 billion in June 1999 to US$5.6 billion in June 2003. However, soft financing outstanding through the ADF as of December 2002 remained at the June 1999 level o f US$4.2 billion. 2.55. The prospective stability o f the real exchange rate and the wide gap that i s emerging between the real cost o f public domestic debt (projected at about 5 percent per year during 2003-07) and extemal debt also suggest the desirability o f greater reliance on external financing (as opposed to the most expensive domestic debt instruments), provided the mix o f external assistance remains heavily concessionary. If extemal borrowing o n appropriate terms can be arranged, it may be feasible to totally eliminate the growth in domestic debt in real terms in a few years and thus lower costs dramatically. 2.56. If the government continues i t s recent economic policy reforms and maintains a strong track record o f implementation, which gives it more credibility in the eyes o f i t s development partners, the country will have considerable opportunities for concessionary assistance and adjustment lending to restructure the present expensive debt. Continued strong reform policies will also facilitate substantial policy and adjustment lending from multilateral institutions. Palustan’s challenge will be to develop strong sector investment and poverty-oriented programs that donors can support through policy-based quick-disbursing assistance, even though balance o f payments support may be disappearing as a rationale for adjustment lending. Meeting this challenge would also require the government to increase the general effectiveness o f public spending and hence increase the emphasis o n sound public debt dynamics. 2.57. Related to the cost o f foreign borrowing are the issues o f choice o f currency denomination and the degree o f interest rate variability on foreign debt. At present, the government has only limited information on the currency composition o f i t s external debt and therefore does not manage this risk in any meaningful way. W e recommend that the DPCO give priority to improving the information base in this regard and also develop a strategy to reduce the currency and interest rate risks (variability) which could include use o f market-based hedging instruments. 29 Between September 2001 and December 2002 the real effective exchange rate appreciated by 4.5 percent and the bilateral US. dollarhpee real exchange rate by 12 percent. 30 See World Bank, Development Policy Review, April 2002, pp. 24-25. 35 Structural Reforms and Growth Revival 2.58. The revival o f economic growth i s critical not only for alleviating poverty, but also for strengthening the debt-canying capacity o f the economy. The debt burden i s determined not only by the absolute level o f debt, but also by the size o f the economy, the level o f government revenues, and exports. But as discussed, the high burden o f debt itself contributed to the slowdown in the average annual growth rate o f real GDP from more than 6 percent to less than 4 percent over the last four years. The government's strong economic reform program aims at exiting the debt trap by pursuing simultaneous progress on macroeconomic stability and structural reforms. The key elements o f structural reform aim at improving governance, the health and efficiency o f the financial system, and the climate for private investment. However, the progress made in pursuing the reform agenda has only with a considerable lag helped revive the economy: the average growth rate during FYs 00-02 was only 3.2 percent but picked up to 5.1 percent in FY03. 2.59. Several factors help explain the sluggish growth and the pick-up in growth in FY03. T h e agricultural growth rate, which averaged more than 4 percent in the 1990s, dropped to 1.6 percent during 1999-2002 mainly as a result o f drought but rose to 4.2 percent in FY03. Gross fixed capital formation has continued i t s downward drift; the rate o f fixed investment reached a l o w point o f 13.1 percent o f GDP in FY02 and FY03. A number o f factors continue to hamper industrial growth including structural problems in industry and exports, continued heavy reliance on cotton textiles, especially cotton yam, and a lack of diversification to growth areas in international trade in manufactured goods. Decades o f l o w investment in human capital are also takmg their toll in terms o f slow growth in productivity. In spite o f these challenges that s t i l l remain, manufacturing and exports experienced rapid growth in FY03. K e y factors were better access to the EU and U S markets and improved competitiveness o f the domestic textile ind~stry.~' 2.60. Some o f the structural problems, such as the l o w levels o f education and slulls and insufficient diversification o f manufactured exports, will take time to resolve. But the remarkable improvement in Pakistan's external resource position in the past three years has fundamentally transformed the prospects of higher investment without an excessive reliance on external flows. H o w will the strength o f the external position translate into accelerated growth? The answer l i e s in the dramatic turnaround in the national savings rate. During 1996-99 the national savings rate averaged little over 12 percent o f GDP as government current expenditures substantially exceeded revenues and capital flight continued because o f fears of continued depreciation and mounting public debt. The strong improvement in private remittances, the sharp reduction in the government revenue deficit facilitated in part by external official grants, and the reversal of capital flight pushed the national savings rate to a record high level o f more than 21 percent o f GDP in FY 2003. If confidence in the currency can be maintained through a further orderly reduction in both public and external debt burdens and an adequate level o f foreign exchange reserves, there i s every reason t o hope that the national savings rate can remain in the 17-18 percent o f GDP range. On the fiscal side, achieving further increases in government savings and total elimination o f the reliance on foreign grants to finance current expenditures are goals that appear within reach in a couple o f years. 2.61. During FYs 2001-03, a significant part o f the increase in national savings took the form o f addition to foreign reserves---partly involuntary because o f the sluggishness o f private sector investment. In the future the need for increases in reserves will be relatively modest and thus most o f the national savings should be available for gross capital formation. Palustan can also reasonably hope to supplement i t s national savings with net foreign capital. The position in FYs 2002 and 2003, when Palustan has had a Elimination of the incentive for under-invoicing of exports after the disappearance o f the premium on exchange rate in the kerb 31 market may also partly explain the strong growth in recorded exports. 36 sizable current account balance o f payments surplus both with and without official transfers, cannot be considered the equilibrium. With the further likely improvement in the extemal debt position, Pahstan can safely run current account balance o f payments deficits o f 1-2 percent o f GDP, financed with a mix of grants, net borrowing, and foreign private investment (see Annex B, Table A.3). Thus, at least from the point o f view o f resource availability, the stage i s set for sustained and strong increases in the level o f investment. An increase in the fixed investment rate from the l o w o f 13.1 percent o f GDP in FYs 2002 and 2003 to around 17 percent over the next four years need not be constrained by financial availability. While public investment should also increase, more o f the increase must come from the private sector (see B o x 2.5). 2.62. I t i s a puzzle why private investment has remained stagnant in the past few years o f reforms even though both foreign exchange availability and the liquidity position o f the banks have improved greatly. Possible explanations are the real interest rate level (until very recently), perceived uncertainty about the outlook for policy reforms in the run-up to elections in October 2002, and the new political environment. Although funding costs (nominal interest rates) have come down, real lending rates have remained high partly because o f very high intermediation costs and lower inflation in the last few years. Acceleration o f banking sector reform and further progress on privatization o f state-owned banks i s clearly indicated. Currently, perceived uncertainty about the continuation o f policy reforms in the new political environment may have led investors to continue with a wait-and-see position. Hence, it i s important that the government reaffirm i t s commitment to reforms that improve the investment climate. 37 I Box 2.5: Growth Revival and Poverty Reduction: Private Investment and Public Spending The focus o f this report i s o n Public Sector Management issues in Pakistan. However, the report has underlined at I various places the critical importance o f reviving private investment. The baseline scenario assumes that 7 0 percent o f the additional investment over 2002-07 would b e in the private sector. This assumes that with the continued privatization efforts, investment by public enterprises w o u l d grow little. Pakistan: Investment in percent of GDP, baseline scenario ” 2001/02 2006/07 Difference Total investment 13.1 16.7 3.6 Public 4.8 5.7 0.9 Government 3.5 5.0 1.5 Public enterprises 1.3 0.7 -0.6 Private 8.3 11.0 2.7 Public investment in % o f total 36.6 34.1 -2.5 Private investment in % o f total 63.4 65.9 2.5 Public investment, Rs b i l l i o n 174.2 333.2 I/ Gross fixed capital formation 21 Assumes continued privatization o f selected public enterprises. Nevertheless, our projections provide for a substantial growth in development budget as w e l l as non-development spending (excluding interest, defense and SOEs). Development budget i s expected t o grow f r o m Rs. 126 b i l l i o n in 2001-2 t o Rs. 292 b i l l i o n in current prices, rising f r o m 3.5 t o 5 percent of GDP. The growth in non-development spending w o u l d b e slower but still t h i s spending would still account for about 60 percent o f the fiscal space created for non- interest n o n -defense spending. H o w would this potential expansion of public spending help the revival o f growth and poverty reduction? H o w will t h i s public sector expansion support the critical increase needed in private investment? The institutional, allocation, and governance options that will best ensure positive outcomes f r o m public spending and w o u l d provide necessary support for the private sector are discussed at length in Chapters 2 t o 5. Only two points should be noted here. First, without substantial increase in capacity at a l l levels o f government and without significant improvements in budgetary and planning processes and fiduciary controls, a large expansion o f public spending would n o t b e justified. Secondly, the government spending should strictly avoid substituting for private sector activities. T h i s can be ensured if the privatization efforts continue and determined efforts are made t o promote public - private partnerships in the delivery o f services especially social services. With the above caveats, there appears to be substantial need o f and r o o m for larger public spending especially in human resource development, maintenance of existing infrastructure, and longer term n e w investments especially n water. The reorientation o f public sector priorities t o facilitate growth and make the growth more pro-poor are i n the next two chapters. discussed i 2.63. The government needs to continue i t s efforts to improve governance and strengthen institutions, including those in the public sector. This entails reforming government institutions and behavior to limit bureaucratic harassment and red tape (especially in the administration o f regulations and taxes). Further institutional strengthening and reform i s also needed in the financial sector, where the government in recent years has initiated improvements in the regulatory environment and enforced more stringent supervision. More broadly across the economy, the government also needs to more strongly assert the rule of law, including law enforcement, and reduce corruption and crime. I t should also improve the provision 38 o f public services such as the quality and availability o f infrastructure (such as power, water, transport, and telecommunications) and-through improved education-raise the quality o f the labor force. Facilitating the availability o f trained and slulled manpower could involve reviewing and improving labor laws and policies. The current anti export bias in some industries could be reduced by lowering duties on intermediate products. 2.64. The challenge for the new government is, on the one hand, to maintain the confidence o f international financial institutions and private investors, both domestic and foreign, and, on the other hand, to use appropriate incentives, policy guidance, and institutional support to ensure that development resources are well directed. There i s a substantial unfinished economic and social agenda, and major problems remain. But for the f i r s t time in many years there i s n o threat o f financial crisis hanging over Palustan’s head. 2.65. The government i s n o w aiming t o revive economic growth to 6 percent per year in the next three or four years. Achieving this rate will require, above all, a strong recovery in the agricultural growth rate and a sharp expansion o f manufactured exports. Some o f the policy and institutional elements required for improved performance in agriculture and exports are listed. 2.66. In agriculture there i s a need’ for improved canal irrigation and maintenance; better incentives for water conservation, specifically sharply higher water rates and better management o f irrigation systems; reduction in adulteration o f seeds, pesticides, and fertilizers through forceful use o f regulatory authority; and shifting o f economic resources, including scarce water, away from production o f sugarcane t o other crops through suitable price support adjustments. 2.67. Increased exports can be achieved by undertalung the following measures to improve Palustan’s external competitiveness: raise productivity by building human capital, upgrading labor skills, and hence more easily absorbing new technology; facilitate the availability o f trained and skilled manpower by improving labor laws and policies; improve product quality and the reliability o f delivery (important contributions would include more reliable power supply and transport systems); reduce the antiexport bias by lowering duties on intermediate products and possibly establishing virtual free-trade zones; attract foreign direct investors for export-oriented production, drawing on their established international marketing channels; focus policy on higher value-added exports rather than volume o f exports and the emphasis on cotton yam; and prepare for the new international regime in garments and textiles trade in 2005. Alternative Scenarios 2.68. T o sum up, the room for maneuver in the public sector i s increasing. I t will be possible to expand public development and social spending at a pace faster than GDP growth, consistent with debt burden reduction, but only if public enterprise losses are reduced, defense spending i s contained, government revenues show healthy growth, and donors are willing to provide substantial grants and concessionary loans. In the medium t e r m the revival o f well-directed private investment will be critical for reviving growth, revenues, and exports. More rapid increases in social spending and public investment than envisaged in the baseline scenario can certainly be justified o n need grounds but will require a much more forceful resolution o f public enterprise issues, especially in power, and reduction in domestic borrowing costs than appears likely at the moment. Given the constraints o n public spending, the institutional implications o f resource allocation and the effectiveness o f spending assume special importance. These are discussed in subsequent chapters. The improvements in the effectiveness and targeting o f spending will contribute to a higher growth rate but, as discussed later, will be particularly significant for attaining poverty alleviation and social development goals. The performance o f social sectors has been poor in the past, as much because of inadequate spending as because o f poor governance and de facto poor targeting 39 o f service delivery (the benefits o f the public services have not accrued to the poor but rather to middle- and higher-income groups). 2.69. The government should certainly explore the policy implications o f doing better than the cautiously optimistic baseline scenario presented here. But i t should be stressed that the base case scenario does assume a continued strong reform effort. If reform falters in just a few areas--revenue mobilization, public enterprise restructuring and privatization, and domestic and external debt management--it will quickly reverse the trend toward a higher growth rate and compound the problem o f poverty alleviation. This i s illustrated in the l o w case scenario discussed below. Low Case Scenario 2.70. The l o w case scenario assumes slippage in maintaining recent proactive and reform-oriented policies, in particular with regard to the fiscal budget, public enterprises, and other structural reforms that have a bearing on the investment climate. The main differences in outcomes over the medium term in this scenario are worsened fiscal balances, less fiscal space, and slower economic growth compared with the baseline scenario (see Table 2.6). A more detailed description i s provided in Annex B. . The main assumptions are Fiscal revenue collection gradually declines as a percentage o f GDP over the medium term (FYs 2004-07) from the relatively high level in FY 2003. The average revenue/GDP ratio i s projected at 16.8 percent o f GDP, substantially lower than in the high-growth scenario (average 17.8 . percent o f GDP). Public enterprise-related costs average 1.6 percent o f GDP during FYs 2004-07 as against 1.0 . percent in the base scenario. Domestic interest payments average 4.0 percent o f GDP during FYs 2004-07 compared to 3.6 percent in the base case. This reflects both that domestic debt i s higher in the l o w case and . interest rates decline more slowly. The Government does not manage to contain defense spending as a share o f GDP. I t rises throughout the projection period, reaching 5 percent of GDP by F Y 0 7 and averaging 4.6 percent of GDP, compared with 3.9 percent in the base case. 40 Table 2.6: Summary of key macroeconomic variables in low case, 1999/00-2006/07 Est. Prov. Est. Proj . 1999100 2000101 2001102 2002103 2003104 2004105 2005106 2006107 (annual changes in percent) Real GDP at market prices 4.2 2.6 2.8 5.8 5.0 4.9 4.9 4.9 Consumer prices (p.a.) 3.6 4.4 2.7 3.1 4.1 4.1 4.2 4.2 (in percent of G DP) Savings a n d investment Gross national savings 14.1 13.6 17.4 21.3 15.1 15.3 14.7 14.3 Gross fixed capital formation 14.4 13.9 13.1 13.1 13.7 14.1 14.3 14.4 Change in stocks (nongovt) 1.6 1.6 1.6 2.4 2.0 1.6 1.6 1.6 Public finances Revenue (including grants) 17.3 17.3 19.3 20.8 18.1 18.3 18.0 17.7 Grants 1.1 1.2 2.3 2.8 1.0 1.4 1.4 1.3 Revenue 16.3 16.2 17.0 17.9 17.1 16.9 16.7 16.5 Expenditure 22.8 21.4 23.6 22.3 22.2 22.4 22.4 22.4 Interest payments 7.8 6.8 6.8 5.2 4.8 4.8 4.6 4.5 Defense 4.8 3.8 4.1 4.0 4.1 4.4 4.7 5.0 SOE-related spending ” 0.5 0.6 1.9 1.2 2.0 1.8 1.7 1.5 Non-interest, non-defense, non- SOE spending 9.8 10.2 10.9 12.0 11.4 11.5 11.4 11.3 Overall balance (including grants) -5.5 -4.1 -4.4 -1.6 -4.1 -4.1 -4.4 -4.6 Primary balance (including grants) 2.3 2.8 3.8 3.6 0.7 0.7 0.3 -0.1 Gross public debt ’’ 100.9 108.3 100.1 91.8 83.6 81.5 78.9 77.9 Gross public debt in % o f revenue 619.6 670.5 588.9 512.0 488.8 482.5 473.0 473.2 Extemal sector Current account incl. off. transfers -2.0 -1.9 2.7 5.9 -0.4 -0.4 -1.2 1.7 Ext. public and publicly-guar.debt 48.6 55.7 52.7 45.8 38.1 35.0 32.5 30.4 Gross reserves in months o f next YearimPofisofgoodsandservices 3.8 1.7 3.7 5.4 4.4 4.4 3.5 3.5 1/ Subsidies and payment o f contingent liabilities. 21 Excludes SBP. 2.71. Consequently, the fiscal deficit grows to an average o f 4.3 percent o f GDP during FYs 2004-07 instead o f declining t o 3.0 percent as in the base case. The average fiscal space for economic and social spending increases only by 0.6 percent o f GDP, compared with 1.9 percent in the base case. 2.72. The policy slippage leads to less strengthening o f confidence. Hence, the reduction in interest rates on domestic public debt i s slower than in the base case. There i s also less foreign financing-both grants and concessional loans--reflecting less confidence by donors and I F I s in the government’s commitment to pursue reforms and address the social and public investment needs. Progress on privatizing public enterprises i s slower, and the projected sales receipts are lower than in the base case. 2.73. Total public debt i s higher and with less concessional terms o n average. By FY 2007 public debt as a share of annual fiscal revenue declines only to about 480 percent, whereas in the base case i t falls to about 390 percent o f revenue. In other words, progress toward reaching public debt sustainability i s insufficient in this scenario. 41 2.74. Furthermore, economic growth, j o b creation, poverty reduction, and improvements in social indicators are slower than in the base case. GDP growth averages 4.9 percent per year during 2004-07 compared with 5.6 percent in the base case. The slower growth reflects lower private investment and lower growth-related public spending o n health, education, and infrastructure-both operations and maintenance and new investments-and hence lower productivity o f public capital stock and human capital. I t also reflects slower improvements in the effectiveness o f govemment spending, govemance, and implementation capacity. 2.75. The differences described in the fiscal outcomes between the scenarios are even more accentuated when one looks at the momentum at the end o f the projection period. By FY 2007 the fiscal deficit grows to 5.9 percent o f GDP (from 4.5 percent in FY 2003) instead o f declining to 4.3 percent as in the base case. In spite o f the larger deficits, the fiscal space for economic and social spending increases only by 0.4 percent o f GDP by FY07 compared to 3.4 percent in the base case. Over the longer term the base case would continue to show improved debt reduction, growth, and social indicators, and the gap between the two cases would continue to widen. The l o w case would basically show stagnation. Table 2.7 Pakistan: Low case scenario, Principal Budget Aggregates Change FY2002-FY2007 (Percent of GDP) Resources Expenditures Government revenues - 0.5 Defense expenditure + 0.9 External grants - 1.O Interest payments - 2.2 Fiscal deficits + 0.3 Public enterprise losses 1/ - 0.4 (including grants) Non-interest non-defense spending + 0.4 olw Development spending -0.7 Total resources -1.3 Total expenditures -1.3 l/Public enterprise losses are approximated by subsidies and payment o f contingent liabilities. High-Growth Scenario 2.76. A scenario with more rapid economic growth and more fiscal space than in the base case i s certainly possible. Achieving this outcome will, however, require three things: First, the government must make a determined effort to lower the real cost o f domestic borrowing. Lowering the real interest rate from 6 percent to 4 percent by F Y 0 7 will save interest payments equivalent to 1 percent o f GDP. Second, i t must virtually eliminate public enterprise losses and privatize these enterprises more successfully, especially in the power sector. Third, it must improve the effectiveness o f govemment spending and institutional capacity t o better utilize the additional fiscal space o n poverty- and growth- related spending. If marked improvements in institutional capacity do materialize, spending on health and education as well as on infrastructure-both O&M and new investments-should increase. Over the medium and long term, this spending could contribute to higher productivity o f the public capital stock and human capital (and improved social indicators), further boosting economic growth. 2.77. In the event, however, that the effectiveness o f govemment spending, govemance, and implementation capacity does not materially improve even with higher economic growth and revenue collection, the govemment should use the additional resources available t o repay those debt items with the highest cost attached (currently long-term domestic debt instruments) to at least capture the gains o f saving future debt-servicing costs. 42 Conclusions 2.78. Pakistan i s at an important economic turning point. The debt problems that have been a major cause o f the sharp slowdown in Pakistan’s economic growth over the past decade, and the consequent stagnation in poverty reduction, have been brought under control. Although debt, especially public debt, i s s t i l l at an unsustainable level and needs to be brought down steadily, the risk that economic and other shocks will lead to a worsening o f the public debt burden in the medium term appear minimal (see the sensitivity analysis in Annex B). N o doubt, there i s a substantial unfinished economic and social agenda, and major problems remain. But for the first time in many years, there i s n o threat o f a financial crisis hanging over Palustan’s head. 2.79. Looking ahead, the issue i s not so much whether macroeconomic stability can be maintained and orderly reduction in debt burden achieved, but whether public expenditure management and restructuring can play a more positive role in the economy than in the past decade. A financial equilibrium at a level and pattern o f public expenditures that does not support provision o f essential public goods and does not further social goals will hardly serve Palustan’s long-term interests. 2.80. With continued reform aimed at increased mobilization o f government revenues, a sharp reduction in the losses o f public enterprises, a reduction in the costs o f domestic government borrowing, and defense spending not exceeding 4 percent o f GDP for the near future, it should be possible to create substantial fiscal space to further the government’s objectives o f reviving growth and reducing poverty while continuing t o make progress on reducing the debt burden. 2.81. This chapter illustrates essentially two economic and fiscal scenarios and discusses, at some length, the policy issues and options involved in bringing about a shift in public spending toward development and social spending. Table 2.8 presents a summary overview o f the main differences between scenarios in the size o f the fiscal space available for non-interest, non-defense and non-SOE spending and the elements that contribute to different outcomes. Table 2.8: Summary of Fiscal Space Alternatives over the medium term Fiscal space altematives l’ Reference: 2001102 estimate LOW case Base case Fiscal space in % o f GDP in F Y 0 7 10.9 11.3 14.3 Fiscal space in % o f GDP - average FY03-FY07 10.9 11.5 12.8 Fiscal space - average annual real growth from FY03 to F Y 0 7 6.3 11.7 Real fiscal space in FY07, in constant F Y 0 2 prices, Rs billion 395.1 531.4 684.0 l/Fiscal space i s defined as spending excl. interest, defense and public enterprise losses. The latter i s approximated by the s u m o f subsidies and payment o f contingent liabilities. 2.82. In the baseline scenario real non-interest, non-defense and non-SOE spending expands robustly over the medium term, and this expansion, together with stringent measures to ensure improved effectiveness and targeting o f the spending, should enable the government to facilitate economic growth and make a substantial dent in poverty, markedly improve health and education indicators, and hence narrow Palustan’s substantial social and gender gap. Non-interest, non-defense spending under the baseline scenario grows o n average by 11.7 percent per year in real terms, be on average 27 percent larger during FYs 2004-07. I n the l o w scenario this growth i s only 6.3 percent, and thus there i s little room for reversing the relative decline in economic and social spending witnessed in the 1990s. I n the baseline 43 scenario non-interest, non-defense and non-SOE spending to 14.3 percent o f GDP, compared with only 10.9 percent in FY02, and recovers t o levels seen in the 1980s and early 1990s. 2.83. In the low scenario, government revenues decline gradually as a share o f GDP over FYs 2004-07, domestic interest payments remain at the FY03 level, public enterprise losses remain high, and defense spending rises steadily relative to GDP. The interesting point i s that even with a higher fiscal deficit, compared with the baseline scenario, the fiscal space under the l o w scenario i s only a fraction o f what could materialize under the continued reform scenario. 2.84. I t i s crucial for successful implementation o f the baseline scenario that governance improves sharply. This means that government should combat corruption, leakage, and deficient revenue collection and control expenditures resolutely and assertively. The benefits will be seen in larger fiscal space to be used for high-priority spending, a reduced debt burden, rising investor confidence, and hence increasing private investment-both domestic and foreign-and growth. In addition, the government i s more likely to get a higher share o f concessional foreign financing if strong measures to improve govemance are part o f an overall reform package. 2.85. A firm commitment by the government i s needed to stay the course with regard to economic and governance reforms in the face o f political opposition, including from vested interests and in particular in the area o f public enterprise reform. If the government can obtain the projected gains for another 3-4 years, the fiscal and external dynamics could be entrenched on a path toward sustainability. 2.86. Pakistan’s development partners have a crucial role to play in providing sufficient foreign financing in the form o f grants or loans on concessional terms. Without such financial assistance in the magnitudes envisaged here, it will be very difficult for the Government o f Palustan to achieve the three central goals (debt burden reduction, acceleration in economic growth, and poverty alleviation) all o n its own. Hence, substantial additional foreign financing o n concessional terms i s an integral part o f this reform and development strategy. 44 3: R CHAPTER EORE I NTI N R G PUBLIC SECTOR PO T IRI EIS Introduction 3.1 The previous chapter outlined the fiscal space that can be created and enhanced by the continuation o f reform efforts. This chapter and the next discuss the issues involved in making optimal use o f the possibilities o f higher economic and social spending. The effectiveness o f public expenditures has several dimensions: 1 Strategic approaches to public interventions Choice o f sectoral and subsectoral priorities 1 Soundness, transparency, and accountability o f budgetary and planning processes 1 Careful screening o f projects and speedy and effective implementation o f programs 1 The existence o f strong monitoring and evaluation systems 3.2 The presence o f hnctioning institutions and a framework o f defined distribution o f responsibilities at various levels o f government are the cross-cutting issues that influence all aspects o f effectiveness o f public spending. This situation requires not only the statues, laws, and rules that determine the establishment and functioning o f various institutions, but also effective management and leadership so that each institution has the flexibility to perform i t s statutory duty in collaboration and not in competition with others and t o adapt i t s roles according to changing needs. 3.3 This chapter focuses principally on the intersectoral and subsectoral priorities, though obviously the issues o f project choice, budgetary and planning processes, and implementation cannot be entirely separated from sectoral allocations and priorities. 3.4 It i s clearly not feasible to undertake a comprehensive sector-wise review o f public expenditures in the context o f this overview report. The attempt here i s to examine, in some detail, a few key sector allocation and implementation issues that will greatly influence the ability o f the government t o achieve i t s stated goals. A l o o k at the selected sectors can also illustrate the constraints public spending has faced, the less than optimal way this spending has been used t o advance economic and social policy goals, and the institutional weaknesses that have contributed to often disappointing and occasionally adverse consequences o f public spending for the economy and the society. Even more important, looking to the future, such an analysis can suggest ways for minimizing the public policy dilemmas (and increasing policy choices) arising from the extremely constrained financial resource position on the one hand and the desperate need on the other hand to step up spending o n education, health, infrastructure, and direct poverty alleviation programs. 3.5 In reviewing the government’s expenditure priorities and options in allocating the fiscal space, this chapter examines the entire public expenditure in the selected sectors and not just the development spending. The notion of development spending, including as it does the investment expenditures, technical assistance, and some initial recurrent expenditures related to new capital spending, may have outlived i t s usefulness as a sectoral and national planning tool. Indeed, to some extent, the implied notion that nondevelopment spending i s o f lower priority has led to an inefficient use o f economic resources, It has contributed to inadequate allocations to the social sectors, woefully inadequate allocations for operation and maintenance, and the creation o f new physical assets at the cost o f the effective use of existing assets. It has also l e d planners to be preoccupied with the development budget while neglecting sectoral analysis and the impact o f so-called non-development spending o n the country’s economic and social goals. The introduction o f LTRSP and the Medium-Term Budget Framework (MTBF) provide important opportunities for moving away from the excessive focus on the development budgets and toward a more comprehensive approach to public expenditure management. 3.6 Using this comprehensive sector-wise approach, the discussion here focuses on allocation and implementation issues in three sectors that will greatly influence the ability o f the government to achieve . i t s stated goals: . . Irrigation water Power Education and health 3.7 Public spending o n a host o f other sectors, such as law and order maintenance and provision o f justice, agriculture, roads, population welfare, rural development, and targeted transfers (food support program, Zakat, etc.) n o doubt also deserves priority. The government includes many o f these (in addition to education and health and a part o f irrigation spending) in the l i s t o f high-priority spending to be monitored under the I-PRSP. The costing o f PRSP-related expenditures and relationships between expenditure inputs and output targets are as yet extremely tentative, but the government has rightly given the protection o f I-PRSP spending the highest priority. Issues related to the adequacy o f funding o f I - PRSP cannot be analyzed in a comprehensive fashion in this report because o f a lack o f costing data. They are touched upon, however, in the next section and discussed at some length in the section on education and health. Since the social sectors account for 60 percent o f identified high-priority spending for poverty reduction in the I-PRSP, there i s a considerable overlap between the latter and the sectors selected here. In any case, the issues o f the coverage and adequacy o f high-priority poverty spending must be revisited in the context o f the finalization o f the PRSP. This spending should have the first claim on the fiscal space that will be created by the continuation o f reform policies. Fundingthe Poverty Reduction Strategy 3.8 The Government o f Pakistan i s following a multiprong poverty reduction strategy, which i s outlined in the IRRSP and will be further elaborated in the full PRSP n o w under preparation. The core principles o f the strategy include engendering growth, implementing broad-based governance programs, improving social sector outcomes, and reducing vulnerability to shocks. The main macroeconomic targets in the I-PRSP that is, growth rates o f GDP, revenue, and inflation--are broadly in line with the baseline scenario presented in Chapter 2. T w o questions related to public expenditures are relevant for the . implementation o f the poverty reduction strategy: Are the indicated YPRSP expenditures consistent with the achievement o f specified goals especially in the social sectors and with the internationally agreed-upon Millennium .Development Goals (MDGs)? Can the financing needed for the poverty reduction strategy in the medium t e r m be accommodated within the fiscal space likely to be created? 3.9 The high-priority expenditures for poverty reduction identified in the VPRSP show a relatively modest planned growth, from 3.4 to 4.0 percent o f GDP, over the period 2001-04. T h i s allocation appears to be totally inadequate in relation to ambitious targets, especially in education. On the one hand, there i s a need to set more realistic but s t i l l ambitious social targets consistent with M D G s . On the other hand, the financing needs for meeting these targets require more careful assessment. 3.10 As discussed later in this chapter, achieving goals in education and health will require substantially larger funding than assumed at present. The full PRSP will probably extend the planning 46 horizon to FY 2006 and i s expected to provide a fuller costing o f social sector goals so that inadequate funding does not frustrate the essential goals o f human development, as often happened in the past. 3.1 1 I t i s suggested that the high-priority PRSP spending also cover l a w and order maintenance and provision o f justice and rural electrification. The governance problems resulting from inadequate spending on police and the justice system have their greatest impact o n the poor. Similarly, i t i s the poor in the rural areas who most lack access to electricity; o f the roughly 40 percent o f Pakistani households that do not have electric connections, most live in the rural areas and are poor. Speedy rural electrification must be an important part o f increasing access to education and health care, dissemination o f new technologies, and general modernization. Rural electrification cannot be undertaken as a commercial proposition and will require special public sector financial support. A system o f capital grants to rural communities and electric cooperatives can be developed that gives priority t o those who mobilize enough own resources t o ensure financial viability after the initial grant. I t i s worth looking into Bangladesh’s experience. Donors will probably show substantial interest in providing grants or low-cost loans to a program whose benefits go overwhelmingly to the poor. The development and expansion o f the rural electrification network should not be limited to the Water and Power Development Authority (WAPDA) and should involve the private sector. 3.12 Using the broader definition o f poverty-related spending suggested, the actual priority poverty spending in 2001-02 was close t o 5 percent o f GDP. Our rough estimates suggest that this spending should increase to about 7 percent o f GDP by 2006-07. Higher education and health spending would probably take up h a l f o f the increase in order to achieve the goals set; the rest would go largely to higher allocations for l a w and order, provision o f justice, and rural and urban infrastructure, including rural electrification. The baseline scenario, which projects that non-interest, nondefense spending could increase from 10.9 to 14.3 percent o f GDP over the period 2002-07, can accommodate the r i s e in high- priority poverty-related spending o f the order o f magnitude suggested here, provided the political commitment to reorient the public spending toward social goals and better governance remains strong. It i s very encouraging that the recently completed PRSP (December 2003) provides both for a broader definition o f poverty spending suggested above and proposes to increase this spending steadily t o 6.8 percent o f GDP by 2007-08. Remedying the Neglect o f Education 3.13 Palustan’s I-PRSP recognizes the importance o f improving education outcomes. Learning from the experience of the Social Action Program (SAP), the government strategy i s to significantly improve the social service delivery system by addressing the underlying governance problems. I t will then gradually increase funding as allowed by the impact o f improvements in tax administration, stronger economic growth, and assistance from donors. In education, the linchpin o f this effort i s the federal government’s action plan in support of education sector reform (ESR), whose implementation lies with the provincial and district governments. 3.14 Educational attainment and health improvements are among the most valuable benefits o f human development and are also--since they constitute vital investment in human capital--important determinants o f growth and poverty reduction. Pakistan has seriously neglected investment in human capital and has paid the price for i t not only in persistently high growth in population for a long time, but also in the slowing o f growth and persistence o f poverty. In education, the problems are the very l o w level o f enrollments not only at the primary, but also at the middle, secondary, and higher education levels and the l o w and declining quality o f public education. 3.15 The irony i s that an increased emphasis o n elementary education and protection o f i t s relative share in public spending in the 1990s did not achieve the desired (primary and middle) results, whereas 47 per-student spending at the secondary and higher education levels declined by as much as 30 and 50 percent respectively in real terms, compounding the crisis o f quality. As i t moves ahead, Pakistan needs to make a major dent in the social gap at all o f levels o f education and accompany it with a demonstrable reversal o f the trend in declining quality. The approach t o remedying the social gap needs to be comprehensive, setting ambitious but attainable targets and backmg them up with adequate funding, clear institutional responsibilities, and a system o f incentives and workable public participation arrangements including private-public partnerships, citizen boards, and parent-teacher associations. 3.16 The basic strategy embodied in the ESR i s sound. I t sets explicit targets for each level o f education (Table 3.1) expected to be realized by provincial and district governments and aims to achieve these targets through increased funding, improved resource mobilization and utilization, and the introduction o f institutional reforms at all levels o f education to improve governance. Table 3.1: ESR Targets by Sub-sector, 2001-2005 2001 Actual 2005 Target (“A) (%I Literacy 49.0 61 Gross Primary Enrollment 83.0 100 N e t Primary Enrollment 66.0 76 Middle School Enrollment 47.5 55 Secondary School Enrollment 29.5 40 Higher Education Enrollment 2.6 5 Source: Government o f Pakistan, Ministry of Education’s Education Sector Refoms document based on provincial Government data. PIHS data has differences. 3.17 As already noted, however, several parameters o f ESR, notably the specific targets, funding requirements, and financing mechanisms, need to be revisited. More important, the education sector plans need to be made more operational in the context o f finalizing the PRSP, using more recent information on actual enrollment levels and the decision t o allow district governments more freedom in choosing expenditure priorities. 3.18 The ESR targets are ambitious and will entail considerable commitments from the government- for example, gross primary enrollments are targeted to increase by 16 percentage points fkom 2001 to 2005. T a h n g into account the population growth, meeting this target would require enrollment o f roughly an additional 5-6 million children at the primary level. Even this figure understates the problem. There i s lack o f agreement on the starting level o f gross primary enrollments: P M S data suggest that current primary gross enrollments may be as l o w as 72 percent compared with a figure o f 83 percent from government administrative data (Table 3.1 .) Similar discrepancies may exist at other levels o f education. 3.19 The ESR had projected a cost of Rs. 100 billion, with roughly 34 percent o f the total to be allocated to primary education,32 over and above the present annual national expenditures o n education. I- PRSP, on the other hand, assumes a growth in public education spending in line with growth in nominal GDP: the percentage of expenditure on education remaining at 1.8 percent o f GDP. Figure 3.1 illustrates the existing difference in projected expenditures between I-PRSP and ESR. 32 T h e original ESR package was prepared at a cost of Rs. 55.5 billion for the years 2001-04. The duration of this package has been extended to 2001-05 to accommodate the President’s Programs, the mainstreaming o f Mmadaris, and the setting up of polytechnics at the Tehsil level. Therefore, the cost estimates have increased to Rs. 100 billion. These estimates, however, do not include recurrent expenditures of the ESR program. The districts and provincialiarea governments are expected to review and assess their own requirements and adjust the non-development budgets. ESR estimates also do not include the provincial education Annual Development Programs (ADPs). 48 3.20 The bulk o f the ESR Figure 3.1: EudgetaryExpenditure o n Educaation expenditure at the primary level i s 1" allocated to improving and expanding ioqmo rundown physical infrastructure, with sqmo additional support for institutional changes, such as the development o f parent-teacher associations. Other features o f the overall strategy include teacher training, an adult literacy campaign, mainstreaming o f the madarassas, a pilot nutrition program, introduction o f a technical stream in secondary schools, reform of examinations, and introduction o f a national education assessment program. With the devolution o f authority t o the district level for the bulk o f education spending, the funding f l o w arrangements for supporting the objectives o f ESR are also not clear. 3.21 As already mentioned, the finalization o f PRSP during the next few months provides a good opportunity to resolve some o f the issues and contradictions noted. 3.22 First o f all, the targets for elementary school enrollment need to be reviewed. The 100 percent gross primary enrollment target by 2004 (I-PRSP) or 2005 (ESR) i s n o longer realistic. On the one hand, firmer information needs to be developed o n the present position (at all levels o f enrollment), and o n the other hand a credible universal education attainment target (say by 2006 or 2007) need to be developed and backed up by a policy package, funding, and institutional arrangements. The ESR middle and secondary school and higher education targets will also need to be moved out in time by a couple o f years. Although all education sector goals appear ambitious in terms o f the effort required to achieve them, they remain modest by international norms. All countries in South Asia except Palustan have gross enrollment ratios o f 100 percent or higher. The ESR, if successfil, will achieve a middle school enrollment rate o f 55 percent by 2006 or 2007. I n Korea three-fourths o f the relevant age group was completing middle school in 1975. Pakistan hopes to raise i t s secondary enrollment rate t o 40 percent over the next four to five years. Malaysia had achieved this rate by 1974. In higher education Pakistan has an enrollment rate (1996) o f 2.6 percent. T h i s i s one o f the lowest proportions in the world--for India the figure i s 6.2 percent (1990); for Iran (1994) it i s 12.7 percent, and for several East Asian counties this rate exceeds 30 percent. In education Palustan appears to be at least thirty-five to forty years behind the East Asian countries and ten to fifteen years behind the South Asian countries. Consequently, an acceleration i s needed to enhance enrollment levels at all levels o f education. T h i s goal will require the government to increase resources allocated to education and to improve the institutional capacity to use the additional resources effectively and efficiently. 3.23 In the very short term, high priority must be given to primary education, including nonformal education, because without rapid universal education, progress toward increasing literacy levels, n o w estimated at 45-48 percent, will remain slow. Expanding public expenditure for primary education i s also crucial because it i s more pro-poor and also helps the rural areas more than the urban areas (see B o x 3.1). But high-quality primary (and secondary) education requires parallel development o f higher education. I t i s from this level that the country gets i t s teachers, scientists, and managers. 3.24 The lessons learnt from the experience o f the past decade, in which expansion o f elementary education spending resulted in only moderate benefits, are already being reflected in approaches to elementary education. The provincial governments o f Sindh and Punjab are n o longer giving priority to 49 the opening o f n e w schools. As emphasized in the ESR, attention i s being given to improving quality by hiring new teachers, increasing accountability o f teachers, improving physical facilities in schools, and raising nonsalary spending. In Punjab special salaries for teachers (and doctors) in rural areas are being considered. The Human Development Foundation set up by the Government o f Palustan as an autonomous corporation i s emphasizing enrollment processes, nonformal education, especially for recent dropouts, and innovative approaches to capacity building. The Pakistan Poverty Alleviation Fund has expanded its mandate to include education and social sectors. But it i s not at all clear that the programs, new initiatives, institutions, and funding commitments are in place o n a scale necessary to achieve the key targets in primary education: 35-40 percent growth in primary enrollments over the next five years, the narrowing o f gender and geographical differences, significant improvements in quality, reduction in dropout rates, and expansion o f programs o f non-formal education. Box 3.1: The Rich-Poor Gap and Public Expenditure on Education According t o Figures 3.2 and 3.3, income groups f r o m the second decile t o the fifth decile receive the highest benefits f r o m public expenditure o n primary education. In this sense, public expenditure for primary education i s pro-poor - an effect that i s partly attributable t o the fact that poor children are more l i k e l y t o attend public primary schools, as opposed t o higher income groups w h o have a higher private-public enrollment ratio. On the other hand, income groups f r o m the seventh decile t o the ninth decile receive the highest benefits f r o m public secondary school education - an effect that arises mostly from the fact that poor children are far less likely t o enroll in secondary school. Figure 3.3: Distribution of Public Incidence of Primary and Expenditure for Primary and Secondary Education Secondary Education 14 zw 12 150 10 $ 8 1w 50 0 4 , , , , , , , , , 1 0 , , , , , , , , , 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 6 Q 10 Per Capita Consumption Expenditue Per Capita Conumption deciles Expenditure deciles / p r i m a r y~ n d a r I y / p i m a wp ~ a r y The Urban-Rural gap and public expenditure on education As Figure 3.4 shows, rural communities, o n average, receive more benefits f r o m public primary education than urban communities. I t is, however, worth noting that poor urban families receive more benefits f r o m public primary education than urban communities. As for secondary school education, urban communities receive more benefits f r o m public secondary schools than rural communities Figure 3.4: Benefit Incidence of public primary education (Urban Public Secondary School (Urban vs vs Rural) Rural) 2% 250 1 I m0 1% 150 100 IW E a 50 - 0 01 , , , , , , , , , A 1 2 3 4 5 e 7 8 9 10 1 2 3 4 5 5 7 8 9 1 0 Per Capita Consumption Expenditure Per Capita Consumption Expenditure 1 deciles deciles I L 50 3.25 In addition to issues related to the devolution process (discussed in detail in Chapter 4), three sets o f issues appear especially important. First, enrollment levels in Pakistan are among the lowest in the developing world, and gender disparities and rural-urban gaps33 are among the widest and point to tremendous unrealized potential and missed opportunities. But there are several incentive programs that have been used around the world to stimulate demand for education and encourage enrollments. Second, the private sector i s the most dynamic segment of education.34I t accounted for more than 40 percent o f the growth in primary school enrollment in the past decade. The average teacher salary in a private school i s actually lower than in public schools, but the quality o f teaching appears to be better because o f much greater accountability. Effective public-private sector partnerships that tap into the strength o f the private sector have yet to be developed. Last but not least, more detailed and realistic estimates o f public funding requirements, not only for primary education but also for the entire education sector, are needed. 3.26 Clearly increasing public funding alone will not meet the urgent challenges in the education sector. But higher public spending i s a necessary if not sufficient condition for improving educational attainments. Without additional funding, devolution will not work, needed quality improvements, especially at the primary level, will not materialize, the scope for incentive schemes will be limited, and the needed recovery in higher and secondary education will not take place. This situation will once again frustrate the attempt t o achieve the ambitious targets. Detailed work i s necessary to firm up the education funding requirements, including for the secondary and higher levels. Our tentative estimates suggest the need to gradually raise public education expenditures from 1.8 percent o f GDP in 2001-02 to 2.5 percent in 2006-07. This growth will translate into average spending o f 2.2 percent o f GDP over the next five years, malung possible an increase o f about 45 percent in real terms over the 2001-02 level. It will also mean that education spending will take up more than one-third o f the fiscal space generated under the baseline scenario. A substantial part of the growth in relative spending will be absorbed by primary and secondary education, though the downward trend in allocations for higher education also needs to be reversed. 3.27 Within the recommended overall increase in education expenditures, a more rational allocation o f spending i s needed. M o r e rational allocation would include increasing resources for nonsalary recurring expenditures for primary and secondary school improvements; provision o f missing facilities in existing infrastructure, especially at the primary and secondary level; provision o f quality services such as teacher training; increasing resources for new infkastructure where there i s a need, that is, at the middle and secondary levels o f education; girls’ incentive programs; and demand-side interventions. 33 In 1999, for example, female enrollment rate was 23 percent lower than that for males. As in the case o f gender, there are significant disparities in educational attainment across regions. Most striking i s the urban-rural gap. Unlike the gender gap, the urban-rural gap has widened over the years. In 1991 primary school enrollment in the rural areas was 22 percentage points lower than the urban areas, b y 1999 it had increased to 30 percentage points. Furthermore, gender gaps in school enrollments vary substantially across regions and between urban and rural areas. In 1999 the primary schooling enrollment rate was more than 99 percent in urban males o f the Punjab but barely 28 percent for rural females o f Sindh. The gender gaps in these enrollment rates ranged from less than 3 percent in urban areas o f the Punjab to about 34 percent in rural areas o f the North West Frontier Province (NWFP). 34 According to a recent Census o f Private Schools (2000), Pakistan has as many as 40,000 o f such private schools, the result o f significant growth in private schooling over the 1990s. The share o f the private sector in total national primary school enrollment increased from 14 percent in 1991 to about 28 percent at present. The private share o f secondary schooling increased from 8 to 17 percent over the same period. T h i s phenomenon was not limited to the higher-income groups. The private share in primary school enrollment grew from 5 to 10 percent for the lowest-income group, while it grew from 35 to 60 percent for the highest- income group. Research on private schooling, based on data from the aforementioned census, also shows that private schools seem generally to offer better gender outcomes, partly because they are coeducational institutions, and also have been successful in recruiting female teachers. 51 3.28 Incentive schemes can be used to promote enrollment in both government and private schools. Voucher schemes, which directly support the students, can help create demand for schooling by meeting some o f the direct and indirect costs o f attending school. They can introduce school choice and competition between schools (reducing unit costs and improving quality) and can be targeted to subpopulations that are underserved or excluded (such as the poor or girls). Other demand-side enrollment incentives have been widely used around the world and take many forms-free textbooks, uniforms, transport, scholarships (Bangladesh has a very successful female secondary school scholarship program), and in-kmd payments to families. Palustan has some experience with these schemes, such as free textbooks in Balochistan and Sindh and some form o f scholarships in almost all provinces. The Food for Schoolgirls program run by the W o r l d Food Program provides inducement in the form o f a tin o f edible o i l to each student if she attends school for twenty days in a month. Initial results are encouraging and point to the wide potential that such schemes hold for improving enrollment and retention rates. The government i s also financing a food support program (Taiwana Pakistan) that should be evaluated after a year of implementation t o assess i t s impact on enrollment and retention. 3.29 While raising the level o f public spending o n the social sectors i s an important medium-term goal, and t w o o f the four provinces (Sindh and NWFP) have economic reform programs underway that hold promise for more spending for the social sectors, an additional concern i s achieving improvements in the efficiency o f present-day spending. Getting budget allocations released in a timely fashion to front-line spending agencies, as well as ensuring that allocations are appropriately balanced across salary, nonsalary, and the development budget, i s a top priority. Adequate provision and timely release o f the nonsalary recurrent budget i s critical to smooth operations at the point o f service provision. W h i l e increased budget allocations are a necessary step, it should also be recalled that in the past actual expenditures have often fallen below planned expenditures-for example, in Punjab actual expenditures have fallen short by as much as a quarter in recent years. Institutional bottlenecks that have disrupted the overall flow o f funds in the past are likely to be even more critical with the devolution o f service delivery to the districts, and the removal o f these bottlenecks i s clearly a high-priority area for reform. According to anecdotal information, the f l o w o f funds, especially nonsalary funds, continues to be unpredictable under the district governments. Predictability o f funding i s crucial, whatever the absolute level o f funding. 3.30 As a part o f increasing public expenditure on education, the government needs to explore new funding mechanisms to increase enrollment, raise literacy, and improve quality. These mechanisms include the Education Foundations, the Pakistan Poverty Alleviation Fund, the Human Development Commission, and education nongovernmental organizations (NGOs), such as Bunyad, which are active in the nonformal education sector. 3.3 1 H o w can public-private partnerships in education be strengthened? One obvious way i s to provide vouchers for students, especially girls, where there i s n o public school nearby (for other possibilities o f public financing o f private education (see Chapter 4). This seems quite practical because in a number o f cases private school fees are not high.35 Access to educational facilities i s an important determinant o f school participation. Physical proximity to a school appears to be quite important and i s a factor in explaining the gender gap in enrollments, since proximity to girls’ schools i s far from universal in rural ~ ~ that the presence o f a school offering primary areas. The regressions o f school p a r t i ~ i p a t i o nshow education within one kilometer increases the probability that a girl in a rural area will attend school by 15 percent. Existing initiatives, such as private sector use o f public facilities in the second shift and concessionary access to public land, should be strengthened. Finally, the government should resist the temptation to strongly regulate the private sector, at least in the next few years when the system desperately needs expansion. To some extent the institutions that will oversee the activities o f the private 35 See “Poverty Assessment, World Bank, 2002. 36 See Chapter 3, Pakistan: Poverty Assessment, The World Bank, 2002. 52 sector should be the same as those that oversee the activities o f the public sector, such as student assessment and examination bodies. 3.32 The fundamental problems that the government must resolve to get better outcomes in education are institutional. I n the absence o f fundamental reform, additional funding by i t s e l f will achieve little. M a n y o f these problems apply to the entire public sector, but they manifest themselves most visibly and with the greatest impact in the education sector because o f the size and multitude o f actors that constitute the system. Palustan has more than 150,000 primary and middle schools and more than 450,000 primary and middle school teachers. T h i s size speaks to the need for accountability, responsibility, service orientation, and incentives that reward people for doing their j o b well. Improving Health and PopulationOutcomes 3.33 During the last decade, Pakistan has made a noteworthy progress in improving i t s health and population indicators. Infant mortality rate declined from more than 100 per 1000 live births in early 1990’s to 82 in 2001/02 and the total fertility rate has declined from 6.4 to 4.8 during the same period. Yet because these improvements are from a very l o w base and a number o f problems in the health services delivery system remain unresolved, therefore there i s hardly in room for being complacent. Eighty percent o f women are s t i l l delivering without assistance from trained personnel; contraceptive prevalence i s s t i l l very l o w (28 per~ent);~’ and due t o poor nutritional status and l o w tetanus toxoid immunization coverage o f pregnant women (48 percent), about one in every three hundred women die in childbirth. Moreover, even the improved aggregate indicators conceal vast gender and location-wise disparities. Mortality among the girls ages one to four i s reported t o be 1.7 times higher than in boys in the same age cohort, and the mortality rate for women between the ages o f twenty and twenty-nine i s about double that o f men. All these indicators being significantly worse in the rural areas. Poor governance and l o w public sector spending on the health sector i s the main cause o f the persistence o f poor health indicators. 3.34 Poor governance, weak health sector management and l o w levels o f public spending are the key factors behind poor health outcomes. During the second h a l f o f 1990s, public expenditures o n health amounted to only about 0.6% o f GDP, among the lowest in the region. The government health services have suffered from insufficient budget allocations t o fund current expenses other than staff salaries. These include expenses for drugs, diagnostics (laboratory tests, x-rays), repairs and maintenance o f facilities, replacement o f equipment, utilities, transportation, in-service training expenses, and health education materials. As a result o f i t s poor quality, public satisfaction with the public health services i s l o w and i s reflected in l o w use o f government’s heath facilities (about 80 percent o f outpatient contacts occur in private facilities, and only slight less for hospitalization). In addition, access to other basic services such as safe water and sanitation i s s t i l l poor in many areas, which, together with poor nutritional status plays a major part in the vicious circle o f illness, l o w productivity and poverty. Health Sector Reforms 3.35 The government i s well aware o f the magnitude and depth o f the problem and has recently tried to address the broader issues o f governance in social sector delivery through major refoms, especially the devolution initiative, while a new agenda for improving health and population outcomes has been defined under the I-PRSP. The I-PRSP contains targets to 2003/04 for child mortality, the number o f Lady Health Workers deployed, prevalence o f malnutrition among pre-school children, percentage o f low-birth-weight babies, contraceptive prevalence rate, and total fertility rate, and it aims to introduce a mechanism t o monitor outcomes on a regular basis. 37 T h i s i s based on PIHS data, the govemment estimates are somewhat higher . 53 3.36 Moreover, the Government’s new health policy i s specifically designed to achieve the medium- term sectoral goals specified in the I-PRSP. The policy has targeted ten specific areas for reform including, reducing the widespread prevalence o f communicable diseases, addressing inadequacies in primary/secondary health care services, correcting male and urban biases, bridging basic nutritional gaps, improved regulation o f private health sector, and capacity building for health policy monitoring. The Government’s Population policy i s designed to achieve population stabilization by 2020 through mainstreaming the concept o f Family Welfare in national and development programs, expanding service delivery and improving i t s quality, fostering public-private partnerships wherever appropriate, promoting health education through advocacy with key influential leaders, training and capacity building o f service providers, decentralizing and de-federalizing the family welfare program, and strengthening health sector monitoring and research. The provinces have prepared plans to implement the policy, which are at various stages o f development. The Way Forward 3.37 The policy directions envisaged in the I-PRSP and the current health and population policies are by and large appropriate to address the issues faced by the sector. Overall policy objectives are however demanding in terms o f local capacity, and there i s as yet n o clear vision o n h o w to improve implementation o f health and population programs to deliver good results o n the ground. In addition, there are n o definite initiatives or plans and means outlined to improve the quality o f health care in both the public and private sectors. The biggest concem however i s that I-PRSP’s projected financing for health sector seems to be totally out of line with the targets laid down in I-PRSP and the sectoral MDGs. 3.38 T o make the public system more efficient and provide better value for money, the government needs to take a series o f strategic actions: The public health services should seek to achieve a sharper focus on service priorities, in terms o f both resources and management attention. These priorities need to be developed keeping in view the level and direction o f private sector involvement in the sector and the appropriate role o f the government. Optimally, the government health services should seek to “crowd in” rather than “crowd out” private health services. 1 Direct provision o f public health services should generally be based o n equity considerations and on existence o f public goods, externalities, information deficiencies, and insurance market failure. Control o f disease vectors, protection o f water and food safety, and health education, preventive health, women and child health care and rural heath are areas where public sector interventions . appear well justified. Establishing better input priorities, for example setting up new facilities o n a well-defined criteria and only in underserved areas, improving the balance between doctors and paramedical staff, and focusing new recruitments only to correct s h l l mix; should increase the efficiency o f scarce . public resources Building on, and learning from, successful initiatives like LHW, preventive and population programs can help lead to better channeling o f public sector efforts and resources. The effectiveness o f these programs may be enhanced by a phased devolution o f management, implementation and financing o f these programs to the district governments. Forging better public-private partnerships to make more effective use o f health facilities and slulls. In this regard, each provincial government should develop a framework for contracting out packages o f services for defined population to NGOs or other private parties individual . entrepreneurs or doctors groups. Making devolution to work for achieving the health outcomes (see Chapter 4). 54 3.39 Most importantly, however, public sector spending in the health sector should increase significantly. At present budgetary allocations for health sector falls well short o f what i s needed to achieve the PRSP targets and MDGs. Although, private sector spending can help close this gap to some extent, the government’s redistributive role requires i t s increased involvement to ensure access o f quality health services t o the poorest and the excluded, through direct intervention or financing health initiatives for the targeted groups through the private sector health delivery system. T h e magnitude o f the government’s spending effort required to achieve the I-PRSP goals, though not yet fully costed, would be considerably larger than that indicated by the I-PRSP estimates. Rough calculations indicate that adequately financing the high-priority programs (like the LHW, Maternal and Child Health, HIV/AIDS, Malaria and TB Control, Health Education, Population Program, human resource development and capacity building, etc.), along with additional finances for public-private partnership programs, enhancing the non-salary budget, and a normal increase in salary would require, at an average, an additional Rs 10 billion per annum. T h i s would increase sectoral financing from the present 0.6% to an average o f 0.8% during 2003-07. Accelerating Water Resource Development 3.40 Agriculture i s a major part o f Pakistan’s economy and one o f the key engines o f overall economic growth. Agriculture today contributes about 25 percent o f GDP, employs nearly 50 percent o f the labor force, and i s responsible directly or indirectly for 60-70 percent o f exports. Within the agnculture sector, crop production (principally wheat, cotton, sugarcane, and rice) accounts for about 59 percent o f value added and livestock about 38 percent. Increases in water resources available for irrigation and expansion o f irrigated area have played a central role in agriculture growth since the 1960s. Agriculture n o w uses an estimated 95 percent o f available water resources, and irrigated land i s responsible for about 80 percent o f agricultural production. Total irrigated area in Pakistan i s about 46 million acres; nearly 97 percent o f this irrigated area i s in the Indus Basin. 3.41 As Table 3.2 shows, water availability at farm head doubled over the period 1960-90 but has remained unchanged since then (hopefully the sharp drop in last two extreme drought years will prove temporary). This vast expansion o f irrigation water supplies was made possible by the large investment in expansion o f the irrigation system and the construction of two major storage siteshydroelectric projects at Mangla and Tarbela in the 1960s and 1970s under the Indus Basin Treaty, as well as by the exploitation o f groundwater resources mainly by private tubewells. I n the 1960s total water and power investments in the public sector exceeded U S 1 . 5 billion, accounting for more than 50 percent o f total public pend ding.^' H a l f o f this amount represented the Indus Basin Replacement works. The long-term investments made in development o f water resources in the 1960s and 1970s sustained economic growth well into the 1980s. Yet the long-term vision of the planners o f that era,39 in which Palustan would further integrate water and power development and engage in highly productive agriculture in which water and other inputs are used very efficiently, has not been realized. Increases in water supplies have tended to extend and prolong traditional low-productivity agriculture. Slow growth in productivity in agriculture i s due to many factors, including insufficient investment in human development and unequal asset ownership. But the l o w level o f public investment in water including drainage and lack o f sufficient progress on institutional arrangements that encourage farmers to pay for water and undertake on-farm development have also acted as major constraints o n the development o f modern agriculture. Public spending in agriculture and water touched a l o w o f 0.8 percent of GDP in 2000-01. Other than the Chashma Right Bank Canal and the Pehur High Level Canal, expansion o f irrigation in the past fifteen years has been limited to small, isolated schemes outside the Indus Basin. More important, the original expectation that a second stage storage on the Indus would be required and built by late 1980s or early 1990s has proved elusive because 38 Parvez Hasan, “Pakistan s Economy at the Crossroads- Past Policies and Present Imperatives,” 1998. 39 See Water and Power Resources o f West Pakistan, a Study in Sector Planning; John Hopkins University Press, 1968. 55 of the political controversy surrounding the Kalabagh D a m and the lack o f timely development of alternatives. Table 3.2: Water Availability, 1960/61-1999/00 Water Availability at Farm Head M A F Irrigated Year Acre-ftl Acre Canal Water Groundwater Total Area M A 1960-61 48.4 4.2 52.5 25.7 2.04 1967-68 56.8 11.7 68.5 30.9 2.22 1977-78 61.6 27.8 89.4 35.1 2.55 1985-87 61.4 40.9 102.3 39.8 2.57 1990-91 65.0 41.0 106.0 41.9 2.53 1999-2000 61.5 41.6 103.1 44.0 2.34 3.42 With the rapid increase in population, the competition for water i s becoming more intense and disruptive among upper and lower riparian provinces and among the agricultural, industrial, domestic, and environmental sectors. Since 1999 water availability in the system has been inadequate to supply the water allocations agreed-to in the 1991 Water Accord owing to drought and inadequate water availability in the major reservoirs as w e l l as canal capacity. Water shortages on the order o f 40-50 percent were experienced during the 2002 rabi and early kharif crop seasons. These shortages created a strong sense o f crisis--they mobilized sector authorities to a degree that other persistent problems, such as deteriorating performance o f infrastructure, l o w productivity, or persistent poverty, did not. 3.43 WAPDA was the first of the key sector organizations to prepare and present a plan--a water resources development strategy document called “Vision 2025” (2001) that calls for huge investments in multipurpose storage reservoirs and several new canals to expand irrigation during the period 2000-25. The goal o f the plan i s to address the problems o f water scarcity, the flow o f flood water to the sea unutilized, safe disposal of drainage effluent t o the sea, tapping o f groundwater, and reduction o f seepage loses in the country’s existing irrigation system. Partly because o f the sense o f urgency created by drought and the political impasse over the Kalabagh Dam, the government approved Vision 2025 in principle, and its main elements were, by and large, included in the Planning Commission’s Ten-Year Perspective Development Plan (TTPP--FY 2001-1 1). Some projects have actually entered the Public Sector Development Program (PSDP) during the past two years. The cost o f federal water projects included in the TTPP during 2003-07 i s estimated at Rs. 202 billion, or roughly Rs. 40 billion per year compared with Rs. 9 billion spent in 200 1-02. 3.44 Because o f great neglect of water investments in the recent past, a substantial acceleration in overall public spending in the water sector appears justified. WAPDA and the Planning Commission must be commended for their initiatives in reviving water sector investments. The thrust o f the present program has several problems, however, that require careful consideration and review by the Govemment o f Pakistan. Many o f the issues involved, the tackling o f which can yield a more optimal and balanced expansion o f irrigation water-related public spending, are discussed at length in Volume 2 o f this report. These issues can be summarized as follows: The need for a clearer vision that integrates long-term agriculture, irrigation water, and hydroelectric development . An excessive emphasis on extending irrigation facilities, some o f whom may prove marginal and all o f which will yield economic benefits only after a number o f years Inadequate attention to the urgent need to rehabilitate existing irrigation assets and on-farm development 56 . Inadequate attention to the political and technical issues that may prevent early resolution o f much-needed additional storage on Indus The need to harmonize ambitious water sector plans with overall resource availability and to avoid preemption o f urgent social investments . The need to balance institutional development needs, including federal-provincial coordination and greater investment in planning and investigation, with higher water sector investment T h e need to identify follow-up work to firm up long-term plans and to seek necessary mobilization o f external assistance 3.45 WAPDA's Vision 2025 and the Medium- Figure 3.6 Public Sector Expnditurer Term Investment Plan (MTIP) o f the Ministry o f Water and Power, completed in October 2002, a "0 year after the TYPP, do not clearly relate the role o f water development either to future agricultural 2 lSaa needs and agncultural potential or to the limits o f s 0 irrigation water availability. Any long-term plan 2 loom for agricultural and water development must come %om to grips with the fact that groundwater development i s already nearly at i t s limit, and surface water resources development will approach i t s overall limit--through increased storage o f flood flows in the Indus River--in the next twenty-five years at the latest. 3.46 Even making the optimistic assumption that two large storage sites on the Indus can be built over the next twenty-five years, the addition to the total water supply in the Indus system would be only about 15 percent, a sharp contrast to the massive additions that were possible during 1960- 90. T h e environmental, political, and economic feasibility o f a second storage, however, remains an open question. O n e o f the concerns, n o w being addressed through environmental studies, i s whether there will be sufficient water below e \ Figure 3.8: Public Sector Expenditures as % of total PSDP K0t1-i.~' The w o n y i s that additions to surface water supplies, without adequate drainage and 35 0 evacuation o f salts, will aggravate the problem o f & 300 8 salt balance in the Indus Basin.41 The finalization 250 B o f a national drainage accord in the near future and f 200 1 0 adequate investments in a national drainage program will help t o mitigate environmental 2 I5O 100 concerns. But the basic fact i s that the past pattern 50 o f agricultural growth, in which insufficient 00 1980/81198~831984/851986/871988/891990/911992/931994/951996/971998/99l000/012002i03 40 The Ministry o f Water and Power and the Indus River System Authority (IRSA) are jointly commissioning a study to determine the flow required below the Kotri barrage for ecologicalienvironmentalneeds and for preventing seawater intrusion. 41 Surface water supplies bring an estimated 33 million tons o f salt each year into the irrigated areas of the Indus plain, and a further 28 million tons are added from groundwater from tubewells. Only a small fraction o f the total salt i s carried out of the basin by the Indus River below Kotri Barrage or the Left Bank Outfall Drain (LBOD) in Sindh. 57 attention was given to water conservation and effectiveness o f water use, i s not sustainable. 3.47 An optimal water strategy must emphasize all thefollowing elements: m Supply expansion: storage expansion (Mangla, Bhasha, andor Kalabagh), small dams, improvements in diversion capacity and function, improvements in overall system operation m System expansion: external expansion (three canals), internal expansion by remodeling and extension o f existing Canal Command Areas (CCAs). 4 Management: on-farm water management interventions (watercourse lining and land leveling, for example), supply management (rehabilitation and modernization, flood protection, improvements in operations and maintenance versus new investments), improved water allocations and scheduling, improved equity at system and distribution levels, groundwater management, incentives, introduction o f new agriculture and irrigation technology, effective access to markets, improved knowledge and information, human resource development and research 4 Sustainability and environment: waterlogging and salinity, reducing drainable surplus, water quality, safeguarding ecology below the Kotri Barrage 4 Productivity: cropping patterns, water rights and markets, water pricing 4 Governance and institutional reform: improved public service and accountability, cost reduction, decentralization, empowerment o f users 3.48 The optimal water strategy for Pakistan's integrated Indus Basin System i s technically complex. The trade-offs-among the set o f alternatives and complements (storage, conservation, reservoir scheduling policy, estuary maintenance policy, sSystem expansion) are not simple because o f the varying degrees o f capital lumpiness, the uncertainty o f costs and benefits, and the dependence o f benefits and costs o n the sequencing o f investments. In addition to the technical complexity, the integrated system includes a cascade o f upper and lower riparians, which introduces political economy considerations in the choice o f the optimal strategy. T o address this technical-cum-political complexity, i t i s essential that provincial governments (lower and upper riparians) as well as intraprovincial stakeholders (system rehabilitation versus system expansion) engage in extensive dialogue to agree o n the fkamework and the rules for sharing the resource and the costs and benefits. T h i s framework may be formalized as water and drainage accords to permit investment choice and sequencing. 3.49 On storage, the decision to raise Mangla D a m appears sound because the increased storage capacity will have a reasonable probability (71 percent) o f filling each year. Also, it i s the only option available to the government in the next five to seven years. 3.50 The decision to accelerate the feasibility o f Basha D a m i s also an important step in resolving the political impasse on Kalabagh. M u c h more work needs to be done, however, during the next year t o deal with the political economy and technical questions before Pakistan i s ready to propose a large storage project to donors for financial support. Political consensus needs to be built on the sequencing o f storage projects, assisted by good economic studies that will throw light on the options available for operation o f new reservoirs. The technical choices between water releases for the agriculture sector and electricity generation and the management options for releases in rabi versus early kharif must be debated and settled as a part o f the decision to build additional storage. This dialogue will, on the one hand, clarify the economic and political rationales for future storage and, o n the other hand, help minimize future conflicts among provinces and in canal commands within provinces. 58 (2016) Onginal Storage Capacity I" - (Tarbela + Mangla + Chashma) L I 12 8 G a E .- 10 4 b Mangla Raising (2008) m P 8 I 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012 2016 2020 2024 Years 3.5 1 The government has initiated a large system expansion program focused o n three canals: Greater Thal, Kachhi, and Rainee canals. These canals will enable diversion o f 4.6 MAF o f water and will cost more than Rs. 90 billion. These projects are driven primarily by the desire o f the provinces t o utilize their respective shares under the Water Accord, though until more storage i s built they will receive perennial supplies for only seventy-five to ninety days during the kharif season. While these projects respond to strong regional interests, the final scale, phasing, and cost coverage o f these projects must be reexamined, especially after the feasibility studies of Kachhi and Rainee canals have been completed. The trade-offs among provinces and stakeholders within provinces must be revisited and agreed-on as part o f the new drainage and water accords mentioned. The conflicting political economy considerations underlying these choices cannot be minimized, but with the help o f technical feasibility studies, consensus should be possible o n the broader national interest. 3.52 The following further points need t o be taken into consideration before fully committing the . funds to the program o f canal expansion during 2004-07: Since these canals would receive water for a period o f only seventy-five to ninety days in a year . (during July to September), the non-availability o f water during April-June (the kharif sowing season) would affect cropping patterns and yields. The command areas include depressions, sand dunes, and large areas with l o w population. Thus . water losses would be high, and agricultural development would take place very slowly as colonization and command area development take place. As currently proposed, the projects do not provide funds for the government's share in command area development, and without this development project benefits would remain elusive. But adding these costs, whether they are borne by the government or the farmers, will sharply lower . the economic rate o f return. Detailed social and environmental assessments should be carried out to determine canal alignments that would explicitly target the poorest areas and small landholders. 59 m The projects would add a substantial burden to already strained operation and maintenance budgets. Full agreement o f the provinces that needed requirements will be funded must be obtained before committing large funds to these projects. 3.53 The Ministry of Water and Power i s considering a “modular approach” for implementing these projects in a phased manner. This approach may move toward satisfying both the political commitments and the TYPP objectives, as well as providing a highly visible demonstration o f the new future o f irrigated agriculture in the Indus Basin. Without prejudging the completion o f feasibility studies and detailed appraisals, we believe these projects also provide an exceptional opportunity to establish a decentralized and farmer-managed system right from the start. Instead o f expanding the existing imgation departments into the new command areas with their traditional modes o f operation, maintenance, and implementation, the government should make concerted efforts to establish Farmers’ Organizations (FOs) through a program of social mobilization and capacity building. This step should be followed by the establishment o f area water boards. Regardless o f their ultimate benefits, the major canals will yield economic benefits only after several years. Short-term benefits will remain elusive. These long-term programs must be strongly supplemented with programs and projects that will yield immediate benefits. 3.54 Various estimates put the potential savings from the current supply at about 4.7 MAF, just under 5 percent o f the current average availability o f surface water in the Indus Basin. These savings would come primarily from lining watercourses and leveling land under the on-farm water management projects o f the provinces. Water conservation and efficiency gains are achievable provided that farmers benefit from the savings--not only because equity would improve and previously unserved farmers would get water, but also because all farmers in the improved canal command would be able t o decide h o w to use the saved water to increase their production and income. A substantially higher allocation for operation and maintenance and fundamental changes in the operation and maintenance regimes are also urgently needed. Current estimates o f system rehabilitation (essentially accumulated deferred maintenance) and remodeling costs exceed Rs. 200 billion.42 Although the ongoing institutional reforms would ultimately address issues o f overstaffing and lack of farmer participation, immediate action i s needed to stem the . deterioration o f the irrigation and drainage system. A well-designed program would: . Establish realistic finding requirements through an independent evaluation; Rationalize fund allocation to meet priority needs; Initiate benchmarking and operation and maintenance performance audits: and Rationalize abiana rates and improve assessment and recovery (empowering farmer organizations not only to operate and maintain their canals, but also to collect abiana and retain a sufficient portion o f the recovered funds to meet their needs would be a key incentive for increased cost recovery). 3.55 Higher allocations for maintenance of existing assets and speedy on-farm development must be underpinned by reforms in the system of governance. Based on the experience gained so far, a new consensus on the way forward within the provinces, and between the provinces and the federal government, i s needed that would essentially re-launch the water management program. This new consensus should focus o n critical steps to be taken and o n piloting and monitoring a variety o f approaches within each pilot canal command over a definite time period. Business as usual would prove costly in terms o f both forgone benefits of participatory irrigation management and further deterioration o f the irrigation system. The lessons leamed from the ongoing reform efforts are that these fundamental reforms require champions at the highest level o f federal and provincial governments and should not be left to the discretion o f the agencies being reformed. 42 PER consultant’s estimate. 60 3 -56 A balanced program o f institutional development, higher investments in water, and appropriate incentives to conserve water and use it more efficiently i s well within Pakistan’s reach. The substantially larger additional funding for the water sector during 2004-07 can be justified provided the program gives due weight to all key components and strikes a balance among short-term, medium-term, and long-term investment and institutional needs. An indicative program o n these lines totaling Rs. 252 billion during 2003-07 i s presented in Volume 2. The suggested program provides for ongoing programs (notably Right Bank Outfall Drain I1 (RBOD2) and National Drainage Program (NDP)), as well as a number o f new projects (Greater Thal Canal, Gomal Zam and Mirani Dams, Mangla). It also allows for substantial acceleration o f nondevelopment and development budgets o f the provinces. I t makes provision for the government share o f command area development for the new canals, new drainage initiatives including institutional reforms, and on-farm development. 3.57 The balanced program that i s being proposed can help initiate the transformation o f the water and agriculture sector even in the medium term. The more than doubling o f the real spending in water over FYs 2002-07 will take about 25 percent o f the additional fiscal space that i s expected to be created under the baseline scenario and will sharply reverse the declining trend in water sector spending. The acceleration will be even faster than proposed in the education sector. Needless to say, such large spending carries risks and could waste resources i f it i s not accompanied by much tighter controls o n program and project priorities and stronger implementation and institutional arrangements than exist at present. A fundamental strengthening o f the sector’s planning, research and development, implementation, and monitoring mechanisms, including much greater interprovincial and federal- provincial coordination, i s also necessary to facilitate decision malung on the more difficult longer-term issues. Transforming the Power Sector 3.58 On the surface it may appear that the power sector investment program will not have major claims o n public resources during 2003-07. The Planning Commission’s TYF’P suggests an allocation in the Public Sector Development Program for the power sector o f only Rs. 90 billion for the next five years, or less than 0.4 percent o f GDP-lower than in the recent past. Comfort i s drawn from the fact that at present there i s surplus power in the country, and with the commissioning o f the large Ghazi Baroda hydroelectric project by the end o f 2004, there will be a substantial addition to capacity. The coming o n stream o f the Ghazi Baroda project will n o doubt be a milestone in the history o f power development in Palustan. It will substantially improve the hydrothermal balance in electricity generation and help t o reduce overall costs. After 2003-04 i t s completion will provide r o o m for other projects in the power sector: PSDP allocations for Ghazi Baroda for FY 2002-04 alone are estimated at Rs. 37 billion. 3.59 In reality, however, there are many unresolved issues in the power sector, which could both complicate the creation o f fiscal space and seriously hamper larger allocations to high-priority sectors. These issues center o n the financial weaknesses in the two state-owned power entities, WAPDA and . KESC. These weaknesses stem from: . Continued large losses in the power system, reflecting a combination o f factors (low level of . efficiency, theft o f electricity, and inadequate investments in power distribution). The high cost o f power from independent power producers (PPs). . . Increasing consumer resistance and political sensitivity to higher electricity tariffs. A high level o f receivables from public sector consumers. An implicit large subsidy to consumers in the Federally Administered Tribal Area (FATA). 3.60 M o r e recently, the financial situation o f WAPDA and K E S C (notably o f KESC) has deteriorated, which has further eroded the entities’ capacity to finance their investment programs. A number o f factors 61 account for this deterioration. Although tariffs have been raised, power purchase and fuel costs also increased substantially, and these increases were not fully passed on to consumers. Targets for reduction in technical and nontechnical losses were not achieved. And the level o f receivables--notably from public sector consumers--remained high. 3.61 Recognizing these challenges, the government has established a reform agenda for the power sector driven by the following key objectives: Mobilize private resources for the development o f the sector, without recourse to the government budget, through the March 1994 Independent Power Producer (IPP) Policy. Improve the utilities’ performance and efficiency through corporatizatiodcommercialization (in preparation for their eventual privatization), enhanced managerial autonomy and accountability, competition, and profit incentives. Introduce a transparent regulatory fiamework (particularly in setting electricity tariffs) to provide comfort to potential investors, while enabling the government to maintain certain socially desirable activities such as “lifeline rates” for poor consumers and rural electrification programs. 3.62 Considerable progress was achieved in the mid-1 990s in mobilizing private resources, through IPPs, for new generation projects. The K o t Addu power plant was privatized in 1996, through the sale o f 36 percent o f shares (along with management control) to a strategic investor. The TPP projects, however, were solicited under fixed-price (or fixed-return), long-term contracts, and the government’s main focus was o n the early commissioning of the projects to overcome the supply shortages being experienced at that time. In the process, too little attention was given to the demand side issues. Moreover, the impact o f these investments o n the overall financial viability o f the sector, including the ability o f WAPDA and K E S C t o pay for power purchases from IPPs, was not adequately analyzed. Although shortfalls in generating capacity have been overcome, there i s actually some excess capacity in the system because economic growth has been unexpectedly slow. Consequently, the power sector i s facing a financial crisis: K E S C does not cover even i t s operating costs with the existing tariffs, and WAPDA i s unable to mobilize adequate funding, through internal cash generation, for i t s investment program.43 While WAPDA has met the internal cash generation target in some recent years, it has done so mainly by deferring critical investments, defaulting o n debt service obligations t o the government, or building up large arrears to fuel and equipment suppliers and contractors. 3.63 Meanwhile, actions have been taken to unbundle WAPDA and to establish the regulatory framework. WAPDA’s Power Wing was reorganized into eight distribution companies (discos; a ninth disco covering the tribal areas in NWFP was recently established); one transmission and dispatch company (NTDC); and three thermal generation companies (gencos). WAPDA i s n o w responsible solely for developing and managing the hydropower generation facilities. Assets and manpower have been allocated to these entities, and most operational functions (including the management o f investment programs) have been delegated. NEPRA has notified i t s main rules (for tariff setting, license applications, etc). It held hearings o n tariff and license applications, which are open to the public, and issued a number o f tariff determinations. A formula for passing the impact o f changes in fuel costs to consumers, through quarterly tariff adjustments, has been approved and i s starting to be implemented. Licenses were issued to all Discos and Gencos by the end o f June 2002 and to NTDC by the end o f December 2002. 43 T h e financial performance targets for WAPDA and KESC, which were agreed-to with the government and the entities under ongoing (and past) World Bank and Asian Development Bank projects, included WAPDA’s financing o f at least 40 percent of i t s investment program through intemal cash generation; the corresponding figure for KESC was 25 percent. 62 3.64 Although the government has taken a number o f steps t o promote private investment in the power sector, the desired goal of ensuring the development o f the sector without recourse to the government budget, except for hydroelectric development and possibly transmission, i s nowhere in sight. The continuing financial distress in the KESC, notwithstanding a substantial injection o f public monies to fund past losses, and international conditions make it highly unlikely that K E S C can be privatized in the near future. It also appears somewhat unlikely that the newly corporatized distribution and generating companies under WAPDA will actually be privatized on a timetable to fund power sector investments during 2003-07. 3.65 Consequently, there i s a disconnect between large national power sector requirements and the amounts that can realistically be expected from the development budget and own resources o f WAPDA. Whereas the TYPP has notionally allocated Rs. 90 billion for WAPDA mainly from external loans, WAPDA’s proposed program for 2003-07, including rural electrification, i s Rs. 261 billion. So there will be an additional requirement, after providing for WAPDA’s own resources, o f Rs. 90 billion to be met from the budget and/or borrowing in the market with a government guarantee. KESC investment needs will be over and above this. The extra requirement for the power sector investment could wipe out one- third of the fiscal space that i s supposed to be available under the baseline scenario. 3.66 What can be done? What should be done? Although the system o f automatic adjustments in electricity tariffs t o reflect costs must be further strengthened, there may not be much scope for increasing real as distinct from nominal, electricity tariffs44. As Table 3.3 shows, the average residential tariff increased by 50 percent over 1996-2002 and i s currently about 6 U.S. cents per Kwh. The industry and commercial tariffs have been difficult to increase because they are already high and adversely affect competitiveness. 3.67 Clearly a much more detailed plan for the power sector for the period 2004-08 urgently needs to be developed and harmonized with the development o f rolling three-year plans and the MTBF. In the absence o f a plan, the power sector could derail the government’s economic and social priorities. Some . elements o f the power sector review should be the following: . . A realistic but proactive timetable for privatization o f corporate entities in the sector . Careful review o f the present plans o f WAPDA, especially for generation Reinforced efforts to reduce technical losses . Solutions for managing load shedding in January when l o w level o f hydroelectric generation will reduce supply . Separation o f rural electrification from normal WAPDA funding and inclusion o f rural . electrification as a part o f the poverty reduction strategy The formal launching o f the new policy for power generation for the private sector Innovative approaches to raise capital for the power sector ahead o f privatization 44 According to recent World Bank estimates, 10-15 percent increase in nominal tariffs over the next few years will be inescapable even with aggressive cost reduction and speedy progress towards reduction o f theft. 63 Table 3.3: Electricity Tariffs (WAPDA System) - FY96-2003 /Paisas/kWh’h) FY96 FY97 FY98 FY99 FYOO FYOl FY02 FY03 Domestic 146 117 223 247 243 259 318 352 Commercial 534 615 702 698 698 704 709 729 Industry 335 391 444 417 417 416 419 442 Bulk supply 275 355 419 437 437 424 448 483 Agriculture 161 192 206 344 281 258 294 345 SYSTEMAverage 223.9 269.6 319 348 339 337 372.7 405.4 Electricity Tariffs in real terms (WAPDA System) -- FY96-2002 FY96 FY97 FY98 FY99 FYOO FYOl FY02 Domestic 146.0 103.1 182.6 191.1 182.9 184.5 216.5 Commercial 534.0 542.1 574.8 539.9 525.3 501.6 482.7 Industry 335.0 344.7 363.5 322.6 313.8 296.4 285.3 Bulk supply 275.0 312.9 343.1 338.0 328.9 302.1 305.0 Agriculture 161.0 169.2 168.7 266.1 211.5 183.8 200.2 SYSTEM Average 223.9 237.7 261.2 269.2 255.1 240.1 253.8 GDP Deflator (1995-96=100) 100.0 113.4 122.1 129.3 132.9 140.4 146.9 Electricity Tariffs (WAPDA System) -- FY96-2003 (in Usckwh) FY96 FY97 FY98 FY99 FYOO FYOl FY02 FY03 Domestic 4.35 3 .OO 5.16 5.28 4.69 4.59 5.44 6.02 Commercial 15.91 15.77 16.25 14.92 13.48 12.48 12.13 12.47 Industry 9.98 10.03 10.28 8.91 8.05 7.37 7.17 7.56 Bulk supply 8.19 9.10 9.70 9.34 8.44 7.51 7.66 8.26 Agriculture 4.80 4.92 4.77 7.35 5.43 4.57 5.03 5.90 SYSTEM Average 6.67 6.91 7.38 7.44 6.55 5.97 6.38 6.94 Exchange Rate (Rs/US$) 33.568 38.994 43.196 46.79 51.771 56.43 58.457 58.457 Conclusion 3.68 The financing o f the poverty reduction strategy should be the first charge o n the additional fiscal resources that are likely to become available to the Government o f Palustan. High-priority poverty spending should be defined more broadly to include l a w and order maintenance, access t o justice, and rural electrification. A large part o f higher spending under the PRSP would also be on education and health. Without adequate funding (and stronger delivery systems), social sector goals will again be frustrated. There i s also a definite need to accelerate spending o n water resource development. But the present government program appears lopsided and has not been carefully thought through. A balanced program o f increase in storage, system expansion, rehabilitation o f existing assets, on-farm development, and concomitant institutional development will yield far better economic and social results. The persisting 64 power sector problems remain a threat not only to government finances, but also to an orderly reorientation o f public sector priorities. These problems need to be analyzed and resolved with the help o f appropriate sector work and policy advice. 3.69 Although there could be several ways in which the additional fiscal space could be allocated, in light o f the government’s priorities and commitments, Table 3.4 illustrates one such way o f allocating additional resources to the high-priority sectors. T o meet the government’s PRSP targets, substantial portion o f the additional fiscal space has been assigned to the poverty reduction sectors and programs, with high-priority given to human development, pro-poor infrastructure and improving the access o f the poor to justice. In the illustrative table, this i s reflected by increasing the pro-poor expenditure by an additional 2.2 percent of GDP by 2006/07. A s for other expenditures, the government’s resolve to correct the past neglect o f the water resources i s reflected by showing substantial increase in the short-, medium- and long-term investments in the irrigation sector. Increases in other infrastructure investments are mainly to improve the efficiency o f port operations, railways, road and power sectors. A considerably higher allocation for general administration i s mainly to cater for higher salaries and capacity building needs o f the civil servants, especially at the district level. Table 3.4: Illustrative Allocation o f Additional Fiscal Space (Percent of GDP) Real Spending (Rs. billion) “ Percent 001102 2006/07 Increasr !001/02 2006107 Increase ’riority poverty related spending 4.8 7.1 2.21 175.5 336.6 91.8 Education 1.8 2.6 0.73 66.0 121.8 84.4 Health and population planning 0.6 0.9 0.29 20.4 40.6 98.6 Law and order 1.0 1.2 0.27 34.5 58.3 68.7 Rural electrification 0.1 0.3 0.15 3.5 11.9 241.1 Water supply and sanitation 0.1 0.2 0.13 2.3 9.1 293.5 Rural development 0.3 0.5 0.17 12.3 24.4 97.6 Irrigation 0.3 0.5 0.22 10.4 24.4 134.6 Roads, highways and bridges 0.4 0.5 0.16 12.8 24.4 90.8 Other ” 0.4 0.5 0.10 13.2 22.0 66.4 Ither non-interest non-defense spending 6.1 7.3 1.21 220.1 347.3 57.8 General administration and pensions 2.5 2.8 0.30 89.3 131.8 47.6 Large scale water development 0.1 0.6 0.48 3.6 27.7 665.0 Railways, ports, national roads and power 0.8 1.0 0.27 27.3 48.7 78.4 Other 2.8 2.9 0.16 99.9 139.2 39.4 rota1 10.9 14.3 3.42 395.5 684.0 72.9 ’ Constant 2001/02 prices. 2/Social security and social welfare, food subsidies, food support program, natural calamities and other disasters, and land reclamation 65 CHAPTER 4: IMPROVING EFFECTIVENESS OF PUBLIC EXPENDITURES Introduction 4.1. Delivery o f most public services in Pakistan, including education, health, water and sanitation and l a w and order, i s poor after several decades o f secular decline. This decline has been only partly the result of inadequate spending on essential public services in the face o f a very high rate o f population growth. I t has also been the consequence o f a lack o f well-functioning institutions and erosion over time o f rules, procedures, and organizational arrangements that ensure the effectiveness o f public expenditures. Poor budgetary practices, lack o f accountability, inadequate management and organizational capacity, and poor governance have all contributed to poor outcomes. The issues o f civil service reform, which are at the heart o f institutional strengthening, are discussed in the next chapter. This chapter focuses on budgetary and planning processes, monitoring mechanisms, the incentive structures for implementing government programs and policies, and the recent devolution decisions, which all will have to play an important role in improving the delivery o f public services, especially in the social sectors. It outlines the important steps the government has taken to improve the effectiveness o f spending and the tasks that remain in implementing the reform agenda. It also pursues the issues o f a further redefinition o f the role o f the state, and public-private partnerships in financing, production, and distribution o f a number o f public services because the public sector capacity problems cannot be resolved quickly. Responsiveness to Citizens 4.2. Before discussing the mechanics o f better and more efficient public service delivery, it i s important to stress that the ultimate test o f effective service delivery i s whether i t meets the expectations of citizens, especially the poor. The usual yardstick used by the government to measure the effectiveness of these services i s the amount o f money spent o n these services and o n creation o f physical assets. Little, if any, attempt has been made to monitor and analyze whether these services are delivered efficiently and whether their provision meets citizens’ expectations. In order to determine the level o f consumer satisfaction with a number o f key public services, the National Reconstruction Bureau (NRB) engaged an international NGO-the Community Information, Empowerment and Transparency (CIET) to undertake a consumer satisfaction survey in all the districts on in the country. The survey attempts to measure the extent to which users are satisfied with these services, aspects o f service that were not effective, and groups o f people most affected by the deficiency in service. This study also provides a database o f information and sets service standards for analysis in future. 4.3. Being a measure only o f people’s perceptions, the results o f the CIET survey need t o be interpreted with utmost care. The results, however, do reveal a strong dissatisfaction within the general public with the quantity, and more important the quality, o f public services (especially education, health, water supply and sanitation, and roads), which clearly demonstrates the l o w effectiveness o f public expenditure. Reaching the Poor 4.4. Implementing the strategy highlighted in the government’s I-PRSP i s important for alleviating poverty. It however, needs to be recognized that I-PRSP was prepared against the background of strict fiscal stringency. With more fiscal space likely to be available in future, the poverty reduction strategy can be funded more adequately and the approach to poverty reduction can, as discussed in chapter 3, be broadened. However, there remains the need, within the broad framework o f poverty reduction, to design programs specifically targeting the poor. 4.5. The W o r l d Bank’s Poverty Assessment4’ has shown access o f the poor to the basic public services i s low. For example, net primary enrollment rates are 37 percent for the poor, as compared to 59 percent for the non-poor; the incidence o f medical consultation for diarrhea among children i s 79 percent for the poor compared t o 84 percent for the non-poor; only 15 percent o f married women o f ages 15 to 49 in the poorest households have ever used contraceptives, compared to 25 percent o f those in the richest households; as many as 24 percent of the poor rely on potentially unsafe sources for drinlung water, compared to 19 percent of the non-poor; about 52 percent of the poor live in households connected to electricity, compared to 76 percent o f non-poor households. 4.6. A determined effort i s therefore required to focus these services toward the poor through enhanced outreach and better targeting o f these services. I t should, however, be kept in view that in Pakistan, supply side bottlenecks are only one side o f the problem o f limited access o f the poor to public services. The opportunity and actual costs, social norms and cultural traits all contribute towards less than full utilization o f these services by the poor. But research has also shown that improving the quality o f services can effectively offset most of the demand side obstacles. Institutional and governance reforms and adequate allocations for key non-salary expenditure items, like textbooks, class material and medicines would be required to enhance the quality of education and health services. Appropriately designed incentive schemes, like scholarships, edible o i l and school lunches (as planned under Taiwana Palustan Program) can thus help alleviate some o f the demand side problems. 4.7. Apart from developing programs in education and health that specifically target the poor, it i s also highly desirable that the government aligns the planned large water sector public investments with i t s poverty reduction strategy. The phasing o f these investments should be determined by the extent to which the poor can benefit from these projects. Also, the land development aspect o f the Thal, Rainee and Katchi Canals require a strong pro-poor focus. Most o f the additional arable land that may become available as a result o f these projects should be distributed among the rural poor. Improving Budgetary Processes 4.8. Despite some recent improvements, the budget processes continue t o suffer from a number o f weaknesses that reduce the efficiency of the budget and therefore the effectiveness o f public expenditure. The main weaknesses o f the process include (1) lack o f strategic focus, and a lack o f clear articulation and ownership o f sectoral priorities; (2) lack of information o n the costs of policies, programs, and services and an intertwining in a single agency o f policy advice, regulation, and service delivery; (3) a primary focus on inputs with performance judged largely by h o w closely spending matched budget appropriations; (4) a short-term horizon for budget decision m a h n g that fails to account for longer-term costs and benefits; (5) an artificial and unwarranted separation o f development and recurrent budgets, making it impossible to prepare an integrated and coherent strategic budget framework; and (6) few incentives for agencies to save budgetary resources because current year spending i s the starting point for the next year’s allocations. 4.9. In the past, the budgetary practices suffered also because the planning and programming mechanisms were unable to withstand the political pressures that resulted in a number o f economically unviable projects being included into the development program (and budget).. As many o f these were large projects they not only pre-empted a sizable portion o f available development funds but also imposed an unnecessary burden on the federal or provincial recurrent budgets (see para [4.77] for some examples o f these projects). Most of these projects were started by ignoring or short-circuiting the regular approval process, with some projects getting included in the development program without even a proper feasibility study. Removal o f key infrastructure sectors (energy, highways, telecommunication, etc.) from 45 Poverty in Pakistan: Vulnerabilities, Social Gaps, and Rural Dynamics, The World Bank, October 2002. 67 the purview o f the Planning Commission’s approval process was one important reason leading to inclusion o f these projects into the development program. The decision o f the government to restore Planning Commission’s authority over project approval process for all sectors i s a step in the right direction. Fiscal prudence and adherence to the established processes and procedures in approving development projects also need to be ensured. In this context, it i s advisable that the government undertake a through review o f all large new projects which are presently in the program. In future, n o project should be undertaken without proper feasibility and need assessment. 4.10. The problems with the budgetary processes were less apparent in earlier years, when the overall levels o f public spending were growing rapidly, planning processes were better integrated with economic decision malung, and the governance problems and the weaknesses in the civil service had not yet assumed alarming proportions. During the past decade or so, the weakness o f the overall fiscal situation and stagnant or declining economic and social spending exacerbated the structural rigidities in the budget process and compromised the allocative efficiency o f expenditure. Arbitrary expenditure cuts undermined program- and agency-level performance, operation and maintenance expenditures were severely under-funded, and development resources were spread too thinly as the overhang o f ongoing projects became very large. Recent Initiatives 4.11. Faced with particular weaknesses in the budget processes and general problems o f effectiveness o f public spending, the government has taken a number o f initiatives to improve planning and budgeting, reduce leakage, and target public spending o n key services and the poorer groups in the society. I t has moved strongly to strengthen fiduciary and accounting controls and improve the workmg of Public Account Committees (see B o x 4.1). It has a set up a Debt Policy Coordination Office to strengthen the macroeconomic work in the Ministry o f Finance, integrate debt reduction goals with the medium-term economic framework, and monitor the implementation o f the government’s Debt Reduction and Management Strategy. I t has also initiated work toward a more comprehensive and detailed Medium- Term Budget Framework (MTBF) with the help o f technical assistance from the United Kingdom. It has made arrangements for estimating the considerable fiscal contingency obligations with respect to civil and military pensions. Last but not least, the government has adopted a poverty reduction strategy with a clear focus o n direct poverty reduction expenditures and initiated the processes o f costing o f programs and monitoring o f both spending and outcomes. Medium-Term Budgetary Framework 4.12. The move toward a comprehensive and detailed MTBF i s especially significant because an appropriately designed MTBF can bring the various initiatives for better fiscal planning together and help remove many o f the problems in the budgeting and programming process. 68 Box 4.1: Budgetary and Financial Management Reforms In recent years Pakistan has undertaken a series o f reforms to improve the effectiveness o f public expenditure through improvements in its budget management and financial systems. T h e main elements of this reform include the following: Devolution: The devolution plan announced by the govemment aims at strengthening the local govemments. By taking the government closer to the stakeholders, this move i s a major step to enhance the accountability o f public service management, and thereby improve the efficiency of the govemment. Given the low institutional capacity of the govemment, however, the reform i s not free of risks. The devolution process, if not appropriately handled, may lead to marginalizing the status and role of provincial govemments; elite capture of local govemments, breaking down o f subnational personnel management, and consequently of the service delivery systems. Although with some expected teething problems, to date the reform seems to be working quite adequately. Budgetary reforms: The government has launched an ambitious reform program to address the numerous budgeting problems. The most important action in this regard i s the initiation of a Medium-Term Budget Framework (MTBF). In addition, accepting the recommendation of the Debt Committee, the govemment has established the Debt Policy Coordination Office, which under the Draft Fiscal Responsibility and Debt Limitation Law would be responsiblefor framing the overall macroeconomic framework, a function relatedto the MTBF. In order to improve budget implementation, significant financial autonomy and responsibility has been devolved to the line ministries through introduction of a New System o f Financial Control and Budgeting (NSFCB), which delegates significant financial powers to the administrative heads of spending agencies. T h i s step has been accompanied by the imposition of hard budget constraints on the line ministries and divisions through expenditure ceilings. At the same time, the spending agencies have also been made responsible for ensuring fiscal and financial discipline, including timely reconciliation of financial data. At the provincial level, the provinces o f Sindh and NWFP, under their economic and structural reform programs have ensured better implementation of provincial budgets through timely release o f funds and removal of redundant procedural constraints. A better system of monitoring and reporting o f budget implementationhas been put in place. Moreover, the use of personal ledger accounts (PLAs) has been severely restricted.Fisca1 transparency: The govemment has announced a plan to improve the quantity and quality of fiscal information disseminated to the public to ensure fiscal transparency, and thereby strengthen financial accountability. The govemment has endorsed a review of fiscal reporting and transparency under the IMF’s Code o f Good Practices on Fiscal Transparency. Consolidated fiscal accounts of the federal and provincial govemments (on the basis of reconciled data) are posted on Finance Ministry’s web site on a quarterly basis. The government of Sindh has started to post even its fiscal monitoring report, which analyzes the implementation of the budget and identifies key problem areas of the provincial government, on its web site every quarter. Financial accounting: The govemment has made significant progress in establishing a n integrated financial management system, which can meet intemationally accepted standards of completeness, accuracy, reliability, timeliness, and usefulness. The medium-term objective i s to accelerate the adoption o f the new public chart of accounts and financial reporting requirements prescribed by the president in December 2000, all under a new computerized accounting system supported by the Bank-fundedProject for Improvement in Fiscal Reporting and Auditing (PIFRA). The govemment seeks to establish the public accounting system a s a robust platform for transparency in the management of public funds for: (a) economic decision making within the govemment to promote program efficiency and effectiveness; and (b) provision of auditable financial information to various stakeholders within the country as well as to donors and intemational financiers. Auditing and legislative oversight. The federal and provincial Public Accounts Committees (PACs) have been made more effective. For the first time in many years, they reviewed the financial transactions o f an incumbent govemment and were also able to significantly reduce the backlog of audit reports. T h e quality o f audits has been improved through systems-based auditing procedures and attention to reporting only on significant matters. And transparency of the PAC process has been enhanced by making PACs’ proceedings open to the public. Expenditure monitoring: The evolving PRSP includes a strong program o f monitoring, including capacity development to gather and analyze information for impact assessment of public policies and investments. The I-PRSP includes institutional mechanisms for the regular/monitoring and evaluation o f poverty reduction expenditures and corresponding outcome indicators at the federal, provincial, and district levels (see 4.28-4.35). These institutional arrangements are presently evolving and will be finalized after the round of consultations with provincial govemments i s completed. Key outcome indicators have also been suggested, and final benchmarks and targets will be set after consultations with national and provincial statistical agencies, federal line departments, and provincial governments. 69 4.13. The rationale for an MTBF i s that it provides the “linlung framework” that allows expenditures to be “driven by policy priorities and disciplined by budget r e a l i t i e ~ . ”It ~~ tries to reduce the disconnect between policy malung, planning, and budgeting. MTBF “consists o f a top-down resource envelope, a bottom-up estimation o f the current and medium-term costs o f existing policy and, ultimately, the matching of these costs with available resources.. .in the context o f the annual budget pro~ess.”~’The “top-down resource envelope” i s fundamentally a set o f macroeconomic projections that indicate fiscal targets and estimates o f revenues and expenditures, including government contingency obligations. To complement the macroeconomic projections, the sectors engage in “bottom-up” reviews that begin by scrutinizing sector policies and activities with an eye toward optimizing inter- and intra-sectoral allocations. The real value added o f the MTBF approach comes fi-om integrating the top-down resource envelope with the bottom-up sector programs. (The steps required in preparation o f an MTBF are given in B o x 4.2.) Box 4.2: Steps in Preparation o f an MTBF A typical MTBF would entail the following steps: +Initiating a process o f rigorous costing and analysis o f existing policy commitments, with a view to improving efficiency and developing a f m baseline cost for essential continuing services of government. Maintaining a continuous medium-term view o f these elements provides an essential foundation for a full MTBF. +Developing a three-year macroeconomic framework as a starting point for the MTBF. +Utilizing improved techniques for macroeconomic analysis and revenue forecasting, and publishing the basis and assumptions for medium-term macroeconomic and revenue forecasts. +Establishing expenditure review teams with the task o f reviewing all o f the government’s existing commitments (Part A o f the budget) through a series o f intensive studies o f different sectors. The overall objective will be to establish a baseline cost for the continuing essential services of government. Once established, this baseline can be projected forward and published as forward estimates in the MTBF framework. +Developing broad aggregates for revenues and sectoral (and, for important subsectors, subsectoral) expenditure ceilings. +Making all ministries and departments responsible for extending their permanent and temporary budget forecasts beyond the budget year by another two years, and separately identifying additional costs over the three-year period that arise from implementation o f government policies and development projects. *Strengthening capacity for sectoral and expenditure policy assessments. These assessments should take into account relevant performance information. QMaking the budget for the following year in a ”rolling” format. That is, in the following year’s budget, the existing policy estimate for the f i r s t outyear w i l l be taken as the baseline estimate for the ministry in negotiating the budget, and similar procedures will be followed each year. 4.14. The MTBF i s intended to deliver a number o f ambitious outcomes: greater macroeconomic balance; improved inter- and intra-sectoral resource allocation; greater budgetary predictability for line ministries; and more efficient use o f public monies. An improved macroeconomic balance, including fiscal discipline, i s attained through good estimates o f the available resource envelope, which are then used to make budgets that fit squarely in the envelope. M T B F s are expected to improve inter- and intra- sectoral resource allocation by effectively prioritizing a l l expenditures (on the basis o f the government’s socioeconomic program) and dedicating resources only to the most important ones. Greater budgetary predictability i s expected as a result o f commitment to more credible sectoral budget ceilings. The MTBF 4 46WorldBank (1998a: 32). T h e World Bank uses the term Medium-Term Expenditure Framework (MTEF) rather than MTBF. 47 For more on the MTEF concept, see World Bank (1998a), Asian Development Bank (1999), and Dean (1997). 70 i s also expected to make public expenditures more efficient and effective, essentially by allowing line ministries greater flexibility in managing their budgets in the context o f hard budget constraints and agreed upon policies and programs. 4.15. While the three pillars o f an MTBF are clear (projections o f the aggregate resource envelope, cost estimates o f sector programs, and a political-administrative process that integrates the two), the blueprint for designing and implementing an MTBF must be developed in each country context. 4.16. I n Palustan several planning and budget process issues will need to be revisited and resolved before the MTBF can become a truly operational document. The relative responsibilities o f federal and provincial governments, the sector ministries and departments, the Planning Commission and provincial planning departments, and the Ministry o f Finance and provincial finance departments in relation to various elements o f an MTBF estimation o f the resource envelope, the costing o f sector programs, the choice o f intersectoral priorities, and the adoption o f specific targets must be looked at afresh and agreed upon because they are in a state o f flux. The arrangements for putting together an MTBF will also need to be harmonized with the preparation and implementation o f the final Poverty Reduction Strategy Paper (PRSP). 4.17. Given that Pakistan i s a federation, i t i s critical for improved budgetary outcomes and enhanced effectiveness o f public expenditure that MTBF process i s “rolled out’, to all the provinces (and ultimately to all the districts). Ideally, the national MTBF would be a consolidation o f subnational and federal MTBFs. I t i s essential, therefore, that the process o f promoting provincial adoption o f MTBF be l e d by a strong commitment from the center. 4.18. Fortunately, two provincial governments (Sindh and NWFP), under their economic reform programs (supported through World Bank structural adjustment credits), have already made a start toward formulating detailed MTBFs. Although the federal govemment, with technical assistance from the British Department for International Development (DFID) (see B o x 4.3) has embarked upon expanding i t s MTBF, there i s a strong need to base all three MTBFs (and indeed the future MTBFs o f Punjab and Balochistan) o n a sound institutional footing. 4.19. The introduction of the new MTBF and PRSP processes provides an important opportunity for strengthening overall planning in Palustan and improving policy coordination among the key economic ministries and departments both at the center and in the provinces and between the federal government and the provinces. T h i s change will not, however, happen automatically. T h e role o f the government needs to be redefined in the light o f new approaches. The objectives should be to ground the new processes in not only better estimation o f resource availabilities and flows, but also greater knowledge about developments in the real economy, sectoral trends, and realities o f project and program implementation. 4.20. The role o f provinces in developing realistic budgetary frameworks and viable poverty reduction strategies can hardly be overemphasized. As discussed later, the federal government must devolve greater authority to the provinces with respect t o economic and social development and eliminate or limit federal projects and initiatives in a number o f sectors, such as education and water, as part o f the overall process o f moving services closer to the people. This devolution from center to the provinces will give impetus to n any case, i t i s clear that i t i s the provincial- and the major initiative o f devolution to the district level. I district-level spending on education, health, l a w and order, provision o f justice, irrigation, rural and urban water supply, and roads that will account for bulk o f the increase in public spending in the near future. The institutional arrangements that will integrate financial, policy, project, and sectoral plans at the provincial and district level clearly need to be strengthened. The role o f the provincial planning departments should be broader than the present excessive focus o n development budgets alone, which are a fraction o f provincial spending. At the same time, planning departments will need t o work closely with 71 both finance and implementing departments and agencies in costing programs, developing yardsticks o f spending, and monitoring o f expenditure implementation, impact, and outcomes. Box 4.3: DFID Assistance to Strengthen the Medium-Term Budget Framework The British Department for Intemational Development i s providing technical assistance (TA) to the govemment of Pakistan (in the form o f intemational and local experts) to strengthen its attempts to develop a Medium-Term Budget Framework (MTBF). At present the budget framework does not go beyond presenting with the budget documents a three-year macroeconomic forecast, which i s updated annually. The TA, which i s expected to be spread over thirty months, will initially focus on the health and population sectors but subsequently will be rolled out to cover other key sectors. The Objectives The main objectives of the TA will be to help the govemment improve the quality, transparency, and accuracy of Pakistan’s medium-term budgeting, provide a firm basis for transmitting policy into effective expenditures, and thereby improve federal expenditure by bringing credibility and stability to budget plans, along with improved estimation of transfers and allocations to subfederal tiers of govemment. T h e TA i s expected to improve the quality of short- and medium-term budget forecasts, including revenue forecasts and rolling estimates of the costs and liabilities o f existing and new policies, drawing on intemational best practices. Moreover, the TA will help the govemment develop necessary institutional capacity to prepare, implement, expand, and deepen the MTBF in the future. Institutional Setting In the first year the MTBF will be piloted in the Ministries of Health and Population Welfare. In subsequent years MTBF planning will be rolled out to other line ministries such as education, and transport and irrigation, which require adequate recurrent resources for the operation and maintenance of large physical assets. As such, although the Budget Wing o f the Ministry of Finance will be the main counterpart agency for the TA, the process i s expected to intimately involve the Expenditure Wing of the Ministry o f Finance, the Planning Commission, the Central Board of Revenue (CBR), the Ministry of Health, the Ministry of Population Welfare, and other key federal l i n e ministries The T A Components (a) Strengthening the Resource Framework: The TA will help the Ministry of Finance, the CBR, and the Planning Commission develop sound analytical foundations for the macroeconomic framework, better estimate contingent liabilities, estimate and incorporatequasi-fiscal activities, and strengthen short- and medium-termrevenue forecasting. @) Establishing InterSectoralPolicy and Strategy: The TA will help the govemment undertake better technical work and help establish stronger institutional mechanisms to link technical work with political adjudication of intersectoral policy priorities. (c) Supporting Line Ministries: T h e TA will help with line ministries (initially focusing on the Ministries of Health and Population Welfare) to: (1) develop sectoral policies and strategies; (2) translate the policies and strategies into programs with well-defined objectives and outputs; (3) develop techniques and procedures to estimate program and project costs; (4) prepare complementary and consistent recurrent budget and development budget proposals in an MTBF format; and (5) prioritize allocations to programs according to resource availability. (d) Securing Improvements in Budget Implementation: The TA will also help the Ministry of Finance, jointly with line ministries, to review budget implementation procedures, specifically the new system o f financial control and budgeting, and secure a more efficient flow of funds mechanism. I 4.21. At the federal level the Finance Ministry in general, and the Budget Wing, in particular, will take the lead in malung the federal MTBF, but much improved coordination would be required between the Ministry o f Finance, the Planning Commission, the sectoral ministries, the Central Board o f Revenue (CBR), the Economic Affairs Division (EAD), and the Accountant General o f Pakistan Revenue (AGPR) for improved formulation and execution o f MTBF. The respective roles o f the Planning Commission and the Finance Ministry in developing a macroeconomic framework and sectoral plans and targets need clarification. Though some overlap in functions between economic ministries may be unavoidable, and indeed healthy, the overall lack o f capacity in all parts o f govemment and systemic problems with the civil service suggest the need for a better delineation o f responsibilities and clearer coordination arrangements. 4.22. Preparation o f macroeconomic frameworks i s a case in point. The Planning Commission has traditionally had a mandate t o prepare medium- and long-term development plans and for that purpose 72 also prepares a medium-term economic framework. Over time, however, the operational relevance o f these medium-term frameworks declined as fiscal crises forced more ad hoc approaches to resource estimation in the Ministry o f Finance. Recently, the Ministry o f Finance has undertaken two initiatives to upgrade i t s macroeconomic work. First, a macroeconomic framework i s to be prepared as a part o f the MTBF in the expenditure wing. Second, the newly created Debt Policy Coordination Office (DPCO) has also been charged with the task o f preparing a macroeconomic framework under the Draft Fiscal Responsibility and Debt Limitation Law. There i s the obvious need for better coordination within the Ministry o f Finance on the macroeconomic framework and between the Ministry and the Planning Commission, The latter can bring to bear i t s expertise in the real economy and the sectors in the preparation o f macroeconomic projections. T o further this coordination, it i s important that at the present juncture, the Ministry o f Finance and Planning Commission remain under a unified leadership, preferably under the same Federal Minister. A unified leadership approach to planning and budgeting has been tried in quite a few East Asian countries with considerable success in accelerating economic decision-malung. 4.23. Development o f an MTBF involves much more than the estimate o f the resource envelope. I t also involves substantial work on sectoral and subsectoral priorities, costing o f programs, monitoring o f public spending, and assessment o f outcomes. In these and related areas the Planning Commission, working closely with the provincial planning departments, can play a major leadership role, quite distinct from i t s responsibilities for appraisal and approval o f large development schemes. Re-invigoratingthe Planning Processes 4.24. The preparation, monitoring, and evaluation o f MTBFs and PRSPs, at both the federal and the provincial level, provide a unique opportunity for invigorating Palustan’s planning processes, which in their heyday in the 1960s were considered a model for developing countries (see B o x 4.4). 4.25. During the past two years progress has been made in providing the Planning Commission with greater authority in project appraisal, shifting the focus o f macroeconomic and sector work away from Five-Year Plans to a ten-year perspective, and setting up rolling Three-Year Plans. Project monitoring has been made systematic and i s shared with public, and the poverty work has been substantially strengthened. The recent improvements in the planning processes have been grounded in the general desire o f the Musharraf govemment to give a larger role t o technocrats in policy and decision-making. But clearly the institutional capacity across economic ministries and planning agencies remains extremely weak. There i s excessive reliance on outside agencies for analysis. Significant improvements in sectoral planning and macroeconomic and budgetary expenditure frameworks are needed both at the center and in the provinces. 4.26. H o w can the planning processes both at the center and the provinces be strengthened? First, at the center there must be both a closer collaboration and a clearer division o f responsibility between the finance division and the planning division. For instance, the planning division could play a more modest role in preparing a macroeconomic framework but should have the major responsibility for setting the sectoral priorities and development allocations. 4.27. The Planning Commission enjoys two distinct advantages vis-a-vis other ministries: (a) It s t i l l has more competent technical staff to deal with all the sectors o f the economy and has a long tradition o f interaction with line ministries and provincial departments. (b) I t i s not subject to the day-to-day pressures that tend to preoccupy the Finance Ministry. 73 Box 4.4: Planning Processes: Sources of Weaknesses Pakistan’s strong economic record o f the 1960s was attributable in part t o the strength o f i t s planning processes and the unique role o f its planning institutions. Contrary t o popular perceptions, it was not the existence o f Five-Year Plans as such that was the m a i n contributor to the quality and speed o f economic policy adjustments and generally sound resource allocation decisions during the 1960s. M o r e important factors were the strong political support t o the planning processes, the high level o f professionalism and independence o f groups o f planners both at the center and in the provinces, and the economic policy coordination mechanisms, which, notwithstanding the centrality o f the MinistryiDepartment o f Finance, provided a clear role for the planning agencies as w e l l as the sector ministries and departments. Serving as high-powered economic and technical staffs, the Planning Commission and the provincial planning departments were very effective in analyzing key economic development issues, recommending adjustments in policies and programs, and assisting the top leadership and the finance minister in both economic policy coordination and foreign aid mobilization. There is general agreement that the planning processes have greatly weakened over time even though Five Year Plans have formally been in existence throughout most o f the past three decades: the Eighth Plan covered the period 1993-98, and the Ninth Plan for 1998-2003, though prepared, was never finalized. Factors responsible for the decline in the effectiveness o f b o t h the planning processes and the planning . institutions have varied over time but include the following: Reduced political willingness over time t o accept the financial and economic discipline o f independent economic and project analysis; in the 1990s large projects in k e y sectors were taken out o f the jurisdiction o f the Planning Commission, and only in the past two years have the powers o f the Planning . Commission t o review a l l large projects have been restored. . Excessive focus o n preparation o f Five-Year Plans and n o t enough attention t o analyzing the causes of continued failures to acheve planning targets in national savings, exports, and social sectors. Frequent lack o f realism o f the macroeconomic frameworks prepared by the Planning Commission, contributing to excessive n e w commitments o n projects and subsequent spreading o f development expenditures too thinly and frequent inadequate finding o f foreign-aided projects. . . Insufficient attention t o sector plans and the mix between policy requirements and investment needs. Poor links between resource needs and physical targets. Lack o f balance o f effort between planning o n the one hand and monitoring and evaluation o n the other; even where monitoring has taken place, it has been excessively concerned with inputs o f money rather . than achevement o f set goals, and impact analyses o f development spending have been rare. Last but not least, professional capacity in a l l agencies concerned with development has become very weak; little n e w professional staff has been inducted into the Planning Commission during the past ten o r fifteen years. 4.28. T h e Planning Commission should cede the existing de facto role o f overall economic coordination to the Finance Ministry but keep primary responsibility for formulating sectoral policies, plans, and strategies. These sectoral plans need to be developed in collaboration with the line ministries and departments. The Planning Commission (and provincial planning departments) would be responsible for intersectoral priorities in the MTBF for development. This implies that the Planning Commission (and planning departments) should be responsible for determining the allocations at the sector and agency level. For this purpose, while the Public Sector Development Program (PSDP) needs to be formulated within the resource envelope provided by MOF, the recommendations o f the Planning Commission on budgetary allocations for individual projects, programs and sectors may be treated as final. 74 4.29. The Planning Commission and the planning departments should lead the effort to integrate recurrent and investment expenditures in a coherent MTBF. The preoccupation o f planners with the development budget i s hampering the pursuit o f balanced strategies and speedy social development. T o ensure progress, the policies formulated under the major initiative o f PRSP need to be consistent with sector-specific policies and total sector allocations. Full integration requires the following: (1) the same strategic objectives should drive the formulation o f the recurrent and investment budgets; (2) the budget planning process at the sector level should consider options to achieve these objectives and implications for recurrent or investment expenditures based on a predefined resource envelope for the sector; (3) project selection should be based o n a range o f economic and non-economic criteria, especially on the defined role o f the government within the sector; (4) the best projects to realize the sector programs should be selected o n the basis o f adequate cost-benefit and cost-efficiency analysis; and (5) monitoring and evaluation o f the budget programs should be based o n a unified budget classification structure that allows sensible analysis. T h e full integration o f the recurrent and capital budget planning processes i s a gradual and long-term process. As a first step, however, there i s a need to improve the estimation o f the recurrent cost implications o f the ongoing investment projects from the l i n e ministries to be reflected in the MTBF sector expenditure proposals. Equally important, organizational arrangements for recurrent and investment planning in the l i n e ministries, the Ministry o f Finance, and the Planning Commission need to be realigned to eliminate the present separation. Only in that way will investment projects be an integral part o f the MTBF. 4.30. The Planning Commission should develop a rolling program o f sector reviews, possibly three to four sectors a year, which will provide depth and credibility to i t s long-term plans. Such sector reviews could be a vehicle for worlung with the World Bank, the Asian Development Bank, and other donors on analytical issues 4.3 1. M o r e concerted efforts to use monitoring and evaluation at the project, sector, and economy level are needed to monitor progress toward key goals and t o better understand the constraints hampering economic and social progress. Systems being put in place t o monitor PRSP can provide an excellent step in this direction. 4.32. The goals listed cannot, however, be achieved without substantially restoring professional strength. The Planning Commission should opt for a lean but highly skilled staff. For some o f the routine functions o f project appraisal, it could rely o n specially constituted technical panels drawing in part o n highly professional experts from outside. Strengthening Expenditure and Program Monitoring 4.33. One key factor behind the declining effectiveness o f public expenditure i s a dysfunctional expenditure monitoring system. The traditional budgetary monitoring system has focused on ensuring the spending o f the budgetary allocations and the procedural correctness o f these expenditures. Little, if any, attention i s paid to service delivery and client satisfaction. Nevertheless, appropriate systems were put in place within the Ministry o f Finance (finance departments at the provincial level), the EAD, and the Planning Commission (planning departments in the provinces) to achieve at least the limited objective o f public expenditure monitoring. For example, the Budget Wing o f the Ministry o f Finance reviews the current and development expenditures on the basis o f data received f r o m the Controller General o f Accounts (CGA) and project entities on monthly basis. The EAD undertakes a periodic review o f donor- assisted projects and programs. The Programming Section and the Projects Wing o f the Planning Commission undertake monitoring o f financial and physical progress o f the development projects. N o t only was the quality of data used by these monitoring exercises declining steadily overtime, but also the interest o f decision makers in using the results o f these exercises in their planning and budgetary 75 decisions. Partly because o f falling demand, but mostly because o f the inadequate human and financial resources allocated to it, the quality o f monitoring i t s e l f has also been going down. 4.34. The Planning Commission has recently revived monitoring o f development projects through in- depth quarterly reviews o f key projects and programs. Although a step in the right direction, this process has yet to factor prominently in inducing the overall economic decision making. 4.35. Meanwhile, the I-PRSP has proposed a more elaborate mechanism to monitor poverty-related expenditures, along with their intermediate indicators and outcomes, vis-A-vis the specific targets set by the government for human development and poverty reduction. When fully functional, the proposed mechanism can provide improved and continuous monitoring o f PRSP inputs and outputs, which would help to validate the choices made in the prioritization process and can reveal the reasons for the success or failure o f the various interventions chosen. Further, it can offer warning signs to those responsible for the management o f the Poverty Reduction Strategy (PRS) when interventions are not having the desired effect or are having undesirable side effects. If undertaken correctly and diligently, the proposed monitoring system can facilitate management decisions about continuation, acceleration, or amendments o f various interventions. Moreover, an effective monitoring system o f this kmd can help develop greater accountability in the use o f public resources. . 4.36. The proposed mechanism includes following areas to be monitored: . . Budget implementation and public expenditure . Policy reforms at a number o f different levels (macroeconomic, structural, and sectoral) Public action choices . Public services (quality, availability, access, satisfaction) Investment programs 4.37. With technical assistance from DFID and the W o r l d Bank, the government has identified indicators at a number o f different levels along the “impact chain,” starting with inputs and outputs, followed by outcomes and impacts. Special emphasis has been given to identifying the intermediate outputs, which bring into the government strategy elements o f accountability, learning and changing actions, and intervention. The system requires monitoring PRSP expenditures on quarterly basis at the federal, provincial, and district 4.3 8. Regarding monitoring longer-term outcomes, the I-PRSP presents benchmarks o n the present-day status of social indicators and establishes medium-term targets for a variety o f indicators, such as infant mortality rates and immunization rates. For these benchmarks and targets to have any meaning, however, they need to be underpinned with a reliable and credible database, as well as backed by the capacity t o interpret and analyze the data. While Pahstan has both administrative and survey data sources, but both have their limitations, and in any case show wide differences in the present status o f health and education levels. Administrative data are incomplete and unreliable, and household survey data are representative only to the level o f the province (not district). There i s a considerable lag between data collection and data availability. 4.39. In order to remove these deficiencies, a new survey instrument Core Welfare Indicator Questionnaire (CWIQ) has been developed t o monitor changes in key social and poverty indicators o n an annual basis at the national, provincial, and district level. The primary objective i s to keep track o f the short-term changes in both the intermediate outputs and outcomes. The Pakistan Integrated Household 48 In addition, the MOF has already started reporting on two indicators o f intermediate outputs on a quarterly basis. 76 Survey (PMS) remains the key vehicle for providing information o n medium-term changes in human and poverty indicators. 4.40. Because PRSP monitoring i s a multidimensional task, i t i s being undertaken by a collaborative effort o f a large number o f government agencies, including the Ministry o f Finance, the Controller General o f Accounts and the provincial accountant generals, the line ministries and departments, and the Federal Bureau o f Statistics. Given i t s vast comparative advantage in poverty research and monitoring o f public expenditure and projects, the Planning Commission has been assigned a key role in monitoring PRSP outputs and outcomes. It i s essential, however, to improve the database and strengthen the analytical capacity of, and cooperation between, the involved agencies to institute an effective monitoring and evaluation system for the government’s poverty reduction strategy. Redefining the Role o f the State 4.41. Improving the effectiveness o f public services and expenditures requires not only mending the institutional framework, but also reviewing the structure and functions o f the government. First, the role o f the state in the economy must be tailored to i t s institutional and financial capability and the market economy alternatives available, because in many areas o f provision o f even public goods and services, the private sector works more efficiently. Second, even in functions that should ideally be in the government’s domain, i t may be possible to move away from the over-centralization o f responsibility at the federal level that has taken place. 4.42. Apart from defense, foreign affairs, and natural resource management, the federal government has a strategic role in economic management that has become and will become highly complex in an integrated world economy. The state at the central level must be responsible for stabilizing the economy, ensuring fiscal discipline, containing inflation, balancing the external accounts, and keeping the debt burden within sustainable limits. It must protect the consumers’ and stakeholders’ interests by regulating the private sector, facilitate investment, provide a level playing field t o economic actors, curtail rent- seeking behavior, preserve i t s autonomy against intrusion by pressure groups, and ensure a competitive environment, vital for survival and growth in a rapidly globalizing world. I n addition, it must provide an effective and reliable legal framework and recourse institutions for conflict resolution. I t must curtail transaction costs by promoting transparency and integnty and imparting greater predictability to the system. It must also reduce information rentals by facilitating access to policies, plans, and data on economic outcomes and trends. 4.43. Just this strategic agenda i s formidable and even the best would be hard put to deliver. Hence, extending i t s energies and resources to activities and functions which could be well-performed by the private sector andor sub-national levels o f government, the central government undermines not only the development o f private sector, but also i t s own ability to effectively deliver o n i t s strategic agenda. 4.44. According to the 1973 Constitution the federal government i s responsible for, key areas like defense, foreign affairs, banlung and currency, postal service, transportation (ports, airports, railways), international trade, macroeconomic management and industrial development. The provinces share with the federal government responsibility for population planning, irrigation, curriculum and syllabus planning, and social welfare, and are solely responsible for roads, highways, l a w enforcement (police) and dispensing o f justice, education and health. Under the LGOs 2001, the provincial governments have devolved a large portion o f their responsibilities to the local (district, tehsil and union) governments. 4.45. In Palustan, while the legislative division o f powers, as set out by the Constitution and provincial statutes, compares favorably t o the assignment based on principals o f fiscal federalism, in practice the federal government has expanded itself in areas o f shared responsibility (and even in fields which 77 constitutionally belong to sub-national governments), and have even taken over some o f the traditional private sector functions. f the Government in Producing and Providing Private Goods and Sewices Reducing the Role o 4.46. Successive recent governments have already attempted to reduce the role o f the state in the economy through privatization o f state enterprises and encouraging private sector into areas that had exclusively been reserved for the public sector. Nevertheless, the government’s role in the economy still transcends beyond the optimal and i t s involvement in many sectors i s s t i l l excessive v i s - h i s the appropriate role defined by economic principal^,^' i t s administrative capacity and fiscal resources. There i s significant r o o m for greater private sector involvement in the fields o f telecommunications; electricity generation and distribution; gas and o i l exploration, production, and distribution; electronic media; banking, finance, and insurance; transportation (including air, sea, and rail traffic), and others. Indeed, the present policy o f the government i s to retain only the regulatory functions and to gradually vacate all o f these areas for the private sector. The government should accelerate the process o f privatization to minimize the deterioration in performance and asset-stripping that may occur once the privatization process i s started. Lack o f an adequate regulatory framework should not be used as an excuse for delaying privatization. Although appropriate regulation o f the private sector, especially o f the privatized state monopolies, i s essential, i t should be possible to develop the regulatory capacity o f the government concurrently with privatization. 4.47. Efficiency o f non privatized organizations can be enhanced by: (1) giving them autonomy over their input and output decisions, thereby increasing their allocative and productive efficiency; (2) protecting them f r o m political interference, to enhance the efficiency o f their personnel decisions; (3) subjecting them to a hard budget constraint to ensure that the resources are used efficiently without any waste or leakage; and (4) providing their management and staff a performance-based incentive structure. 4.48. Other than the sectors already cited for reduced public sector involvement, i t may be appropriate for the government to reassess the needs and roles o f different kinds o f public sector entities (such as government departments, attached bodies, autonomous bodies o f various types, and commercially oriented public enterprises) with a view to making the overall size o f government more consistent with available resources and focused more intensely o n the govemment’s strategic agenda. The decision-tree in Figure 4.1 can help in making a quick assessment o f whether a particular governmental function should be shed or continued. 4.49. There i s also a need to re-examine the role o f government as a producer/provider vis-A-vis as a financier and facilitator o f services. An area, which has not been fully explored in Palustan, i s the possibility o f inducting the private and non-government sector in service delivery through public-private partnership. Despite the emerging possibility o f larger fiscal space becoming available, both from financing but more importantly from efficiency point o f view it i s necessary that the government should actively pursue private sector in all those sectors where it can deliver service more efficiently. In all these areas, the government responsibility may be limited to regulating, and in some cases financing, the delivery o f service through private providers, so as achieve i t s equity and other social objectives. In many cases, i t i s more efficient for the government responsible to provide only finances to the private sector to provide the service. The financing mechanisms therefore need to be developed that attract the private sector in provision of public services but at the same time are able protect the poor and other vulnerable groups, for being excluded from private delivery system. B o x 4.5 provides examples o f such financing mechanisms in the case o f delivery o f education. 49 Steps required to decide on what are the core functions o f a particular govemment are highlightedin Figure 4.1. 78 Box 4.5: Options for Public Financing of Private Delivery of Education Public funding for private schools: In many countries, such as Canada, Chile, and the Netherlands, the government promotes education by providing f i a n c e s to the private schools. The recent spread o f private schools in the urban and, t o a lesser extent, rural areas o f Pakistan offers a possibility for the government of better use the private sector t o further i t s objective o f a rapid spread o f school education. This objective requires providing public financing t o the private schools in targeted poor and backward areas and neighborhoods o n a per-student basis. A s these payments are based o n the average cost o f education in the public sector, while expenditures per student (for comparable education) are significantly less in the private sector, the private schools have a strong incentive t o vie for new students, thereby raising school enrollment in private schools. Public schools under private management: The government may choose t o contract one of more qualified nongovernmental organizations (NGOs) t o manage a certain number o f public schools. An ideal contract would give the managing NGO the right t o appoint the head teacher and school staff. The teachers would have qualifications and salaries comparable t o those o f the contract teachers o f public schools. The government would bear the salary cost and provide t o the NGO the nonsalary operational cost o n the basis o f a per-student average o f the public schools’ cost. Such a mechanism would imply n o additional budgetary impact t o the government, while the NGO would have every incentive t o run the school in a most efficient manner and raise the quality o f education to bring in more students. Such an arrangement would b e ideal for enhancing the outreach o f education to remote o r backward areas where it i s difficult for the public sector to operate. Giving parents a choice (the voucher system): One option for improving school quality i s to give parents a choice through a voucher system. Under this system individual students would be given vouchers funded by public revenues but redeemable in either public or private schools. Schools would have to compete for student patronage, mainly through raising the quality o f education. Such a system has been tried with considerable success in many developed countries (such as the United States) as w e l l as developing countries (such as Puerto Rico). All these means of public-private partnership require strong monitoring t o protect the interests o f the stakeholders. Such monitoring can b e established only through the participation o f the community. 79 Figure 4.1 :Decision tree for identifying core governmental tasks I Disaggregate functions Policy Co-ordinanon Service deli\ ery support Reeulation No I Does thefunction support other governmental priorities? No be achieved through regulation, taxation, transfers to sub- oes national or internationa law or the Constitution require (iii) No Yes Source: Methodology for Functional Reviews, Paper produced by the Russian Federation Govemment, World Bank, July 2001. 80 4.50. Because o f their public goods nature, many health services (preventing and treating infectious diseases, providing safe drinking water, generating and disseminating health information, and preventing non-communicable diseases) will be underprovided if left exclusively to the market forces, as public benefits far exceed private returns. The government has a natural role in providing all these services. Moreover, the government should also get involved in providing a safety net for the poor and for hard-to- reach and difficult-to-treat, unprofitable segments o f the population. Thus the public sector role in heath care should center on activities such as providing immunization, controlling disease vectors such as mosquitoes and rats, and guaranteeing basic care for the poor and excluded. Beyond these areas, there are ample examples o f high-quality health care services being provided and financed efficiently and equitably by both public and private sectors. The government has only to ensure appropriate incentives to the providers, public or private, to guarantee efficiency. A recent attempt to promote public-partnership in health care i s in the area o f detection and treatment o f cases o f tuberculosis (see B o x 4.6) Box 4.6: Public-Private Partnership in Tuberculosis Control Pakistan has more than 250,000 new cases o f tuberculosis every year, o f which only 20 percent are detected. In 1994/95 the World Health Organization recommended a new strategy, “DOTS” (Directly Observed Treatment - Short Course) to eradicate TB globally. DOTS is an inexpensive and effective approach, but unfortunately it covers only 10 percent o f TB patients in Pakistan. Only 20 percent o f Pakistani TB sufferers seek care from the public sector, and about 40 percent go to the private sector, mainly to general practitioners and NGO health units. The remaining 40 percent either do not seek any care or go to other nonprofessional providers. Unfortunately, the nonstandardized TB practice in the private sector leads to very l o w cure rates. In addition, the infectious nature o f the disease leads to increased transmission o f the TB germs to others, and inferior-quality treatment creates the risk o f multi-drugresistant (MDR) TB. Given the nature o f care sought by most TB patients, it i s suggested that a public-private partnership would be an ideal approach for treatment of TB. The National TB Program (NTP) i s seeking acceptable and feasible mechanisms to ensure closer working between the government and private partners to extend TB care according to national standards. Pooling human and financial resources in the two sector has enormous potential for nationwide TB control. Involvement o f private sector providers in implementing DOTS in Pakistan has been recommended in a nationally agreed strategic framework. Consultative processes are ongoing, and there are a few good experiences in some provinces. Successhl implementation of the DOTS strategy through collaboration between the M a r y Adelaide Leprosy Center (MALC) and the Government of Azad Jammu and Kashmir, under the World Bank-financed Northern Health Project, i s one such example. The key components o f the public-private partnership project in TB control, to be implemented mainly at the district level, include the following: Capacity-building ofthe private and public sectors: Because the majority o f TB patients seek care from private practitioners, the ”IT aims at training the private doctors, paramedics, and laboratory technicians in standardized good-quality TB treatment. In addition, the private sector will be provided with the latest equipment (microscopes) and chemical reagents to facilitate detection. The public sector, too, w i l l be upgraded: district health offices, supported by provincial authorities, will be equipped to develop and sustain partnerships needed to increase DOTS implementation. 1 Advocacy: Media messages to promote awareness about health issues w i l l be developed at the federal level. 1 Research and evaluation: Provincial officials w i l l participate in development, research, and evaluation activities in their respective provinces. 4.5 1. Similarly, given the importance o f agricultural extension in enhancing agricultural productivity, and the virtual collapse o f public extension services, there i s considerable scope for public financing o f private delivery o f extension services. In the Palustani context, such private extension service providers might include unemployed agricultural graduates, agricultural consultants, consultancy firms, progressive farmers, farmers’ organizations, cooperatives, nongovernmental organizations, agribusiness companies, 81 input dealers, newspapers, private television channels, and private sector banks. The key i s to involve all or some o f them in the agricultural development process through appropriate incentive mechanisms, including direct financing from the budget. 4.52. The private sector can also be inducted in the road sector through appropriately designed concessions. Information technology i s a fast-developing area in Pakistan, having relatively large claims on public funds. According to international experience, the sector requires public involvement in regulation and facilitation more than in development and provision o f information technology, which can be well and efficiently performed by the private sector. 4.53. A number o f services even in the subnational domain might be better provided by the private sector, including f i r e protection, refuse collection, street maintenance, traffic management, local transit service, refuse disposal, and electric power. I n addition, private sector participation i s useful in primary and secondary education, transportation, water supply, sewerage, and health services (see World Bank 1994 for details o n such participation). Furthering Devolution 4.54. Perhaps the most important initiative o f the Figure 4.2: T h e 2-step logic of decentralization government i s i t s comprehensive plan o f restructuring the -1 government through devolving the state authority Table Structural changes: 4.1 presents the areas o f responsibility o f each level o f local g~vernment.~' The plan attempts to create full- fledged district and municipal (tehsil) governments administrative having legislative and financial powers, with the main objective o f changing the political and administrative culture o f the government by taking government closer to the people and therefore getting communities better N e w accountabilities for integrated with the government. This, it i s hoped, that the senior staff I devolution process would improve decision-making and government's accountability, particularly o f service I I I providers, to the citizens. Service delivery improvement arrangements: 4.55. The overall objectives o f the current devolution supply-side program in Palustan can be discemed from the various demand-side publications o f the NRB, the local government ordinances, and their amendments. Some interpretation i s required, as the objectives are not set out in any single document. One point i s clear from government ces for the integrated development and management o f the following 50 In addition, city districts may set up district municipal offi responsibilities jointly with the Tehsils: (1) water source development and management, storage, treatment plants, and macro- distribution; (2) sewage tertiary and secondary network, treatment plants, and disposal; (3) storm water drainage network and disposal.; (4) flood control protection and rapid response contingency plans; (5) natural disaster and civil defense planning; (6) solid waste management, treatment, and disposal, including landfill cities and recycling plants; (7) industrial and hospital hazardous and toxic waste treatment and disposal; (8) environmental control, including control o f air, water, and soil pollution in accordance with federal and provincial laws and standards; (9) master planning, land use, zoning and classification, and reclassification; (10) urban design and urban renewal program; promulgation o f building rules and planning standards; (1 1) parks, forests, playgrounds, and sporting and other recreational facilities; (12) museums, art galleries, libraries, and community and cultural centres; (1 3) conservation o f historical and cultural assets; (14) landscape, monuments, and municipal omamentation; (1 5) urban and housing development, including urban improvement and upgrading, and urban renewal and redevelopment, with care being taken to preserve historical and cultural monuments; and (16) regional markets and citywide commercial centres. Moreover, unions can receive functions from districts and Tehsils, provided that they are also given the funds for managing those functions. 82 descriptions o f devolution: it i s intended t o improve service delivery and i s not an end in itself. Keeping a focus o n the intended end product--improvements in service delivery--is crucial, as the literature i s clear that decentralization, per se, offers n o guarantees o f improvement in accountability or service d e l i ~ e r y . ~ ' 4.56. Devolution i s an attempt to change incentives within the public sector through a pragmatic combination o f administrative, fiscal, and political changes. These changes are intended t o change how senior staff o f subnational governments understand their tasks-that is, changing their perceived accountabilities so that they feel responsible for achieving service delivery improvements. Those service delivery improvements are likely to be variable between districts and across sectors and thus difficult to specify precisely in advance. They are also likely to include both supply-side reforms, improving the quality o f health or education services that are produced, and demand-side, strengthening citizens' ability to pressure government for better-quality services. 4.57. The key assumption i s not that these service improvement arrangements could not happen within a centralized system. The logic i s that malung any such arrangement work requires effort and ingenuity o n the ground, and changing the accountabilities o f senior staff in local government motivates them to install and sustain these arrangements. 4.58. Primary education offers an example o f this chain o f connections. Passing authority over teacher recruitment and discipline to competent local school management committees can be a step toward better educational service delivery. Establishing local govemance arrangements for schools could in principle, however, be undertaken by central government as well as by subnational government. The decentralization argument i s that a judicious mix o f political and fiscal changes, coupled with some decentralization o f hire and fire authority over teachers to local government, will be the trigger that makes nazimeen and senior district officials more concerned with making these arrangements work. 4.59. The same example illustrates why service delivery improvements require better local policy as w e l l as improvements in service quality. Improvements in the quality and attendance o f primary school teachers are rather marginal if patronage and elite capture significantly direct public spending toward secondary education for the middle classes, rather than primary education for the poor. 4.60. Initial evidence concerning community oversight at local level i s encouraging. M a n y school management committees seem quite well developed and have had considerable success in reducing the n Sindh, under the Khairpur Shelterless School Scheme, thirty-three schools cost o f school construction. I have been completed at a third of the cost of a comparable government school, in quicker time, and without compromising on the quality of construction. It seems that the elected officials were crucial for n addition, some mobilizing the communities and ensuring the effective functioning o f these institutions. I early signs show that the education monitoring committees are having an impact through surprise visits to schools and checks o n teacher attendance through direct observation. It i s important to build o n these early successes and, mindful of the risks of overclaiming, to publicize concrete examples o f successes if some momentum i s to be maintained. 5' A recent review o f the literature found that the empirical evidence that decentralization increases allocative efficiency in public expenditure, accountability, or cost recovery i s quite meager. T h e literature suggests that decentralization can have disadvantages, including failures o f policy coordination, "elite capture" o f local govemments, and overburdening o f weak local govemments. Overall, it seems that while moving political authority over services and over resources to subnational govemments can indeed improve service delivery, positive results are contingent on a large number o f other factors, not least ensuring the ability o f civil society to exercise discipline over govemment both through voice and exit as mediated by the electoral arrangements and access to information; preventing monopoly control o f information being exercised by community leaders; locating functional responsibilities for public goods at the right levels; limiting the retention o f hierarchical authority by national govemments (even if this i s rarely used); and improving the flow o f information from governments to their constituents. (See Azfar, Omar. 2002. "Conditions for Effective Decentralized Governance: A Synthesis o f Research Findings" (http://www.inform.umd.edulIRIS/IRISIPEG/synthesisgaper.pdf). 83 4.61. Although Pakistan has made considerable progress in devolution, there i s a considerable distance s t i l l to travel. In terms o f distribution o f state responsibilities among various levels o f government, i t i s apparent that the actual assumption o f responsibilities at the higher levels has departed substantially from the de jure assignment, with the federal government increasingly encroaching on provincial governments’ responsibilities. For example, although education i s a subnational responsibility in Palustan, the federal government has assumed responsibility for formulating national policy, planning, developing curriculum, managing centers o f excellence, and financing university education through the Higher Education Commission (formerly through the University Grants Commission). Table 4.1: Functional Responsibilitiesof Districts, Tehsils and Union Councils City District District I Tel~siYTaluka/Town~~ Union *Public transportatio: *District planning *Spatial plans *Statistical information for and mass transit *Schools (boys & girls) *Control over land-use, land- socio-economic surveys *Passenger and freight *Girls Schools subdivision, land development *Consolidation o f village and transit terminals *Technical Education and zoning by public and private neighborhood development *Traffic planning, *Colleges (other than sectors for any purpose needs engineering and parking professional) *Prevention o f encroachments *Union-wide development *Industrial estates and *Special Education. *Regulation o f sign-boards and *Coordination for removing technological parks *Literacy Campaigns & advertisements deficiencies in the delivery o f Cottage, small and Continuing Education *Municipal infrastructure and TMA services medium sized enterprise *Vocational Education services *Registration o f births, deaths promotion *Public Health *Water supply and control and marriages ohvestment promotion *Basic & Rural Health *Development o f water sources *Libraries and protection *Child & Woman Health *Sewerage, treatment and *Sports toumaments, fairs, *Population Welfare disposal shows and other cultural and *District and Tehsil (HQ) *Storm water drainage recreational activities; hospitals *Sanitation *Cattle fairs and cattle *Agriculture (Extension) *Solid waste collection and markets *Livestock disposal *Public open spaces, public *Farm Water Management *Roads and streets, other than gardens and playgrounds *Soil Conservation roads falling under the *Public sources o f drinking jurisdiction of, and maintained water, including wells, water *Soil Fertility by, the District Government or pumps, tanks, ponds and othei *Fisheries Government and streets works for the supply o f water *Forests (local) maintained by the Union *Street lighting, public ways *Community Organization and Administration or Village and public places (through registration Council mutual agreement with the *Labor safety regulation TMA) *Traffic planning, engineering *Social Welfare and management (including *Facilities for the *Sports and Culture traffic signalling systems, signs handicapped, destitute and *Cooperatives o n roads, street markings, poor *Enterprise and Investment parking places, transport station *Protection against stray Promotion stops, stands and terminals) animals and animal trespass *Civil Defense *Street lighting C a t t l e pounds *Information Technology *Fire fighting *Regulation o f grazing areas Development & Promotion *Parks, playgrounds, open spacr *Assistance in disasters and Environmental regulation and arboriculture natural calamities, and relief *Land Revenue & Estate *Slaughterhouses activities (including de-silting *Excise and Taxation. *Sports, cultural, recreational o f canals) *Spatial Planning and events, fairs and shows Development *Regulation o f markets and *District Roads and Buildings services *Energy and Transport I 4.62. Although the devolution program provides the best opportunity in years for improving the structure o f governance and the effectiveness o f public services, the system needs considerable fine-tuning to achieve these objectives. Some o f the areas that require early attention o f the govemment are the following: 52 In urban areas, these responsibilities are undertaken by c i t y districts, n o t the c i t y towns. 84 . Extending administrative and fiscal autonomy: Local governments do not yet have full control over the staff and finances o f devolved departments. For efficient delivery o f service, local government should have full autonomy over personnel decisions (hiring, posting, performance monitoring, and rewarding) o f facility-level and other related staff. Local governments at all levels will only start takmg their responsibilities for workforce management seriously as and when they consider themselves to be fully the employer o f their staff. Completing devolution means, first and foremost, transferring budgetary responsibility for salaries to the district governments. . Thus salary budgets should be transferred to the districts via account IV. Extending facility-level oversight: Useful steps have been taken in the development o f school management committees, but few significant employer functions have been transferred to the school level yet. Currently SMCs only have very limited funds, resulting from the transfer to the districts o f some nonsalary operating budgets, although in a few comparatively richer areas SMCs are raising their own funds as well. The devolution program presents an opportunity to further delegate decision making authority over critical facility-level matters to bodies such as SMCs, which can ensure better oversight over service facilities, perhaps reducing staff absenteeism. Staff absenteeism i s not a major issue for private sector schools and health facilities as immediate accountability i s ensured. For public facilities, community monitoring provides one avenue to . ensure facilities remain functional with their staff present. Improvingpolitical careers. It was hoped that the new system would introduce a new class o f politicians who had not previously stood for office and who would permeate the entire political system. Certainly, a large number o f women have for the first time participated in the political process. One issue, however, concerns the limited time that citizen activists will have to contribute to a large and growing range o f participative bodies, including Citizen community boards, citizen police liaison committees; village, union, district, and tehsil councils; participation in rural support programs and in district development committees for Khushal programs. Since the number o f new elected members at various levels i s large, allowances paid are very modest--a problem o f special concern to many low-income councilors (especially those elected on reserved seats). The . travel costs they must incur to attend a council session can be a much as a day’s wages. Reducing institutional complexity: To be motivated to deliver public goods, politicians must be able to effectively claim credit from voters for providing such goods t o them. The institutional complexity o f the new system suggests that voters might have little knowledge about which tier o f local govemment i s responsible for which functions. The high proportion o f politicians that are indirectly elected complicates the linkage between voters and their representative^.^^ Enhancing awareness and transparency o f roles o f each level o f government i s crucial for effective delivery o f . public services. Improving capacity: The NRl3 and local civil society groups have been providing training support to local councilors. T o date training has focused on constituency and coalition building, social mobilization, strategic negotiation, legislative oversight processes, and legal rights awareness for elected representatives. Further training could usefully include s k i l l s building for new local . government politicians in publicizing local achievements. Raising the eficiency offiscal management: Only a few provincial taxes, including the property tax, and some user charges have been devolved to the local government. T h i s situation i s in distinct contrast to the well-accepted logic that revenue assignments should be closely matched to 53One third o f tehsil and district council members are indirectly elected b y the 126,462 union councilors. (The ballot contained 126,462 seats, 21 seats for each o f the 6,022 unions. Not all seats were filled, however.) 85 revenue needs.s4There are also many local tax administration issues to resolve, such as clarifying the mechanism o f returning property tax revenue from the district to tehsils and education and . health sector user charges from the province to the district governments. Improving PFC formulae: The interim Provincial Finance Commission Awards need to be improved t o avoid adverse incentives that they might create. For example, according to the Sindh PFC Award, districts will be given an additional allocation matching their baseline collection o f certain items. This arrangement, however, may motivate districts to raise taxes regardless o f h o w well they are spent. Districts with wider tax bases and lower efficiency may gain more than those with a narrow base and higher efficiency. The PFCs could be encouraged (or required) to transfer additional resources through mechanisms that convey federal and provincial policy priorities. . Clarifiing mechanisms for conveying federal and provincial policy priorities: Provincial micromanagement i s enabled by the involvement o f provincial and federal governments in personnel management decisions in the districts (and to a lesser extent in the tehsils). I t i s partly justified because there are few transparent avenues available to the federal government and to provinces to influence local government policy priorities. In order for the devolution initiative to be in complete sync with I-PRSP, mechanisms, national and provincial policies, local priorities, and financial resources must be better integrated. T o achieve this, incentive mechanisms need to be instituted to help local governments prepare and implement their polices and plans in accordance with the national and provincial priorities, without compromising the autonomy o f local governments. Increasing the tenure o f senior district staj? Inter-district transfers o f provincial cadre staff are an enormous problem. As long as provincial and federal cadres retain positions in local governments, then some minimum time must be placed o n the tenure o f these . officers. Without this, the temptation t o use transfers as a covert policy instrument i s too strong. Improvingfinancial management andfiscal reporting: One approach to reducing the incentive to use development funding for achieving political benefit i s to improve the quality o f public reporting on the effectiveness o f development expenditures. For more effective service delivery, . the financial management system should be improved (see C F A A for detailed recommendations). Reviewing alternative funding mechanisms: Decentralization o f services i s usually justified on the basis improving the delivery o f services that are valued by the community and for which they are willing to pay. In the case o f water and sanitation, for example, devolving capital expenditures from the provincial level to the district or tehsil level, accompanied by good intentions t o tie them to capacity for operations and maintenance, i s unlikely to be sufficient t o discipline investment. Without a relationship between the quality and quantity o f services delivered on the one hand and the cost o f those services o n the other, infrastructure will suffer from underinvestment and worsening problems in operations and maintenance. Thus reliance on block grants t o finance water i s clearly influencing the institutional options that city districts or tehsils are contemplating . for delivery o f services. Strengthening the Mushawarat Committees: The weak management links between the districts, tehsils, and union councils do seem t o provide some challenges in collaboration. The only coordinating mechanism between districts and tehsils i s the Mushawarat Committee. Since each local body can make development decisions independently, without consulting or informing the others, there i s potential for considerable waste due to duplication o f efforts and development 54 T h e argument i s that govemments have few incentives to maximize their effort in raising revenues that they do not spend, and when they spend revenues that they do not raise, they have few incentives for spending them efficiently. See Shah, Anwar. "Fiscal Federalism in Pakistan: Challenges and Opportunities". World Bank. 1996. 86 schemes. The Mushawarat Committee can be strengthened to serve as a coordinating body for budgetary decisions. 4.63. As the system beds down, it will be important to harness the potential that i t offers for strengthening user control over local policy and over the supply o f local services. This will require that government take a distinctly pragmatic approach to the development o f citizen community boards, amending their structure and their governance arrangements as practical experience i s gained. 4.64. More strategically, devolution will require some adjustments in the new political environment. An active role for the provinces must be found within the new arrangements. One area that the provinces might well be encouraged t o pursue i s in actively choosing the form o f devolution that they encourage. They could choose among several different paths. Early signs are that Sindh i s likely to seek to maintain stronger guidance over policies at the local government level, whereas N W F P may feel more comfortable in delegating authority. If this i s the case, then the key i s clarity and honesty about what approach provinces are taking. Radical devolution might not be the approach chosen by each province. 4.65. Most important, however, local governments need adequate fiscal and human resources to fulfill their duties appropriately and efficiently. Considerable efforts need to go into developing the managerial and technical capacity o f the local governments and providing them with adequate facilitating office technology. Nevertheless, without a significant increase in fiscal resources o f local and provincial governments, there i s little hope for any major improvement in public services through the devolved delivery system. Federal- Provincial Responsibilities 4.66. With the recent devolution, the distribution o f responsibilities at lower tiers o f the government has been corrected to a large extent. Yet there i s significant scope for devolving additional functions from the federal to the provincial governments, which in many ways remains overly centralized. Over- centralization o f functions i s also reflected in the planning process, which s t i l l exhibits a strong top-down format. A large portion o f decisions regarding development expenditure are taken by the federal government. A clear example o f this i s inclusion of a large number o f schemes in the PSDP that are regional in nature. 4.67. The simplest way to proceed o n this front i s to have a serious review o f the actual responsibilities in relation to spirit o f devolution and, encourage the provincial governments to take back as many o f those functions that have been assumed by the Federal Government during the last three decades either because o f financial constraints or lack o f technical and administrative capacity (actual or perceived) at the provincial level. Devolution o f some o f the functions presently carried out by the Federal Government would reduce the provincial government concern o f becoming redundant in the wake o f large devolution to the district governments, will lift the morale o f the provincial civil service and would enable the federal government to focus more intently o n i t s key strategic functions. In most o f the federal - provincial concurrent l i s t areas, the federal government should remain concerned with providing strategic direction, setting national minimum standards and providing financing to the provinces conditional on their compliance with these standards. Inter- Government Financial Flows 4.68. The adequate funding o f the devolution initiative i s only a part o f the important issue o f inter- government financial flows. W e have argued above for much greater devolution o f responsibility t o the provinces in order to underpin and strengthen the devolution to the district level. But even with the present rather ad hoc distribution o f responsibilities, it i s clear that it i s the provincial/ district level 87 spending on education, health, law and order, provision o f justice, irrigation, rural and urban water supply, and roads that will account for the bulk o f the increase in real public spending in the near future. 4.69. The share o f provincial/ district level public spending i s likely to rise from around 23 percent o f the total public spending in 2002 to over 35 percent in 2007. In terms o f GDP, provincial/ district level spending should increase from 5.6 percent t o over 8 percent over the period, if our assumptions about the fiscal space and priorities materialize. This has major implications for allocation o f national financial resources. At present, the sub-national governments are facing a severe fiscal squeeze. Whereas the National Finance Commission (NFC) has been seriously considering a significant increase in the share o f provinces in the federal divisible p o l l from the present 37.5%, greater flow o f funds would be required to achieve the targets o f the PRSP. Clearly, mechanisms will need to be developed to allow a large portion of the fiscal space created at the federal level to be passed down to the sub-national levels. This will require two things: (1) new NFC and PFCs arrangements which assign much greater formula-based transfers to the lower levels o f governments than at present and (2) putting in place a system o f financial incentives and special transfers to the provincial and local governments so that national priorities can be protected consistent with decentralization and devolution. 4.70. Although institutional frameworks under the NFC and PFC exist to devise appropriate revenue sharing formulae, presently there i s n o such mechanism that can help the federaYprovincia1 governments to induce the local governments to pursue the national/provincial policy objectives. For example, achievement o f PRSP targets, require a sharp increase in sub-national spending in high-priority areas like education, health, water supply and sanitation, etc. I t is, therefore, imperative to strengthen the administrative and technical capacity and financial management systems at the sub-national levels and develop mechanisms to channel sub-national expenditures towards meeting national objectives (e.g. PRSP targets). One such mechanism could be institutionalizing the conditional grants, whereby the federal government provides additional funds to the sub-national governments conditioned upon them meeting some clearly defined targets. For instance, the federal government can finance full or part cost o f an adequately designed scholarship or any other incentive program to promote girls education in the province or district. Care must be exercised to make the grant criteria very transparent to avoid misunderstandings and frictions among the governments. There i s a particular need for targeted national programs to generate employment and income among the poor and for promoting specific national agenda in health and education sectors. Improving Governance and Fighting Corruption 4.71. Poor governance and rampant corruption have been responsible for significant wastage of, and leakage from, public resources (see B o x 4.8). Table 4.2 provides key indicators o f governance for selected South Asian countries. Although based o n perceptions rather than real data, these indicators show that Pakistan performs rather poorly in terms o f accountability, effectiveness o f government, and corruption. All these weaknesses have a strong bearing o n the effectiveness o f public expenditure. Areas where governance has been a particular problem are personnel and resource management; these include staff absenteeism, hiring o f staff on grounds other than merit, frequent transfers, and maldistribution o f staff to the disadvantage o f rural areas. Closely related to weak governance i s the phenomenon o f ghost facilities (especially schools and health centers) and staff, which siphoned a significant amount o f public resources with n o service being delivered. 4.72. One aspect o f weak governance in Pakistan has been the widespread political interference in the planning and budgeting process, which has eroded the quality and credibility o f the government’s expenditure decisions. I n Pahstan, the informal role o f politicians in the budgeting process went beyond the ideal o f establishing the policy agenda for prioritizing the expenditures, malung final decisions o n 88 broad expenditure allocations consistent with these priorities, and monitoring the use o f budgetary resources and associated outcomes (such as via the Public Accounts Committees).” 4.73. Senior government officers operate in a political environment, where absence o f j o b security forces them to rely on their political patrons. The 1973 civil service reforms removed constitutional protection given to civil servants against being prosecuted for unintended adverse consequences o f their official decisions. Another significant change was the shift o f disciplinary proceedings from the Federal Public Service Commission (FPSC) t o the Establishment Division, which, although was later retrieved by the Federal Shariat Court, had left a sense o f insecurity among the civil service. 4.74. The varied and unstructured responsibilities o f government officers; complex, out-of-date rules and procedures; and unclear institutional structures create a system that on the one hand gives government officers immense discretionary powers while on the other i s extremely opaque and within which it i s difficult to determine responsibility and accountability. This situation creates an ideal environment for corruption to flourish. In addition, the financial management and accountability system has broken down (see para 4.80). 4.75. Furthermore, according to anecdotal evidence, politicians frequently interfere in recruitment, transfers, and posting o f government employees. M o r e important, they exert pressure, and often with success, to incorporate projects and schemes in the development program that do not merit inclusion on grounds o f technical, locational, or economic reasoning. Lahore-Islamabad Motorway, Saindak Mineral Project, and Chashma Nuclear Power Plant are some o f the many projects that were undertaken mainly on the insistence o f politicians or for political reasons. Similarly, the Peoples’ Works Program (PWP) and Tameer-i-Watan (TJW) Programs in the PSDP are standard examples o f politically motivated expenditure programs. These programs were included in the PSDP to provide the elected representatives with development resources to undertake projects o f their liking. Although justified on the grounds that these programs offer an ideal mode o f bringing local preferences into the planning and budgeting process, the schemes undertaken under these programs circumvented the regular screening processes o f the government and thwarted the policy f r a m e ~ o r k . ’ ~ ’’ For a variety of reasons, the Public Accounts Committees (PACs) failed to meet regularly, leading to a large backlog of undiscussed audit reports. As a result, two- to three-year delays between when expenditures are incurred and when the PACs take up the related audit reports for discussion are quite common. 56 For example, the rural water supply schemes initiated by the govemment departments ask the benefiting community to sign a memorandum o f understandingwith the department indicating that it will assume responsibility for operating and maintaining the scheme. Rural water supply schemes started under PWP and TIW, however, did not adhere to this sectoral policy, affecting not only the sustainabilityo f these schemes, but also making application o f policy on regular schemes that much more difficult. 89 Box 4.7: Corruption in Pakistan According to a Transparency International perception survey, Pakistan falls among the bottom few countries in prevalence o f corruption. With the establishment o f national and regional accountability institutions, however, and the military govemment’s stated objective of reducing corruption from the society, the situation has improved significantly over the past three years. Presently Pakistan scores about the same as India and significantly better than Bangladesh on the corruption perception index. Nevertheless, there i s significant room for further improvement. A recent survey by Transparency International gives a stark review o f corruption in Pakistan. Based on perceptions and experiences o f individuals, the survey attempts to explore the level o f corruption in seven main departments o f the government. The results o f the survey are reproduced below: Education: About 70 percent o f respondents who interacted with educational institutions cited the existence of irregular methods o f gaining admission. Teachers and members o f the management committee were cited as the most involved. The main causes o f corruption identified were a lack o f accountability and l o w salaries. Health: Sixty-five percent o f all patients visiting a hospital reported irregular admissions, and 96 percent o f those who were admitted said they were victims o f corruption. Hospital staff were identified as the key facilitators of corruption by 65 percent o f the users, and direct extortion was reported in 60 percent o f the total cases o f corruption. Lack o f accountability and monopoly power were quoted as key contributing factors. Electricity: A very high percentage (65 percent) o f users with access t o electricity reported irregular processes in acquiring it; a much higher percentage reported corruption in regular interaction with the department (96 percent). Meter readers and billing employees were identified as the key facilitators; extortion was reported by 72 percent of the victims. A lack o f accountability and l o w salaries o f employees were identified as major contributory factors. Land administration: Only 8 percent o f respondents reported interactions with the land administration department. All o f those who used the services o f the department, however, reported incidences o f corruption. Surveyors and revenue officers were identified as the major facilitators. Bribes were directly demanded in 64 percent o f cases, and a lack o f accountability, discretionary powers, and l o w salaries were cited as the key causes o f corruption. Tax administration: Eight percent o f respondents interacted with the tax department during the past year. Corruption was perceived to be rampant, with 99 percent o f users reporting corruption. A third o f all incidences o f bribes were paid to reduce a customer’s assessed tax. Tax officers and employees o f the department were identified as the major facilitators, and in most cases (61 percent) bribes were extorted. A lack o f accountability was identified as the major reason for the prevalence o f corruption in the tax department. Police: Nearly a third (32 percent) o f all respondents reported using the services o f the police during the past year. All o f those who interacted with the police reported encountering corruption; one in two victims identified the police officer as the key perpetrator. Extortion was reported by 74 percent o f the respondents. A lack o f accountability and l o w salaries were quoted as the major contributory factors. Judiciary: Only 6 percent o f respondents reported using the services o f the judiciary during the past year; almost half of the users interacted in their capacity as a complainant. Ninety-six percent o f users reported encountering corruption; 44 percent o f cases involved bribing court officials. Court officials (41 percent) and judges (24 percent) were identified as key facilitators; in most cases (61 percent), bribes were demanded directly. One-third o f the respondents said that a lack o f accountability was the main factor contributing to corruption in the judiciary. 90 Table 4.2: Key Governance Indicators in South Asia Polity Press Voice & Law & Govemment Graft Score Freedom Accountability Order Effectiveness Corruption Bangladesh 6.00 60.00 -0.20 2.00 -0.54 -0.64 India 9.00 42.00 0.66 4.00 -0.17 -0.39 Nepal 6.00 57.00 -0.06 NIA -1.04 -0.3 1 Pakistan -6.00 57.00 -1.43 3.00 -0.48 -0.79 Sri Lanka 5.00 74.00 -0.23 3.00 -0.44 0.00 Source: Transparency International, Corruption in South Asia, December 2002. 4.76. Recently, the newly elected government has announced it will provide development funds (Rs. 10 m i l l i o n for each member o f the national Parliament). N o t only are there concerns that this program will follow the same route o f PWP and TIW and lead to wasting o f public resources, but in view o f the devolution program, the argument that this program brings local preferences into the budgeting process i s much weaker and may even go against the devolution objectives o f the government. Moreover, unlike the past when the deputy commissioner (generally a federal government employee) coordinated all development activities, including PWP and TIW schemes, in the district, the present district coordinating officer (DCO) i s a local government employee under the district Nazim-a politician. There are therefore concerns that coordination of the M " P A schemes with other local and provincial development activities may not materialize, leading to duplication and further wastage o f development funds. 4.77. Governance i s a multidimensional issue, and improvements in all dimensions are required to improve the effectiveness of government. Given the correlation between increase in corruption and the erosion o f real income o f government employees, an increase in the salary o f civil servants would play an important part in reducing corruption from within the government. In the short run, the govemment should ensure that rules and regulations governing financial transactions are strictly adhered to. Deviations from these rules should be kept t o a minimum, with the need for doing so adequately spelled out. The decision t o allow such a deviation should rest with the responsible agency and not with the politicians (for instance, the Planning Commission in terms o f including development projects in the development program), and the decision should be made only after thorough consultation with all concerned parties. The National Accountability Bureau (NAB) needs to develop a medium-term anticorruption strategy, which should not only significantly improve present methods o f combating corruption, but also formulate a more effective and systemic approach toward monitoring and prevention o f corrupt practices. For any anticorruption strategy t o be effective, i t needs to promote the following features in the government's operations and decisions: transparency, accountability, meritocracy, deregulation, standardization, automation, and community and public involvement. Strengthening Financial Accountability 4.78. Although, Palustan has a reasonably sound accountability system o n paper, in fact weak accountability i s the factor underlying most, ifnot all, the problems that impact the effectiveness o f public expenditure. The procedures governing financial accountability are neither overly cumbersome (given the circumstances of Palustan) nor obviously flawed. Nevertheless, accountability has been a serious problem, for the following reasons: a) Parliamentary and executive control o f the public purse has been ineffective; audit reports o n the annual accounts are heavily qualified; and federal, provincial, and district accounts do not currently meet the expectations o f the Constitution and associated laws and regulations. 91 Substantially strengthening the effectiveness o f the oversight function provided by the public accounts and other parliamentary committees can improve this control. b) The accountability o f the executive at present i s not sufficiently supported by information that would enable i t to become focused on results and outcomes. This improvement in focus can be accomplished by progressively introducing a performance-based management system with incentives that reward outputs and outcomes and hold the secretaries o f govemment departments to account for meeting departmental objectives and performance standards. The responsibilities need to be enforced by strong performance reporting and review functions and intemal control responsibilities. c) The separation o f audit from accounting functions i s s t i l l not complete. The Auditor General o f Palustan (AGP) continues to exercise control over staff o f Pakistan Audit and Accounts Service (PAAS), which include the accounting staff. This implies that the Controller General of Accounts (CGA) does not have clear administrative authority over his own staff. This problem could be effectively resolved by separating P A A S into an auditing cadre, controlled by AGP, and an accounting cadre, controlled by CGA. d) The devolution process has brought to light the institutional wealmesses in the accounting system. The initiative for a more devolved system o f governance for the central, provincial, and local governments and their financial reporting arrangements needs substantial further strengthening in financial accounting, reporting, and auditing and in arrangements for improving public scrutiny. e) The public audit function does not currently meet the standards required by the Constitution and supporting authorities. The audit reports are too narrowly focused and do not contain any information on the efficiency and effectiveness o f program or service delivery. The focus i s mainly o n pointing out irregularities in individual transactions without sufficient attention to significance, and systemic weaknesses, o f financial management system. The legislative oversight o f accountability system too i s weak. The P A C meets too infrequently and lacks the expertise to undertake an effective review o f audit reports. Public scrutiny o f the govemment’s activities and financial transactions i s seriously impaired because o f lack o f timeliness and comprehensiveness o f legislative reviews o f audit reports and because o f poor follow-up to the PAC’s recommendations. The auditing function can be strengthened by focusing it in the short term more directly o n attest audit and improving intemal controls, ensuring i t s independence by malung the Auditor General more independent o f the executive; improving the P A C with institutional support; and making the govemment responsible for following up on the PAC’s recommendations. f) Although a penalty system exists for misdemeanors in public service (such as unauthorized expenditures and misuse or theft o f resources). A range o f sanctions are applicable, including dismissal. The procedures for exercising such sanctions, however, are inordinately cumbersome and lengthy, with the result that they are ineffe~tive.~’ There i s a need to simplify the penalty procedures. g) While dismissal o f staff through regular procedures appears to be virtually unachievable, anecdotal evidence suggests that informal intervention from politicians, at least in some cases, 57 In contrast to the lack o f timeliness o f such sanctions, staff can be transferred to different departments and different areas o f the country at relatively short notice. Transfers are regarded informally within the public service as an effective power of sanction, although this method i s not used officially for this purpose. Instead, i t increases the scope for breakdowns in accountability b y virtue o f the power that the authority to transfer bestows. 92 can result in swift dismissal (and also in appointments that circumvent procedures), with adverse implications for accountability and morale. h) Procedures for recruitment on merit at the senior level appear to be sound, but promotions depend on Annual Confidential Reports (ACRs), which apparently have failed to establish a sound performance system. A very large proportion o f staff i s given an “outstanding” rating by managers, irrespective o f their performance and in contravention to the A C R guideline. Moreover, the completion o f ACRs i s not without bias. i)The recent reversal o f the previous government’s decision to recruit all federal government staff in BPS 11-16 through the FPSC i s a step in the wrong direction. The FPSC’s purview over this recruitment should be restored, or alternatively a foolproof system should be developed by which this recruitment could be made by the government departments without compromising merit. 4.79. The government has recently embarked on a comprehensive reform program to restructure i t s financial management system.58Although significant steps have been taken in this regard (see Box 4.1), the government must carry the program forward and involve the subnational governments, public corporations, and autonomous bodies more intently in the reform process. The continuation, expansion, and deepening o f the reform program can ensure substantial improvement in the government’s financial management system within the next three to five years. The CFAA-Phase-I1 has outlined (among other things) the following key actions for improving the financial management system. 4.80. The immediate actions include (1) enhancing the transparency o f the budget by disclosing fixed- cost and contingent liabilities and the transparency o f the accountability system by publishing annual and quarterly financial reports o f all levels o f government; (2) developing provincial budget implementation systems along the lines o f the federal government’s Revised System o f Financial Management and Control and making the federal and provincial systems fully operational; (3) establishing a strong internal control mechanism for each level o f government; (4) building the capacity o f the auditing and accounting establishments and o f the PAC; and (5) renewing the government’s commitment t o implementation o f PIFRA. 4.81. The medium-term measures include: (1) developing and operationalizing the MTBF; (2) undertakmg c i v i l service reforms; (3) implementing a phased transition from cash to accrual accounting; (4) giving provinces control over their own accounting system by redefining and clarifying the role o f CGA and AGs; (5) ensuring a transition to international auditing standards; and (6) malung public accountability more viable through formal P A C reviews and training the public and the media o n the power and utility o f financial information. These reforms are based on issues and recommendations that appeared in the IMF’s Accountable Fiscal Management Framework and the World Bank-supported Project for Improvement in Fiscal Reporting and Auditing (PIFRA; and the Country Financial Accountability Assessment (CFAA ) -Phase I and II. 93 CHAPTER 5: STRENGTHENINGTHE CIVIL SERVICE K e y Issues 5.1 As discussed in the previous chapters, Pakistan has a unique opportunity to expand and restructure i t s public sector spending in support o f i t s economic and social goals. But while public expenditure o n infrastructure and social development must increase significantly for reviving growth and reducing poverty, this effort could be wasted i f public funds are not utilized effectively. In the previous two chapters, we discussed the principal issues, which normally influence the effectiveness o f public spending, strategic approaches to public interventions, the role o f the various levels o f government, sectoral allocations, budgetary, planning and monitoring procedures and fiduciary controls. But all these aspects o f design and implementation o f public service delivery require functioning institutions including a clear framework o f laws and regulations. A competent, honest and independent civil service provides the backbone o f these institutions. 5.2 In Palustan the strength o f public institutions has been greatly eroded over time. M a n y inter- related factors have contributed to this decline including increasing centralization and non-transparent decision malung, lack of accountability, l o w or declining compensation, and the dilution o f the role o f merit in selection and promotion o f civil servants with undue political interventions in normal functioning o f government departments and agencies. This has contributed to declining technical and professional competence o f officials, and a culture o f risk-averse behavior. 5.3 Some steps have been taken to reform the civil service aimed at improving the performance o f public servants and reducing corruption. In the interest o f merit based recruitment, a more autonomous Federal Public Service Commission (FPSC) was given responsibility for recruitment for all middle level and high level professionals. Promotion panels for higher grades are n o w chaired by the Chairman FPSC and to ensure quality of high grade staff, acceptance to advanced training programs are n o w subject to entry examinations. The first phase o f pay and pension reforms was implemented in January 2002. All federal civil servants including military were given new pay scales which restored 75 percent o f the lost purchasing power after 1994 and also revised pension benefits (lower commutation rates and more actuarially fair commutation factors). The pay reform also reduced the share o f monetary allowances from 40 to 27 percent and de-compressed the pay scales. The ratio o f the highest to the lowest increased f r o m 1:9.2 to 1:9.9 including all monetary allowances. 5.4 Palustan shares the symptoms o f poor performance and l o w accountability stemming from an unresponsive bureaucracy with many other low-income countries. Indeed, the problems o f compressed pay scales offering few incentives for slulled senior managers to j o i n the public sector, the excess o f unslulled lower level employees relative t o senior staff, and the extensive patronage and politically motivated appointments and transfers are found in many settings. However, some o f Pakistan's administrative reform challenges are distinct - as are the opportunities that are available for addressing them. 5.5 The structural rigidities in the federal and provincial administrations stem from strongly vertically-oriented occupational groups. These groups foster the distinctively deep-seated concern with status. Inter-cadre rivalries significantly determine the shape o f federal divisions and attached departments, as structures are created t o meet the needs o f the various groups and cadres for promotion opportunities - rather than o n the basis o f policy priorities or operational needs. 94 5.6 However, government in Pakistan also has some significant advantages. Above all, the public sector i s not bloated in the absolute sense o f too many employees representing an insupportable fiscal burden. The wage bill i s not high in absolute terms, and has been reduced over the last decade as real pay has declined. This i s not to say that it has not crowded out other non-salary recurrent expenditure - but this owes more to the reduction in these expenditures than to an increase in employment costs. Also, this i s not to offer false comfort - public employment i s certainly excessive relative to what the public sector delivers and i t i s undoubtedly bottom-heavy, with a disproportionate number o f employees at the lower levels. The government also finds itself with three ‘low-hanging fruit’ that can be picked at relatively modest political costs. Specifically, government has: some consensus between the Establishment Division and the NREI o n an approach to restructuring; the fiscal space to afford pay reforms; and a well developed, but incomplete devolution agenda. 5.7 This chapter sets out the need for administrative reform, but argues against an over-ambitious or over-elaborate set o f proposals. History suggests that such proposals will generate many opportunities for lengthy reports, but will lead to little sustained progress. Instead, it argues for three steps. First, it suggests that there i s an opportunity to move ahead rapidly o n the creation o f a National Executive Service that would greatly increase incentives both for improved performance and upgrading o f s h l l s through learning and education. The two critical elements o f this reform, o n which there i s seemingly a substantial consensus, are the much greater emphasis o n merit and performance for selection to top grades, including a high bar beyond grade 18, and opportunities for higher education and learning for civil servants especially early in their careers. Second, it suggests sharp pay increases (and monetization o f many o f the large in kind benefitss9) for higher levels o f civil servants so that the serious decompression of pay that has been going o n for decades can be reversed. Finally, it notes that completing administrative devolution, by transferring budgetary responsibility for salaries to the district governments and by progressively passing appointment, promotion and transfer (APT) authorities to the districts, will address a necessary, if not sufficient, condition for local governments to value professionalism more than patronage. 5.8 These or similar steps are central to the achievement o f both the narrower objective o f more n efficient use o f public resources and the broader objective o f “reviving growth and reducing poverty”. I a very basic sense, civil service strengthening i s on the critical path o f the reform process. Structure, not size, i s the problem 5.9 Government in Palustan starts from a significant advantage - the public sector i s not bloated in the sense o f too many employees or excessive fiscal weight. Total public employment i s about 3.2 million (Figure 1) and i s around 2.3% o f the population6’. Historically, provincial employment in the provinces has been driven more by political and patronage concerns rather than identified social and 59 In kind benefits, subsidized housing, free cars, telephones, etc are large part o f compensation at the highest levels. Such in kind benefits are wasteful because they generally lead to excessive use o f facilities provided, free telephones, cars, gasoline, utilities, etc. An equivalent cash compensation w i l l both increase the discretion o f civil servants about the pattem o f their consumption and reduce wasteful spending. 60Total Public Employment includes both state-owned enterprise (SOE) employees and General Govemment. General Govemment refers to employment in “all government departments offices, organizations and other bodies which are agencies or instruments o f the central or local authorities whether accounted for or financed in, ordinary or extraordinary budgets or extra- budgetary funds. They are not solely engaged in administration but also in defense and public order, in the promotion of economic growth and in the provision o f education, health, cultural and social services.” (International Standard o f Industrial Classification o f A l l Economic Activities (ISIC), Series M No. 4, Rev 3- 1990.) By tradition, General Govemment contains six mutually exclusive categories: Armed Forces; Civilian Central Govemment; Subnational Government; Health employees; Education employees; and Police. I t excludes State-Owned Enterprise (SOE) employees: employees o f enterprises that are majority owned by govemment. 95 development priorities. Sanctioned employment in provincial governments grew dramatically in the early to mid 1990s but has since leveled off or reduced slightly. Federal employment i s only about 20 percent o f general government employment (excluding military) reflecting the federal character o f the government. Figure 5.1: The Main Components of Public Sector Employment in Pakistan Powlation (0001 134,790 37.2% TOTAL PUBLIC EMPLOYMENT Total public emp. as % of pop. 2.3% 3,164,900 General govt emp. as % of pop. 2.1% 382,800 2,782,100 I Total subnational Armed Forces (ii) government (iii) government (iv) 1,765,600 Education (v) Education 10,700 Health (iv) Health Police Police I 1 106,700 Subnational Government excluding Government excluding education, health. education, health, B police B police ources: (i) Government of Pakistan, Management Services Wing. Establishment Division, "Annual Statistical Bulletin of Employees of AutonomousISerni- Autonomous CorporationslBodiesunder the Federal Government. 2001-2002; (ii) International Institute for Strategic Studies, The Military Balance 1996'99; data for 1997; (iii) Government of Pakistan, Management Services Wing. Establishment Oiyision. "Annual Statistical Bulletin of Federal Government Employees,2001- 2002"; and "Annual Statistical Bulletin of Employees of AutonomouslSemi-Autonomous CorparationsIBodiesunder the Federal Govemment, 2001-2002"; (iv) Data for2000 provided by the Ministry of Finance. Government of Pakistan; (v) Government of Pakistan, Planning Commission Report, April 2001; (IV)Data does not include 70,000 community health workers hired on one-year renewablecontracts under the "National Program for Family Planning & Primary Health care." 5.10 Excluding state-owned enterprises, general government employment i s some 2.1 percent o f population, which - bearing in mind that international comparisons are only indicative because of practical and conceptual problems relating to the accuracy and consistency o f all public employment data - compares to an average of some 2.3 percent for other low-income countries in the region6' 5.11 The wage bill, at Rs 160 billion, i s about 4.3 percent of Pakistan's GDP. This i s similar to the level o f this ratio in India, China, and Bangladesh although higher than Korea and possibly other similar 61Source: (de Tommaso, Giulio and Amitabha Mukherjee. 2002. "World Bank Administrative & Civil Service Reform Website: Cross-NationalData on Govemment Employment & Wages"; World Bank. (http://wwwl .worldbank.orglpublicsector/civilservice/cross.htm. Access Date: March 24,2003). Data on public sector employment are notoriously slippery and other sources make the regional average rather higher. Schiavo- Campo et a 1 indicate that general civilian govemment employment (Le. public sector employment excluding the military as well as state owned enterprises) has accounted for about 2.5% o f the population in Asia and the Pacific in recent years (Schiavo- Campo, Salvatore and P.S.A. Sundaram (eds). 2001. To Serve and Preserve: Improving Public Administration in a Competitive World. Manila: Asian Development Bank.) 96 emergent economies (Table 5.1). I t i s not high in any absolute sense and it has gone down in the 1990s as real pay declined (notwithstanding recent adjustment). However, the share o f establishment costs in total recurrent spending increased during the last decade as overall spending was squeezed. As a result, non- salary expenditure in the social sectors such as education and health, and O&M spending o n irrigation, roads etc. declined. 5.12 The average public sector wage as a multiple o f per capita GDP at 2.9 for Pakistan i s about the middle o f the range between the high level in Korea, Malaysia, and Bangladesh and l o w level in China, Indonesia, Sri Lanka and the Philippines. (in percent of total population) " F , Ei a 9 g 5 E ; o E: &' i;' 2 Wage bill as % o f GDP 3.7 3.8 4.2 3.0 2.2 5.9 7.7 4.3 5.5 5.1 Average wages as multiple o f per capita GDP 4.1 1.3 N.A. 1.6 4.6 2.0 3.4 2.9 2.3 1.8 Bottom heavy 5.13 Public sector employment in Palustan may not be large in this comparative sense but it i s certainly excessive in many areas in relation to what i t delivers and certainly it became large in the first h a l f o f the 1990s in relation to the fiscal resources (see Table 5.2). Table 5.2: Sanctioned Posts by Province (selected years) 1988189 1993194 1996197 2000 d Punjab 722,916 8 18,647 880,781 888,796 Sindh 285,042 424,974 445,577 457,494 NWFP 177,106 b l 227,679 254,587 1 n.a. Balochistan 98,942 122,921 127,069 1 128,132 ai Figures are actual posts as o f January 2001. b/ NWFP data for 1988/89 has been estimated Figure 5.2: Composition o f public employment 5.14 Also, public employment i s heavily skewed towards lower grades (Figure 5.2), which often typify l o w productivity levels: about 95 % o f the public employment i s in grades 1 to 16. For just the Federal level-this share i s 97 percent. Rigid I1 5.15 kgidities arise because decades o f patronage-based employment and inattention to skill mix in the civil service has undermined the .grades 17-22 11,491 7.378 61,334 29,807 3,894 24,125 ~ work ethic throughout the civil service and has agrades 12-16 13,050 8.309 124,537 38.760 6,371 68,912 Ogrades 1-11 203,138 112,445 702,925 376,813 53,197 741,710 Source: WB staff files 91 created an entitlement culture, in which status considerations and inter-cadre rivalries far outweigh any concerns about service to the public or to government. Various reforms have attempted to counter what are considered to be the self-serving nature o f the bureaucracy. Most radically, the 1973 reforms sought t o simplify the grading system, encouraging mobility between different parts o f the service and encouraging lateral entry for talented individuals from the private sector. The reforms seem to have had surprisingly little impact on the vertically-oriented occupational groups, and despite these reform efforts, occupational groups remain vertically separated, distinct career streams. 5.16 The structure o f federal divisions and attached departments i s unnecessarily restricted by the need to ensure that the various occupational groups that have a claim to positions within that division or department are satisfied. Positions are created to satisfy the needs o f the various groups and cadres for promotion opportunities - rather than on the basis o f policy priorities or operational needs. In some cases set proportions o f various groups have also to be maintained. 5.17 Recently lateral induction on fixed term contracts have been introduced (MP grades systems). However, i t i s unclear h o w many organizations are talung advantage o f this flexibility. I t appears that the proportion o f these positions in the senior management remains low. Generally, the appointments made at senior levels reflect the need to accommodate generalists who tend to be promoted in large numbers because their cadres are larger and thus promotion i s faster. At times the career paths define different years o f experience for different cadres malung promotions slower or faster in them. Also for some o f the technical cadres the promotion path narrows down with seniority and ends before reaching the highest levels. For this reason, senior positions tend to exclude staff who have technical backgrounds in specific disciplines. Mismatch Between Skills and Rewards 5.18 Compensation has become increasingly insufficient to attract skilled entrants at the higher levels but may be too generous at lower levels. The average pay masks a serious problem with pay compression.62 As noted earlier, and as shown in Table 5.3, the present ratio o f the highest to the lowest paid (BPS22:BPSl) i s about 1:lO. In OECD settings, this ratio might be reasonable because o f the structure o f the labor market. At higher levels, public sector pay in both developing and O E C D countries may be compared with slulled private sector positions. However, at lower levels, while in industrial countries the usual comparators for public sector pay are clerical jobs in private companies, this comparison may not be relevant in developing counties where the true employment alternatives o f many public sector workers are in the informal sector o f the economy. "Compression Ratio" i s used here to refer to the ratio o f the highest salary to the lowest on the central government's main salary scale. T h i s definition differs from that adopted by the OECD, which measures wage compression in OECD countries as the mean o f ninth decile salaries divided by the mean o f first decile salaries. T h e OECD approach ensures that a handful o f salaries w i l l not dramatically skew the compression ratio. However, the data quality needed to calculate the OECD compression ratio makes it impractical for extensive cross-country work. 98 Table 5.3: Impact of the January 2!002 Pay Reforms (monthijpay, Rs) - Before pay reform (1994 2001) Base Day a/ I Estimated I Total I Allowances ~~ ade 2 Source: W E3 staff files a! These rates show the pay o f staff who are 2 increments below the scale maximum 5.19 There i s substantial evidence that real compensation for the senior levels has declined over time and the gap between public and private salaries and emoluments has steadily increased. For the highest grade (BPS22), the nominal salary and monetary allowances have increased from Rs. 3,000 per month in the 1950 to Rs.38,000 at present while consumer prices have arisen thirty f o l d over the period. Undoubtedly, the share o f in lund benefits at higher levels has increased but certainly not enough to compensate for the sharp drop in real pay. Real total compensation for the highest c i v i l servants probably decreased over the last half century by at least one-third. 631n the meanwhile, real incomes in the private sector have at least doubled reflecting the general increase in per capita incomes. Even if one assumes that the gap between the top management positions in the private and the public sector was small initially, it would appear to have grown several fold. The increasing use of M P grades, originally meant for top managerial positions in the public corporations and approximating more closely to private sector pay, to attract managerial and technical s k i l l s to govemment also indicates the problems with the present levels of compensation. Finally, the fact that there has been a historic decline o f the pay compression (excluding in kind allowances), estimated at 1:4664in 1962, also seems to confirm the hypothesis that pay at the lowest levels has increased significantly over time while i t has declined at the highest levels.65 5.20 This pattern o f compensation i s undoubtedly contributing to two significant problems. First, relatively generous remuneration at lower levels increases the pressure for patronage-based employment o f unskilled staff. Second, below market rates at senior levels are unlikely to be attracting the quality o f entrant that i s needed at the highest levels o f govemment. 63 See Hasan, 1998 (p.359). 64 Table 7.8, "Govemment That Works: Reforming the Public Sector in Bangladesh" 1996. World Bank. Dhaka. Indeed if one assumes that there has been no real pay decline at the highest levels, the lowest level pay must have increased very sharply to explain. 99 Box 5.1: Monetary Allowances and in kind Benefits Before pay Post Dec. R s (million) reform 2001 Annual Base Pay 11,728 18,774 Annual Allowances 7,765 6,85 1 Total 19,493 25,625 Notes: This refers to federal ministries and attached departments - it does not include autonomous bodies Table 5.4 shows that January 2002 pay reforms provided consolidation o f monetary allowances within base pay. The table shows the impact o f consolidation on the federal civil service wage bill. The impact on the pr will have been broadly similar. Even after the January 2002 pay reforms, in kind benefits for BPS-22 (value of accommodation, transport, etc.) were estimated to be much higher than the basic pay. Further, monetization o f these allowances and in-kindbenefits i s highly desirable as has been done in the State bank of Pakistan and i s under consideration in the Central Board o f Revenue. Persistent Patronage 5.2 1 Politically-motivated patronage appointments to the civil service and transfer decisions have had a damaging impact on performance. For this reason, the decision o f the Federal Cabinet in early February t o recruit grades 11-16 without involving the Federal Public Service Commission sends a disturbing signal. Delegation o f lower level recruitment to line departments, per se, i s not problematic. Most Public Service Commissions delegate recruitment decision for junior and middle-level t o line departments. However, generally they do this while maintaining tight regulatory supervision over the recruitment process.66 Thus while they are not themselves recruiting, they maintain full responsibility for the quality of the recruitment process. 5.22 The recent decision in Palustan seems to remove responsibility for the process as well as the function from the Commission. Since the original decision was made with the intention of providing some safeguards against identified problems o f patronage in a range o f more sensitive federal bodies, it begs the question as to why this will not simply recur. However, in addition this will send an unfortunate signal to the provincial Commissions. I nNWFP, for example, PSC oversight has been extended down to grade 11 in some corruption-prone departments (mirroring the equivalent reform at the federal level). I t i s improbable that such arrangements will remain in force given the federal changes. 5.23 In addition to appointments, political pressures are also behind the continuing high level o f transfers o f senior staff. While data o n the rate o f transfers o f senior staff are not available at the federal level, the problem i s clearly chronic at the provincial levels. In Sindh, the average tenure o f Secretaries i s some 15 months. In NWFP, it i s 7.5 months. 66 See Polidano, Charles and N i c k Manning. 1996. Redrawing the Lines: Service Commissions and the Delegation o f Personnel - Management. Managing the Public Service Strategies for Improvement Series. London: Commonwealth Secretariat. 100 5.24 Similar problems pertain at district levels. The Province retains the ability to transfer senior staff in any o f the provincial occupational groups (i.e. above grade 16 and thus not transferred in December 2002 to the districts). T h i s i s a highly politicized arrangement to influence transfers throughout the district. By the threat o f an unpleasant transfer, or the promise o f an attractive one, the senior staff member can be put under pressure to accede to a transfer o f the junior staff. A Pragmatic Approach 5.25 The track record o f incomplete and ultimately ineffective civil service reforms in Palustan suggests that there i s little advantage in devising yet another large and elaborate strategy. T h i s report suggests instead that some momentum towards reform i s created by taking advantage o f three fortuitous circumstances. The government finds i t s e l f in the unusual position o f having: some consensus between the Establishment Division and the NRB on some restructuring; the fiscal space to afford pay reforms; and a well developed but incomplete devolution agenda. 5.26 By seizing the low-hanging fruit that these provide, government can set in motion some movement towards a longer-term reform agenda. Other steps can be taken that will achieve short-term efficiency gains and send significant signals at n o political cost. For example, on recruitment, the FPSC should be given the unambiguous responsibility for oversight o f the recruitment process at all grades, even if the decision stands that i t should not i t s e l f undertake direct recruitment for positions below BPS16. 5.27 Similarly, since present compensation levels are inadequate at senior levels, use o f fixed term consultants with maximum t e r m o f say three to four years at near market salaries should be greatly encouraged both for professional and managerial positions. Liberal use o f consultants i s especially needed to strengthen the federal ministry o f finance (including the debt office and the economic adviser’s wing), the federal planning commission and the ministry o f commerce as well as planning and finance departments o f the provincial governments Seize the emerging consensus 5.28 The NRB has proposed the creation o f five new broadly-based services: . . 9 The Federal Civil Service . The National Executive Service . The Provincial C i v i l Service, including the Provincial Executive Service The District C i v i l Service The Tehsil Municipal service 5.29 Within the NRB proposal, both the Federal and Provincial C i v i l Services will continue to be divided into occupational groups below grade 18. The National Executive Service (NES) would, however, be an All Palustan Service comprising a pool o f officers selected on merit from amongst officers o f the Federal, Provincial and District Services for critical policy and management positions beginning at Grade 19. A maximum o f 20 percent o f the entrants would be recruited laterally from the private, NGO and academic sectors through a competitive process. The N E S posts would total between 1,000 and 1,500 and would encompass all grade 19 to 22 positions in the federal secretariat, heads o f important departments in the federal government, and senior positions in provincial and district governments (e.g. Provincial Chief Secretaries, some posts o f provincial secretaries and additional secretaries, and DCOs in District governments). 101 5.30 The NRB propose that the N E S should have three broad streams o f professional specialization: . Economic Management Social Sector Management General Management 5.31 The compensation package for N E S will be attractive. It i s proposed that N E S officers (and officers from the proposed parallel Provincial Executive Service) would be required to undergo specialized training in foreign and domestic universities and later in the newly created National School o f Public Policy. 5.32 The Establishment Division’s proposed reforms are similar. The Establishment Division has also proposed the creation o f three streams o f professionalized civil servants to some 1,000 policy formulation positions in the Federal Government. These civil servants would be competitively selected and specially trained, and would enter at grade level 19 with up to 20 percent recruited laterally. They would be attractively compensated - the Establishment Division i s proposing to double the basic salary plus relevant grade allowances. Establishment Division also proposes three broad professional streams for these senior staff. 5.33 These reform proposals are appropriate and can be a significant contribution towards increasing institutional capacity at the higher levels o f government. It i s not clear, however, that i t i s necessary to create a dual structure o f compensation. Compensation for grades 19 to 22 should be raised substantially and combined with monetization o f most in-kind benefits. Those selected for N E S will have a fast track advantage for promotion so there need not be separate pay scales. The key t o both proposals i s the National Executive Service, with a high bar before grade 19, sufficiently robust t o deter any attempt to use seniority as a justification for access, and a system o f very broad occupational groups beyond that point. The proposals will encourage improved performance and greater s h l l s development because promotion beyond grade 18 will not be at all automatic. The pursuit o f the professionalization objective will be greatly helped by the full development the new National School o f Public Policy that has been created with the following functions: . . T o provide training and research in public policy, law, economics, finance, and management science T o provide consultancy and advisory services for the government T o award degrees in Public Policy, public sector management and related disciplines 5.34 The broad agreement between the NRB and the Establishment Division provides a rare opportunity to move ahead on a pragmatic first step. The National Executive Service i s undoubtedly the priority. Over time the need for provincial executive services may emerge. Government should seize the initiative and decide the outlines o f the N E S quickly. W e also suggest that the reform should be comprehensive that i s it should create the three agreed occupational groups immediately. Managing the Transition Address pay 5.35 For the reasons noted above, the retrenchment o f large numbers o f employees i s not an overriding priority. Large-scale retrenchment may not, in any case, be politically feasible at this time because o f slow j o b creation in the rest o f the economy. However, the government finds itself in a situation when significant pay reform, combined with highly selective future hiring, can be achieved largely through fiscal space and attrition in the l o w priority civil service groups. 102 5.36 There i s a prima facie strong case for significantly adjusting upwards the compensation for higher managerial and professional grades, say grades 19 t o 22. The numbers involved are small. For instance, Federal employees in the grades 19 to 22 are only about 3,200, less than 1 percent o f total Federal employment. The share i s somewhat higher in the provinces. But overall these grades account for less than 3 percent o f the government wage bill. So financial costs o f say a 100 percent real pay increase over the next five years would be only 0.1 percent o f GDP. A substantial part o f this increase could be financed fkom natural attrition o f lower grade staff focused o n areas where employment i s excessive and services are n o longer needed. Restraints on recruitment coupled with attrition might be easier politically than retrenchment as a method for creating fiscal space, but it i s recognized that the de-compression in compensation that i s being recommended here i s nevertheless politically challenging. It i s however urgently needed if high-level managerial and professional slulls are to be attracted to the public sector. 5.37 It i s much less clear whether there should be a general pay increases at lower levels without loolung at j o b content, s l u l l s required, and the comparator salaries in the private sector. The recent moves both at Federal and provincial levels to selectively give large special pay increases (for instance to tax officials and doctors in rural areas) also suggest the advisability o f tailoring compensation to the nature o f the responsibilities, hardship o f posts and reducing the incentives for corruption. There are other cases, primary school teachers, lady health workers, and junior police officials who are extremely poorly paid because their jobs are graded low. Detailed comparator studies o f private sector equivalents will, o f course, be a necessary to guide future pay reforms for all grades. 5.38 Whether created by attrition and restraints on recruitment or otherwise, the opening o f the fiscal space can provide an opportunity for reversing the squeeze in real salaries and de-compressing the compensation differentials and further consolidating monetary allowances and in lund benefits into base pay (See B o x 5.1). However, before a new compensation policy can be formulated the future pension liability must be assessed and actions taken to avoid a crisis in the future. The Ministry o f Finance has initiated a comprehensive review o f reform options for the pension system. Such a review should identify alternatives for controlling pension expenditure and designing and financing the new system, including proposed methods for data collection and for an actuarial valuation o f cash f l o w and l i a b i l i t i e ~ . ~ ~ 5.39 Provided future pension liability can be contained and civilian employment growth can be kept down to say 1 percent per annum, there should be room for increasing real salaries significantly over the next five years. The challenge will be to use the fiscal space, o n the one hand, for restoring compensation closer to private sector for the higher grades, and o n the other t o adjust pay scales for groups for instance primary school teachers, policemen, and lady health workers who are particularly poorly paid. Flat rate increases, that do not decompress salaries or provide motivational important occupational groups, will not serve the objectives o f the c i v i l service reform. Complete devolution 5.40 The unfinished administrative decentralization agenda provides a further opportunity for government to initiate movement on an otherwise intractable issue. Completing devolution means, first and foremost, transferring budgetary responsibility for salaries t o the district governments. Thus salary budgets should be transferred to the districts via account IV. In theory, from that time, they can select 67 Further work should include a review of the proposal by a Commissioner o f the Securities and Exchange Commission o f Pakistan that would move employees o f autonomous bodies onto privately managed defined contribution pension schemes. T h e scheme i s a little unclear, without a clear definition o f coverage or o f the fate o f existing (largely defined benefit) schemes. Apparently the proposed law seeks to replace all existing schemes with a mandatory defined contribution scheme to be administered by Pension Administration Companies set up for the purpose. The Pakistan Society o f Actuaries has written to the govemment expressing some concern that the above and other significant issues have not been addressed. 103 more o f one type o f staff or s k i l l and less o f another as once the salaries are in account IV, then the interim PFC awards to the districts require a single l i n e item transfer that i s up to the districts to allocate - with n o reserve powers to the provinces to interfere.@ 5.41 This i s not a sufficient condition for local governments to value professionalism more than patronage, but it i s undoubtedly a necessary one. Local governments at all levels will only start talung their responsibilities for workforce management seriously as and when they consider themselves to be fully the employer o f their staff. 5.42 Furthering devolution might also assist in reducing the scale and frequency o f politically- motivated transfers. One source o f the problem i s the willingness o f higher levels o f government to intervene in the affairs o f lower levels. There are many reasonable arguments for maintaining a federal presence at the provincial level, and similarly for maintaining a federal and provincial presence in the districts. However, this provides a point o f entry for manipulation and, particularly for the new district governments, leaves them subject to provincial pressures to transfer staff. As feasible, the appointment, promotion and transfer (APT) authorities for staff should be progressively passed t o the districts. Currently, staff at grade 17 and above are appointed, promoted and transferred by the provincial governments. APT authorities for grades 17 and 18 can be transferred to the districts within a year, although this will require some clarification o f the role o f the Provincial Service Commissions or, more logically, by the creation o f District Service Commissions. 5.43 In addition, minimum times in post for senior federal, provincial and district staff should be clearly specified with a requirement that public explanations be provided if staff are to be moved before they have completed the minimum. By themselves, however, such specified minimum tenures are unlikely to be sufficient to withstand the considerable political pressures for inappropriately rapid transfer^.^' Also, as noted in the discussion o f devolution, one pretext for provincial micromanagement in the districts can be removed by providing any increases in provincial-district fiscal transfers in the form o f specific purpose grants. T h i s would counter the argument that, since transfers are made o n a block grant basis, provinces must resort to various methods t o ensure that districts address federal and provincial policy priorities. T h e SLGO establishes as a "function and power o f the Z i l a council" to "approve the proposals o f the District Govemment for changes in the number o f posts o f officials and employees o f the decentralized offices o f District Administration and Taluka Municipal Administration and Town Municipal Administration as part o f the budget statement" (Art. 39). 69 A requirement that staff remain in post for at least 3 years i s already included in the Govemment Rules o f Business (Rule 35 (ii) Schedule 9). T h i s i s not implemented however. 104 Box 5.2: Incomplete Devolution On August 14,2001, District, Tehsil and U n i o n Councils were created as administrative entities. These entities inherited a complex array o f staff, with complicated and overlapping authority relationships. The consequence of the creation o f the districts i n August 2001 was a transfer o f responsibilities for various fimctions - it was n o t initially a transfer o f the ‘employer’ role f r o m provincial to district governments. The newly created districts found that they had inherited three groups o f staff: 1. They found themselves with some staff f r o m federal employment groups (primarily District Management Group (DMG) and the Audits and Accounts Group). 2. In addition, employees who had earlier belonged t o the rural district councils were posted t o the districts. 3. The bulk o f their employees came f r o m provincial employment groups - particularly Public Health Engineering, Rural Development, Local Government, and Education. M o s t provincial groups had some staff in the districts. The “employer” o f the f i r s t group remained undoubtedly the federal government. The “employer” o f most of the second group became unambiguously the district, although some staff at lower grades were assigned as administrators t o the U n i o n Councils (maximum o f 3 staff per UC at Grade 9). The employer for the third group i s distinctly ambiguous.’ Performance evaluation responsibility for staff was transferred t o the district as soon as they were created, by the simple logic that Annual Confidential Reports (ACRs) are initiated at the same level o f government as the administrative location o f the staff. District officers (DOs) heading each o f the functional service categories in district government are to have their ACRs initiated by their respective EDOs, countersigned by the DCO. B e l o w t h i s level, ACRs are initiated by an employee’s immediate superior. Career management responsibilities have just been transferred. Legally, a l l the formerly-provincial staff below grade 16 n o w have their appointment, promotion and transfer (APT) decisions made through the districts. Senior staff o f the districts (Executive District Officers (EDOs) and other staff at above BPS16) will remain subject to the province or the federal govemment for appointment, promotion and transfer decisions. The ambiguities arise because budgetary responsibility for salaries has not yet been fully transferred. Currently, the staffs o f the Public Health Engineering, Rural Development, and Local Government employment groups - as w e l l as other formerly provincial employee groups posted to the districts - are paid by the province through the mechanism o f Account I(a provincial account). Under this arrangement, sanction i s sought in advance f r o m the province for the release o f the funds for their salary payments. Thus, the salaries are effectively paid f r o m the provincial budget. Until district governments consider themselves t o be unambiguously the employers o f their own staff, then they will have few incentives t o manage their workforce effectively. 5.44 If education and health expenditures are to be increased and MDG outcomes achieved (see Chapter 3), there will necessarily be a significant increase in the number o f teachers and health care personnel. Clarifying the employment authorities for district governments, and gearing up their capacity for the significant challenges that lay ahead, will be crucial. 105 REFERENCESD3IBLIOGRAPHY Azfar, O., Kahkonen, S. and Patrick Meagher. 2001. “Conditions for Effective Decentralized Govemance: A Synthesis o f Research Findings”. Center for Institutional Reform and the Informal Sector, University o f Maryland. Government o f Pakistan. 2001a. A Debt Burden Reduction and Management Strategy. Debt Reduction and Management Committee, Finance Division. March 2001. (Pakistan Debt Report). Government o f Palustan. 200 lb. Interim-Poverty Reduction Strategy Paper (I-PRSP). Jointly prepared by the Ministry o f Finance and Planning Commission, Islamabad. November 2001. Government o f Palustan. 2001c. Summary of the Ten Year Perspective Development Plan 2001-11 and Three Year Development Program 2001-04. Planning Commission, Islamabad. September 200 1. Government o f Pakistan. Economic Survey. Various Issues. Halcrow and ARCADIS. 2002. “Pakistan Water Resources Strategy Study”. ADB TA 3130 PAK, (Draft Report), June. Hasan, Parvez. 1998. Pakistan’s Economy at the Crossroads - Past Policies And Present Imperatives. Karachi. Oxford University Press. International Monetary Fund. “Accountable Fiscal Management Framework”. Washington, DC. Shah, Anwar. 1996. “Fiscal Federalism in Pakistan: Challenges and Opportunities”. World Bank. Washington DC. State Bank o f Palustan. 2002. Annual Report 2001-2002. World Bank. 2002a. Pakistan Development Policy Review: A New Dawn?. Report No. 23916-PAK. South Asia Region. Washington, DC. World Bank. 2002b. Pakistan Poverty Assessment - Poverty in Pakistan: Vulnerabilities, Social Gaps, and Rural Dynamics. Report No. 24296-PAK. South Asia Region. Washington, DC. World Bank. 2001a. Reforming Punjab’s Public Finances and Institutions. Report No. 20981-PAK. South Asia Region. Washington, DC. World Bank. 200 lb. “Methodology for Functional Reviews”. Paper produced for the Russian Federation Government. Washington, DC. World Bank. 1998a. Pakistan: A Framework for Civil Service Reform in Pakistan. Report No. 18386- PAK. South Asia Region. Washington, DC. World Bank. 1998b. Pakistan Public Expenditure Review: Reform Issues and Options. Report No. 18432- PAK. South Asia Region. Washington, DC. World Bank. 1998c. Pakistan - Towards A Health Sector Strategy. Report No. 16695-PAK. South Asia Region. Washington, DC. 106 World Bank. 1995. Pakistan Poverty Assessment. Report No. 14397-PAK. South Asia Region. Washington, DC. World Bank. 1994. Pakistan - Irrigation and Drainage: issues and options. Report No. 11884-PAK. South Asia Region.. Washington, DC. World Bank. 1968. Water and Power Resources of West Pakistan: A Study in Sector Planning. John Hopluns Press. 107 Annex A Page 1 of 4 GOVERNMENTOF P A K I S T A N FINANCEDV I S I O I N 14,2001 MARCH REDUCTION AND MANAGEMENT DEBT STRATEGY 1. Pahstan faces grave debt issues, which have been in the malung for decades and threaten the economic future o f the country. Moreover, the country faces not one but two distinct debt problems: the overall public debt problem and the external debt problem. The external debt problem has greater immediacy because o f the concentration o f large external debt service payments, amounting to nearly US$2 1 billion over July 2000 - June 2004, which cannot be met without exceptional financing from IMF and other multilateral institutions and further debt rescheduling from the Paris Club. 2. Pakistan’s total external debt obligations (including debt obligations to residents in foreign exchange) increased from less than US$ 10 billion in June, 1980 to US$ 20 billion in June, 1990 and peaked at US$43 billion in M a y 1998, just before the imposition o f economic sanctions forced Pakistan to freeze all individual foreign currency accounts. Even after these large-scale foreign exchange liabilities (totaling approximately US$ 9 billion) were extinguished, external debt and foreign obligations, including those to residents, stood at US$ 37 billion at end-2000, equivalent to nearly 300 percent o f foreign exchange eamings (the normal sustainable level i s 200 percent). External debt, narrowly defined, stands at US$ 34.5 billion. The nominal value o f overall public debt grew from Rs. 155 billion in 1980 t o Rs. 800 billion in 1990 and rose to Rs. 3,570 billion at end-2000, raising the burden o f public debt from 66 percent o f GDP in 1980 to well over 100 percent at present. In terms o f a more relevant indicator, public debt i s n o w around 625 percent o f government revenues compared to 400 percent in 1980 (whereas the sustainable limit i s approximately 350 percent). Pakistan has a higher public debt burden, by a large margin, than any large developing country. Moreover, the above figures do not include any estimate o f contingent liabilities o f the Federal Government, which comprise guarantees to P P s and other private and public bodies; and large unfunded losses o f public corporations including WADPA, KESC, Pakistan Steel Mill as well as nationalized banks and public financial institutions. 3. Given the absolutely unsustainable level o f the country’s debt and urgent need to reduce the debt burden, the Chief Executive announced in his public address on December 15, 1999 the establishment o f the Committee on Debt Reduction and Management (Debt Committee) headed by Dr. Parvez Hasan, a noted Pahstani Economist and former director o f the W o r l d Bank. The Debt Committee’s task was t o analyze (a) factors underlying debt growth, (b) impact o f high public debt interest payments o n Government development and social spending (c) implications o f large contingent liabilities o f the Government for future debt and fiscal management; and (d) short-term consequences o f large gross borrowing requirements for balance o f payments management. The Committee was also asked to review the existing framework for debt contracting, recommend medium and long-term goals for the reduction o f burden o f public as well as external debt and specify institutional arrangements for a debt management system. The Committee has worked for nearly a year. I t submitted i t s interim recommendations t o the Minister for Finance in April 2000, and has n o w finalized i t s Summarv Report. The Chairman o f the Committee briefed the Cabinet on November 1,2000 and February 15,2001 upon submission o f the draft Summary Report. 4. The Committee has analyzed the evolution o f debt burden in Palustan and found that extraordinarily high public debt and very large external debt have stemmed from the following factors: (a) Large and persistent fiscal and current account balance o f payments deficits. Annex A Page 2 of 4 (b) Imprudent use o f borrowed resources (such as wasteful government spending, resort to borrowing for non-development needs, undertaking o f l o w economic priority development projects and poor implementation o f foreign aided projects) resulting in l o w productivity o f resource use. (c) Weakened debt carrying capacity in terms o f stagnation or decline in real government revenues and exports. (d) Rising real cost o f government borrowing, both domestic and foreign. 5. The work o f the Debt Committee has confirmed that the high and growing debt burden i s a fundamental cause o f many o f Palustan’s problems, including the slowdown o f overall economic growth, increase in poverty incidence, the l o w levels o f social welfare indicators, declining investor confidence in the long-term prospects o f the country and the recurrent financial crises. The Committee has especially highlighted the close links between high debt burden and slow economic growth. First reduced net external resources are constraining investment. Since reliance on net external flows in the past was totally unsustainable, the Committee has concluded that national savings must finance 90 percent o f investment at least for the next decade. Second, large government domestic borrowing i s keeping real interest rates high and discouraging private investment. Third, private investment i s also being adversely affected by the shortage o f complementary infrastructure and inadequate s h l l development, attributable directly to shortage o f fiscal resources. 6. Since the debt problem has arisen because o f poor economic decisions, postponed reforms and weak governance over a sustained period, there i s n o easy remedy for debt burden reduction. The satisfactory resolution o f debt issues will take both time and aggressive policy action. It will also involve painful choices. The Debt Committee recognizes that, desirable though it may be, reduction o f debt to a sustainable level cannot be the only economic goal. There i s also an urgent need to revive economic growth, which on a trend basis has been only around 4 percent per annum over the last five years. The Committee has therefore stressed that Pakistan needs a comprehensive and integrated economic revival and debt reduction strategy. The twin goals o f this strategy should be a notable reduction in the debt burden and a significant increase in the rate o f economic growth over the medium term. This implies that in the short-run (that is, for up to two to three years), some difficult tradeoffs between debt reduction and economic growth will be unavoidable. 7. The present Government has already been vigorously pursuing many elements o f the strategy recommended by the Debt Committee, notably, reducing fiscal deficits, increasing government revenues, expanding exports, accelerating privatization and strengthening the o i l and gas and information technology sectors. It i s also attaching high importance to maintaining a program with the IMF, with a view t o moving towards a three-year Poverty Reduction and Growth Facility (PRGF), which will provide the basis of exceptional assistance from the W o r l d Bank and Asian Development Bank and further rescheduling from the Paris Club till 2004. 8. Nevertheless, the Committee’s work has provided many fresh insights concerning the nature o f Pakistan’s debt problem and the economic challenges Pakistan faces in this decade. These include: i) A focus on productivity enhancement through both structural reforms and improved governance i s crucial for reviving economic growth, given financial constraints. ii) W h i l e steady reduction in the fiscal deficit and sharp increase in government revenues are critical, the pattern o f fiscal adjustment and effectiveness o f public spending will determine whether government investment in human and physical capital can increase significantly and whether the economic and social returns to public borrowing w i l l become significantly Annex A Page 3 of 4 positive. The balance between development and defense expenditure needs to be restored and the Government must eliminate borrowing for non-development spending, which i s lowering national savings and compounding the debt problem. iii) Real costs o f domestic and foreign borrowing by the Government have risen to very high levels in recent years and need to be brought down sharply. iv) The huge public debt burden poses as big a threat to Pakistan’s long- term financial and economic health as does the large external debt. 9. The Debt Committee i s confident that with strong financial discipline, forceful structural reforms and improved governance, economic growth can be revived to a soundly based 5.5 percent per annum by 2003-04. At the same time, with adequate international support, the real external debt burden can be reduced by 30 percent and the need for exceptional assistance from the IMF and Paris Club rescheduling phased out by 2004. I f Palustan can safely navigate the turbulent economic waters during the difficult transition period o f the next three or four years, and if major unanticipated exogenous shocks do not occur, i t will have laid the basis o f high, self-sustaining and equitable economic growth, a combination which it has never achieved in the past. Pakistan’s population growth rate i s expected to drop gradually from the present level o f 2.2 percent to 1.8 percent per annum by 2009-10, reflecting the demographic transition that has commenced. There are thus good prospects o f achieving a per capita GDP growth rate o f over 4 percent per annum in the second half o f this decade -higher than in the 1980s and with a much better quality o f growth. 10. The Committee has examined the issues involved in debt default and debt reduction. For clearly stated reasons, the Committee believes that default i s not a real option. Debt reduction also faces many practical problems. Takmg a pragmatic approach, therefore, the Committee has not ruled out seelung softer terms at Paris Club reschedulings. It hopes, however, that Palustan can resolve i t s external debt problems while maintaining and improving i t s creditworthiness. 11. The Government has accepted the broad thrust o f the Debt Committee’s recommended strategy and has decided on a number of steps to implement i t s key recommendations. T o this end, the following steps are being taken to formalize a Debt Reduction and Management Strategy: 1) Public and external debt reduction goals will be made an integral part o f the Government’s economic revival and poverty alleviation strategy. Specifically, the Government has adopted the goals o f reducing external debt burden to 200 percent o f foreign exchange eamings by 2005 and reducing public debt burden to 350 percent o f government revenues by 2010. These targets will become part o f the ten-year macroeconomic framework being prepared by the Planning Commission. ii) The building o f gross foreign exchange reserves to a level o f U S $ 5 billion by mid-2004 will be a part o f external financing plans. iii) The attainment o f the admittedly ambitious revenue and export growth targets over 2000-2004 (60 percent and 50 percent respectively) will remain a top Government priority. iv) The Government will also aim at eliminating the imbalance between current revenues and current expenditure (that is, borrowing for non-development purposes), which at present i s 3.7 percent o f GDP, by 2004. Consequently, the overall fiscal deficit will be reduced t o 3 percent of GDP by 2004 and further t o 2 percent by 2010. v) T o facilitate the above goals, defense spending will be kept constant in real terms, for the next four years. Annex A Page 4 of 4 vi) As the overall fiscal deficit i s gradually reduced, the share o f interest payments in total expenditure will decline from 33 percent in 2000 to 16.4 percent by 2010. The fiscal revenues thus freed from lower interest payments and defense will be used to sharply increase the share o f social and development spending in total expenditures. vii) Bringing down the burden o f external debt significantly by 2005 will permit an increase in normal extemal borrowing thereafter. However, net reliance on extemal resources will be kept to less than 10 percent o f total investment during 2000-10. This compares with net borrowing from extemal resources o f over 20 percent in the 1980s and over 30 percent in the 1990s. viii) The Government will make utmost efforts to obtain U S $ 6 billion in exceptional assistance from the IMF, World Bank and ADB o n soft terms in order both to avoid debt default and reduce the onerous burden o f extemal debt payments. ix) During the next four years normal term borrowing from a l l sources, including IBRD and ADB, will be kept to a minimum. x) The Govemment will aim at reducing domestic borrowing costs in real terms to three to four percent per annum, from the average o f nearly 9 percent during past two years, by sharply reducing i t s net domestic borrowing and restructuring i t s borrowing plans. xi) The Govemment will be extremely selective in undertaking fresh foreign aided projects in the short to medium term and apply strict criteria o f financial and economic viability for screening these projects. Terms and conditions o f foreign loans will also be disclosed to the public at large. xii) Borrowing for local currency from international sources will be phased out in the next two years. Thus, development projects requiring mainly rupee funding will rely on indigenous financial resources and local consultants. xiii) T h e Govemment will set up a Debt Policy Coordination Office (DPCO) in the Ministry of Finance, which will report to the Cabinet every year on progress in attaining debt reduction goals, provide broad leadership o n debt management issues to all parts o f govemment and closely monitor and evaluate contingent fiscal liabilities. Annex B Page 1 of 20 MEDIUM-TERM FRAMEWORK Technical Annex 1. T h i s annex provides technical information about the medium-term macroeconomic and fiscal framework in the two scenarios in Chapter 2, and a sensitivity analysis o f the base case. 2. The medium-term budgetary framework presented in this report consolidates the budgetary accounts o f both Federal Government and the Provinces. The presentation o f a consolidated fiscal framework i s for illustrative purposes and allows for an assessment o f the total fiscal impact on macroeconomic balances and expected progress towards public debt sustainability. The consolidated fiscal space estimates allow for an assessment o f the room for increased spending in priority areas by all levels o f government with progress towards the government’s goals o f growth facilitation, poverty reduction, and social development. Drawing on this framework we present findings about the total countrywide resource availability, illustrate the public expenditure level and pattern for the country as a whole and highlight future expenditure options. 3. In order to operationalize the consolidated budgetary framework for use by the central and provincial governments, the authorities would need to break i t down into separate real budgetary frameworks for the central government, the provinces o f Punjab, Sindh, North West Frontier, and Balochistan, and the areas o f A X and the Northern Areas. This will include making overall and sector- specific spending projections at the provincial and district level consistent with the consolidated totals we present and the associated necessary transfers from the center to the provincial governments. I. THEBASELINE SCENARIO 4. The baseline scenario for macroeconomic policies for the period FY 02/03 - FY 06/07 assumes a continuation o f the reform-oriented policies pursued by the GoP in recent years. The gains seen in achieving macroeconomic stabilization and reducing fiscal imbalances are expected to be consolidated and strengthened further. However, i t also envisages a concerted approach to reach the goals for poverty reduction and facilitation o f economic growth. The macroeconomic framework for the first projection year, FY 02/03, i s the same as in the PRGF program agreed with the IMF. 11. K E Y ASSUMPTIONS FOR THE PROJECTIONS A. Overall macroeconomic framework (Table A.1): Real GDP growth (at market prices) r i s e s from 4.4 percent in FY 01/02 to 5.9 percent o f GDP in FY 06/07. Inflation stays at 4.0 percent throughout FY 02/03 to FY 06/07, up from 2.7 percent in FY 01/02 w Gross national savings r i s e from 16.4 percent o f GDP in FY 01/02 to17 percent o f GDP in FY 06/07, reflecting continuing fiscal consolidation Investment (gross fixed capital formation) rises from 12.3 percent o f GDP in FY 01/02 to 16.6 percent o f GDP in FY 06/07 with proportionally similar increases in investment by the public and private sectors (the latter includes public enterprises in the statistics) w The ICORs (Incremental Capital Output Ratios) decline slightly from the F Y 0 2 level (3.1) to 3.0 by FY07. The three-year average ICORs decline from 3.5 to 3.1. Constant real effective exchange rate - which in the relatively l o w inflation environment projected here implies gradual adjustment in the nominal exchange rate in small steps. Annex B Page 2 of 20 B. Fiscal projections (Table A.2): . 5. f sufficient fiscal space are: The most critical assumptionsfor creation o Revenue rises from 16.6 percent o f GDP in FY 01/02 to 18 percent in FY 06/07. The revenue . increase comes entirely from tax collection, in particular tax collection by the C B R (1.4 percent o f GDP). Contingent liabilities are contained and sharply reduced, reflecting in particular improvement in . operational performance o f public enterprises. Payments o f contingent liabilities decline from 1.4 percent o f GDP in FY 01/02 to 0.3 percent in FY 06/07. Defense spending i s constant in real terms at the FY02/03 level. Interest payments decline from 6.6 percent o f GDP in FYO1/02 to 4.6 percent o f GDP in FY06/07. T h i s decline reflects mainly the positive public debt dynamics in the overall - favorable - macroeconomic framework and stagnant domestic public debt in the face o f a growing economy. It also reflects slowly declining interest rates o n domestic public debt (the weighted average interest cost o n domestic debt declines by 0.1 percentage point per annum). 6. Thefollowing assumptions apply to the use o f thefiscal space: Non-interest non-defense non-contingent liability spending - which includes the high priority . spending o n growth revival, poverty alleviation and social development - rises from 11.1 percent of GDP in FY 01/02 to 14.5 percent in FY 06/07. . Social and poverty-related spending rises from 3.6 percent o f GDP in FY 01/02 to 5 percent in FY 06/07. Development spending i s targeted to increase from 3.4 percent o f GDP in FY 01/02 to 4.7 percent in FY 06/07. In this regard it i s critical to improve governance (inter alia in procurement), implementation capacity, and expenditure control and monitoring. To the extent that such improvements cannot be achieved in line with the spending plan, actual spending would need to be lower to avoid waste and unnecessarily large public debt 7. Other assumptions: The government i s successful in containing overall expenditure pressures. Total spending . declines from 23.4 percent o f GDP in FY 01/02 to 22.3 percent in FY 06/07. This requires strong political commitment and improved institutional capacity for effective expenditure control. . Foreign grants decline from the high level in FY 01/02 but remain in the range o f 1.0 - 1.2 . percent o f GDP (the same as in FY 99/00 - OO/Ol). N e w foreign loans are mainly on concessional terms. Net privatization receipts to the fiscal budget average 0.3 percent o f GDP per year Annex B Page 3 of 20 Table A.l: Pakistan: Baseline Scenario, 1999/2000 - 2006/07 Est. Prov.Est. Proj. 1999100 2000/012001/02 2002103 2003104 2004/05 2005/062006/07 (annual changes in percent) Output and prices Real GDP at factor costs 3.9 2.2 3.4 5.1 5.3 5.5 5.7 5.9 Real GDP at market prices 4.2 2.6 2.8 5.8 5.3 5.5 5.7 5.9 Consumer prices (p.a.) 3.6 4.4 2.7 3.1 4.0 4.0 4.0 4.0 (in percent of GDP) Savings and investment Gross national savings 14.1 13.6 17.4 21.3 16.7 18.0 18.3 18.5 Govemment -2.5 -2.2 -0.9 1.6 0.1 0.9 1.7 2.2 Nongovernment 16.5 15.8 18.3 19.7 16.7 17.1 16.5 16.3 Gross capital formation 16.0 15.5 14.7 15.5 16.5 17.1 17.7 18.3 Govemment 3 .O 2.6 3.5 3.2 3.5 4.0 4.5 5.0 Nongovemment l/ 11.3 11.3 9.6 9.9 11.0 11.5 11.6 11.7 Change in stocks (nongovt) 1.6 1.6 1.6 2.4 2.0 1.6 1.6 1.6 Savings - Investment balance -2.0 -1.9 2.7 5.9 0.2 0.9 0.6 0.2 Govemment -5.5 -4.8 -4.4 -1.6 -3.4 -3.1 -2.8 -2.8 Nongovemment (inclchange in stocks) 3.5 2.9 7.0 7.5 3.6 4.0 3.4 3.0 (inpercent of GDP) Public finances Revenue (including grants) 17.3 17.3 19.3 20.8 18.5 19.3 19.6 19.8 Revenue 16.3 16.2 17.0 17.9 17.3 17.6 18.0 18.3 Expenditure 22.8 21.4 23.6 22.3 21.8 22.4 22.4 22.6 Non-interest, non-defense, no-SOE spending 9.8 10.2 10.9 12.0 11.6 12.6 13.4 14.3 Overall balance (including grants) -5.5 -4.1 -4.4 -1.6 -3.4 -3.1 -2.8 -2.8 Primary balance (including grants) 2.3 2.8 3.8 3.6 1.4 1.5 1.4 1.0 Gross public debt 21 100.9 108.3 100.1 91.8 83.4 80.3 74.3 71.0 Public debt in percent o f revenue 619.6 670.5 588.9 512.0 481.3 456.9 412.1 388.1 (in percent of GDP) Extemal sector Merchandise trade balance -2.3 -2.2 -0.5 -0.8 -1.4 -1.6 -1.7 -1.8 Merchandise exports 13.4 15.2 15.4 15.9 15.4 15.4 15.4 15.4 Merchandise imports 15.8 17.4 15.9 16.7 16.8 17.0 17.1 17.2 Current account excl. off. transfers -3.5 -3.3 0.2 4.4 -0.9 -0.7 -1.0 -1.3 Current account incl. off. transfers -2.0 -1.9 2.7 5.9 0.2 0.9 0.6 0.2 (in percent of exports of goods and nonfactor services) Extemal public and publicly-guarantd debt 2/ 308.7 289.4 282.6 229.6 211.3 199.2 184.3 172.5 Debt service 46.9 28.2 34.0 26.4 27.6 22.7 17.0 15.1 Gross reserves (in US. dollar ml) 908 1,679 4,330 9,486 9,212 9,569 10,679 11,669 In months o f imports o f goods and services 3.8 1.7 3.7 7.0 6.3 6.0 6.2 6.2 Memorandum items: Real effective exch. rate (% change) -0.6 -2.6 -1.2 6.0 3.9 0.0 0.0 0.0 Real per capita GDP (“hchange) 2.0 0.3 0.7 3.6 3.1 3.3 3.5 3.7 GDP at market prices (Rs bl) 3,147 3,423 3,629 4,018 4,400 4,828 5,307 5,845 Source: Data provided b y the Pakistani authorities; and World Bank staff l/ Includes state-owned enterprises. 2/ Excludes SBP. Annex B Page 4 of 20 Table A.2: Pakistan: Consolidated Government Budget, 2000/01- 2006/07 2000/01 2001102 2002103 2003/04 2004105 2005/06 2006107 Prov. Act. Proj. Proj. Proj. Proj. (in percent of GDP) Revenue and grants 17.3 19.3 20.8 18.5 19.3 19.6 19.8 Revenue 16.2 17.0 17.9 17.3 17.6 18.0 18.3 Tax revenue 12.9 13.0 13.8 13.6 14.0 14.4 14.6 Federal 12.3 12.5 13.3 13.0 13.4 13.7 13.9 CBR revenue 11.5 11.1 11.5 11.6 11.9 12.2 12.4 Petroleum surcharge 0.5 0.81 1.2 1.o 1.1 1.1 1.1 Gas surcharge 0.4 0.49 0.5 0.3 0.4 0.4 0.4 Provincial 0.6 0.5 0.5 0.6 0.6 0.7 0.7 Nontax revenue 3.3 4.0 4.1 3.7 3.6 3.6 3.7 Grants 1.2 2.3 2.8 1.1 1.7 1.6 1.5 Expenditure 21.7 24.0 22.3 21.8 22.4 22.4 22.6 Current expenditure 19.6 19.0 19.7 17.7 17.9 17.7 17.5 Federal 14.8 14.2 14.9 13.0 12.9 12.3 11.8 Interest payments 6.8 6.8 5.2 4.8 4.6 4.2 3.9 Domestic 5.4 5.1 4.0 3.9 3.7 3.5 3.3 Foreign 1.5 1.7 1.2 0.9 0.9 0.7 0.6 Defense 3.8 4.1 4.0 3.6 4.0 4.0 4.0 Running of the civil govemment 2.1 1.6 1.5 1.6 1.7 1.7 1.8 Pensions for defense and civil govemment 0.9 0.7 0.9 0.9 0.9 0.9 0.9 Subsidies 0.6 0.5 1.2 1.4 0.9 0.6 0.3 Power Sector 0.8 0.6 0.4 0.2 Other SOEs 0.5 0.2 0.1 0.0 Non SOEs 0.1 0.1 0.1 0.1 Grants 0.5 0.6 0.6 0.7 0.7 0.7 0.7 Provincial 4.9 4.8 4.8 4.7 5.0 5.3 5.7 Development expenditure and net lending 2.1 3.5 2.7 3.7 4.1 4.5 5.0 Public Sector Development Program 2.6 3.5 3.2 3.5 4.0 4.5 5.0 Net lending -0.5 0.0 -0.6 0.2 -0.1 0.0 0.0 Payment o f contingent liabilities 1.4 0.5 0.3 0.2 0.1 Statistical discrepancy -0.3 -0.4 0.1 0.0 0.0 0.0 0.0 Budget balance (excluding grants) -5.2 -6.6 -4.5 -4.5 -4.8 -4.4 -4.3 Budget balance (including grants) -4.1 -4.4 -1.6 -3.4 -3.1 -2.8 -2.8 Primary balance (including grants) 2.5 2.0 3.6 1.4 1.5 1.4 1.o Financing 4.1 4.4 1.7 3.4 3.1 2.8 2.8 Extemal 2.3 1.4 -0.2 1.o 0.1 0.6 0.4 Domestic 1.7 2.7 1.6 2.1 2.7 1.9 2.2 Privatizationproceeds 0.0 0.2 0.3 0.2 0.3 0.3 0.3 Memorandum items: Primary balance (excluding grants) 1.6 0.1 0.7 0.2 -0.2 -0.2 -0.5 Revenue balance (excluding grants) -3.5 -2.0 -1.8 -0.4 -0.4 0.3 0.8 Social and poverty-related expenditure I/ 0.0 4.8 5.1 5.4 6.0 6.5 7.1 Spending excl interest, defenses, subsidies, cont. liabilities, % GDP 10.2 10.9 12.0 11.6 12.6 13.4 14.3 Share o f total spending (percent) 47.4 46.0 53.5 52.9 56.2 59.9 63.3 Nominal GDP (in millions o f Pakistani rupees) 3,423 3,629 4,018.1 4,400.3 4,828.0 5,307.4 5,845.3 ~~ Source: National authorities; and staff projections and calculations. 11Education, health and population planning, Law and Order, Rural electrification, Water supply and sanitation, Rural Development, Irrigation, Roads, highways and bridges, and “Other poverty-relatedexpenditure”. Annex B Page 5 of 20 8. Key results: . Reflecting the totality o f the above assumptions the primary balance (excluding grants) rises from broadly balance in FY 2001/02 to 0.9 percent o f GDP on average in FY03 - FY07. The revenue balance (current revenue - current spending) marks an even stronger improvement by a full 2.7 percent o f GDP over the projection period, from a deficit o f 2 percent o f GDP in FY 2001/02 to a surplus o f 0.7 percent o f GDP in FY 06/07. Importantly, this leaves room for financing parts o f public investment from savings on current operations which marks an important turnaround from many years o f government dissaving. The overall fiscal deficit including grants and payment o f contingent liabilities drops from 4.2 . percent o f GDP in FY 2001/02 to 3.3 percent o f GDP in FY 06/07, consolidating the fiscal adjustment achieved so far under the PRGF program with the IMF. This scenario shows a steady improvement in public debt indicators. Annual public debt service in percent o f fiscal revenue falls fi-om 55 to 32 % by the end o f the projection period. The public debt/GDP ratio drops from 99 to 72 % and the Present Value o f the public debt/GDP ratio drops from 80 to about 60 %. The structure o f public expenditure changes considerably towards high priority areas. The share o n non-interest non-defense non-contingent liability spending in total spending rises from 48 percent in FY02 to 65 percent in FY07. C. Balance o f payments (Table A.3): 9. Key assumptions: . Stable real exchange rate Average annual export volume growth: 6.1 percent. This reflects the exchange rate assumption . and continued structural reforms to remove impediments to improvements in productivity and to diversifying into higher-value added exports Average annual import volume growth: 9.5 percent, reflecting the pick-up in economic growth, . the expanding government development spending and increasing private sector investment- related imports. Private transfers: it i s assumed that the surge in ‘workers’ remittances in FY02/03 i s temporary and reflects a portfolio reallocation. Hence workers’ remittances in FY03/04 are projected t o fall . back to the FY01/02 level (about $2.4 billion) and then grow slowly in line with international inflation. Official grants are projected to decline from 2.2 percent o f GDP in FY 02/03 to 1 percent o f GDP . in FY 06/07, throughout the period lying considerably below the average for the last three years (1.8 percent o f GDP). Continued program financing (on concessional terms) from the World Bank and the AsDB throughout the period whereas the IMF’s PRGF financing stops in FY 04/05. Total net foreign financing (disbursement o f grants and loans plus rescheduling minus amortization payments) i s projected to decline gradually from 4.6 percent o f GDP in FY 01/02 to 2.7 percent o f GDP in FY 06/07. Actual disbursements o f new loans are projected to remain in the 2 % - 3 1/3 percent o f GDP range seen in recent years reflecting willingness from donors to provide financial assistance for the government’s reform program. The projections imply that disbursements grow in foreign currency terms but remain fairly constant in terms o f percent o f donors’ GDP. Annex B Page 6 of 20 Table A.3: Pakistan: Baseline Scenario: Balance of Payments, 2000/01-2006/07 Proj. Proj. Proj. Proj. Proj. 2000101 2001102 2002103 2003104 2004105 2005/06 2006107 (in millions o f US. dollars) Current account (excluding official transfers) -1,951 96 3,029 -651 -602 -847 -1,198 Current account balance (including official transfers) -1,112 1,591 4,029 190 793 570 239 Trade balance -1,268 -292 -536 -1,040 -1,306 -1,507 -1,762 Exports f.0.b. 8,934 9,140 10,889 11,759 12,663 13,648 14,738 Imports f.0.b. -10,202 -9,432 - 11,425 -12,800 -13,969 -15,155 -1 6,500 Services (net) -3,143 -2,617 -2,172 -3,230 -2,977 -3,082 -3,239 of which: interest payments -1,657 -1,579 -1,272 -1,117 -1,053 -902 -799 Private transfers (net) 2,460 3,005 5,737 3,619 3,682 3,742 3,803 Official transfers (net) 839 1,495 1,000 841 1,395 1,417 1,437 Capital account -623 -2,322 -1,529 -2,481 -1,426 -676 -418 Public medium- and long-term capital -652 -1,613 -1,497 -1,590 -1,254 -832 -733 Project and nonproject loans -332 -983 -1,696 -1,012 -879 -501 -592 Disbursements 1,463 53 1 723 788 855 967 1,052 Amortization -1,795 -1,514 -2,419 -1,800 -1,734 -1,468 -1,644 Commercial banks and IDB (net) -76 -224 -158 -33 -16 -16 -16 Other -244 -407 357 -545 -359 -315 -125 Public sector short-term (net) -59 -1,064 -233 -564 -1 17 -100 -100 Private medium- and long-term 343 -80 221 218 475 670 744 Private short-term (including errors & omissions) -256 435 -19 -551 -530 -415 -329 Overall balance (before debt relief) -1,735 -73 1 2,501 -2,297 -633 -106 -179 Financing 1,735 734 -2,501 -2,297 633 106 179 Reserve assets (increase -) -1,088 -3,082 -5,267 -1 1 -892 -1,541 -1,442 State Bank o f Pakistan (including FE-25s) -729 -2,717 -5,916 99 -540 -1,309 -1,194 Deposit money banks -359 -365 649 -1 10 -352 -232 -248 Fund repurchases -194 -194 -418 -526 -362 -135 -111 Net exceptional financing 3,017 4,007 3,185 2,842 1,941 1,850 1,800 Arrears (increase +) -525 0 0 0 0 0 0 Rescheduling 1,587 1,210 1,009 100 100 100 100 of which: Private Sector Involvement 0 0 100 100 100 100 100 Rollover o f foreign deposits with banking system 1,676 1,314 900 1,200 500 500 500 Program financing from 1FIs 279 1,367 1,090 1,342 1,121 1,000 950 World Bank 0 698 213 500 500 600 600 AsDB 0 185 408 400 400 400 350 JMF 279 484 469 442 221 0 0 Privatization receipts 0 117 186 200 220 250 250 Memorandum items: (in percent of GDP) Current account balance (including official transfers) -1.9 2.7 5.9 0.2 1.o 0.6 0.2 Current account (excluding official transfers) -3.3 0.2 4.4 -0.9 -0.7 -1 .o -1.2 Current account (excluding all transfers) -7.5 -4.9 -4.0 -5.6 -5.2 -5.2 -5.2 Interest payments1GDP -2.8 -2.7 -1.9 -1 .5 -1.3 -1.0 -0.8 External primary balance (incl. off. transfers)/GDP 0.9 5.4 1.7 1.7 2.2 1.7 1.1 Official grants1GDP 1.4 2.5 1.5 1.1 1.7 1.6 1.5 Total foreign financing (net) incl. resched./GDP 4.0 4.5 2.0 1.7 2.1 2.3 2.0 Trade balance -2.2 -0.5 -0.8 -1.4 -1.6 -1.7 -1.8 Exports f.0.b. 15.2 15.4 15.9 15.4 15.4 15.4 15.4 Imports f.0.b. -17.4 -15.9 -16.7 -16.8 -17.0 -17.1 -17.2 End-period gross official reserves 1,679 4,330 -9,486 9,312 9,773 10,989 12,089 (in weeks o f imports o f goods and nonfactor services) 7.5 15.9 30.4 27.5 26.7 27.6 33.2 Sources: State Bank o f Pakistan; Ministry o f Finance; and World Bank staff estimates. Annex B Page 7 of 20 . 10. Results: . The trade deficit widens gradually but remains within moderate levels (in all years at less than the FYOO/Ol level o f 2.2 percent o f GDP). The current account balance (including official transfers) returns t o deficits - albeit modest - in FY03/04 after two years o f surpluses. T h e surpluses have been driven by the massive surge in private transfers that in our assessment reflect a portfolio reallocation and hence are not o f a long- term nature. The current account balance excluding all transfers (official and private) shows a much less marked improvement in the last two years and over the medium term, only a slight increase in the deficit i s projected. The current account deficits (including official transfers) are in . line with the needs o f a fast-growing low-income economy and fully sustainable as shown by the rapidly declining external debt burden. . External public and publically guaranteed debt declines from 54 percent o f GDP in FYO1/02 to 36 percent o f GDP in FY06/07. Gross official reserves continue to accumulate in nominal terms, reaching U S $ 9.9 billion by FY06/07. They level out at about 5.8 months o f import cover (of goods and services) over the medium term. 111. Low CASE SCENARIO 11. The main alternative scenario assumes less stringent reform actions in particular with regard to the fiscal budget, public enterprises and (other) structural reforms. D. Overall macroeconomic framework (Table A.4): Real GDP growth (at market prices) rises from 4.4 percent in FY 01/02 to 4.9 percent o f GDP in . FY 06/07. Inflation rises gradually to 4.3 percent over the medium term, up f r o m 2.7 percent in FY 01/02 . Gross national savings remain at the pre-FYOU02 level o f about 14 percent o f GDP compared to the base case reaching 17 percent o f GDP by FY 06/07. Investment (gross fixed capital formation) remains at the pre-FYOU02 level o f about 14 % . percent o f GDP; substantially lower than the base case reaching16 % percent o f GDP by FY 06/07. The ICORs (Incremental Capital Output Ratios) are somewhat higher (averaging 3.3) than in the . base case (average 3 .O) since productivity gains to investments are lower due to lagging structural reforms. Constant real effective exchange rate. E. Fiscal projections (Table A.5): . 12. M a i n assumptions: Fiscal revenue collection i s considerably lower than in the base case: the government proves less effective in tackling challenges in C B R and vested interests. Revenue collection in percent o f GDP stagnates at the FY 01/02 level averaging 16.6 percent o f GDP over the medium term. By FY 06/07 revenue collection i s 1.1 percent o f GDP lower than in the base case. In particular, the improvement in C B R revenue collection seen over the last year ceases and C B R revenue stalls slightly below the FY03 level (in percent o f GDP). There i s less foreign financing - o f both grants and (concessional) loans, reflecting less confidence by donors in the government’s commitment t o pursue reforms and address the goals Annex B Page 8 of 20 . for poverty reduction, social development and growth revival through appropriate economic policies. . Improvements in effectiveness o f government spending, governance and implementation capacity are much slower than in the base case. Total spending levels are only slightly higher than in the base case but the composition i s . radically different with much higher interest payments, defense spending and payments o f contingent liabilities than in the base case and consequently, much less fiscal space. Interest payments are much higher: at 4.6 percent o f GDP by FY06/07 they are 0.9 percent o f GDP higher than in the base case. This reflects a slower reduction in interest rates o n domestic public debt because o f less strengthening in domestic confidence, and also the larger fiscal deficits and hence higher borrowing requirements that lead t o substantially larger public debt than in the base case. Defense costs grow in line with nominal GDP instead o f in line with inflation only in the base . case. They remain at 4 percent o f GDP, and by FY06/07 defense spending i s 1.1 percent o f GDP higher than in the base case. The government i s less successful in reducing public enterprise losses and other contingent liabilities. Payments o f contingent liabilities average 0.9 percent o f GDP in FY 02/03 - FY 06/07 . compared t o 0.4 percent o f GDP in the base case. The progress on privatizing public enterprises i s slow. Annual average sales receipts are h a l f o f those in the base case. F. Key results: 13. There i s substantially less r o o m for non-interest non-defense non-contingent liability spending (stable at 11.9 percent o f GDP throughout the medium term) than in the base case (averaging 13.2 percent o f GDP and increasing to 14.5 percent by FY07). The structure o f spending remains almost unchanged from FY03. The share o f non-interest non-defense non-contingent liability spending in total spending declines slightly from 54 to 53 percent. Social and poverty-related spending stagnates at 3.8 percent o f GDP, below the FY03 level. Development spending i s cut over the medium term since most o f the fiscal adjustment needed t o . avoid excessive fiscal deficits takes place in this category. By FY06/07 development spending has fallen to 2.8 percent o f GDP; i.e., 1.9 percentage point lower than in the base case. The lower spending on health and education as well as o n infrastructure - both O&M and new investments (growth-related spending) - result in lower productivity o f public capital stock and human capital and slower economic growth than in the high case. Social indicators, poverty reduction and productivity o f human capital and public infrastructure improve less than in the base case since the effectiveness o f government spending improves less and the government finds less room for spending in those areas. * Total public debt i s higher and with less concessional terms o n average (since there i s less external debt and more nonconcessional domestic debt); hence, the prospects o f reaching fiscal sustainability are considerably worse with much slower improvement in public debt indicators. Annual debt service in percent o f fiscal revenue rises f r o m 38 percent in FY03 to 41 percent by FY07 as compared to 32 percent in the base case. The debthevenue ratio drops from 602 percent in FY02 to 485 percent by FY07 as compared to 396 percent in the base case. Progress i s hence slow in the l o w case on reaching the sustainability threshold o f 300 percent. Annex B Page 9 of 20 Table A.4: Pakistan: Low Case Macroeconomic Framework, 2000/01-2006/07 Est. Prov.Est. Proj. 2000101 2001102 2002103 2003104 2004105 2005106 2006107 Output and prices (annual changes in percent) Real GDP at factor costs 2.2 3.4 5.1 5.0 4.9 4.9 4.9 Real GDP at market prices 2.6 2.8 5.8 5.0 4.9 4.9 4.9 Consumer prices (p.a.) 4.4 2.7 3.1 4.1 4.1 4.2 4.2 Savings and investment Gross national savings 13.6 17.4 21.3 15.1 15.3 14.7 14.3 Government -2.2 -0.9 1.6 -0.8 -1.0 -1.4 -1.8 Nongovemment 15.8 18.3 19.7 15.9 16.3 16.2 16.1 Gross capital formation 15.5 14.7 15.5 15.7 15.7 15.9 16.0 Government 2.6 3.5 3.2 3.3 3.1 2.9 2.8 Nongovemment 1/ 11.3 9.6 9.9 10.4 11.0 11.4 11.6 Change in stocks (nongovt) 1.6 1.6 2.4 2.0 1.6 1.6 1.6 Savings - Investment balance -1.9 2.7 5.9 -0.6 -0.4 -1.2 -1.7 Government -4.8 -4.4 -1.6 -4.1 -4.1 -4.4 -4.6 Nongovemment (incl.change in stocks) 2.9 7.0 7.5 3.5 3.7 3.2 2.9 Public finances (in percent of GDP) Revenue (including grants) 17.3 19.3 20.8 18.1 18.3 18.0 17.7 Revenue 16.2 17.0 17.9 17.1 16.9 16.7 16.5 Expenditure 21.4 23.6 22.3 22.2 22.4 22.4 22.4 Non-interest,non-defense, non-SOE spending 10.2 10.9 12.0 11.4 11.5 11.4 11.3 Overall balance (including grants) -4.1 -4.4 -1.6 -4.1 -4.1 -4.4 -4.6 Primary balance (including grants) 2.8 3.8 3.6 0.7 0.7 0.3 -0.1 Gross public debt 21 108.3 100.1 91.8 83.6 81.5 78.9 77.9 Public debt in percent of revenue 670.5 588.9 512.0 488.8 482.5 473.0 473.2 Extemal sector (in percent of GDP) Merchandise trade balance -2.2 -0.5 -0.8 -1.4 -2.0 -2.5 -2.8 Merchandise exports 15.2 15.4 15.9 15.4 15.2 14.9 14.7 Merchandise imports 17.4 15.9 16.7 16.9 17.2 17.4 17.5 Current account excl. official transfers -3.3 0.2 4.4 -1.6 -1.8 -2.5 -2.9 Current account incl. official transfers -1.9 2.7 5.9 -0.6 -0.4 -1.2 -1.7 (in percent of exports of goods and nonfactor services) Extemal public and publicly-guaranteeddebt 21 289.4 282.6 229.6 208.1 196.4 186.7 179.0 Debt service 28.2 34.0 26.4 27.7 20.0 14.9 10.8 Gross reserves (in millions o f US. dollars) 1,679 4,330 9,486 7,920 7,024 6,000 4,979 in months o f imports of goods and services 1.7 3.7 7.0 5.4 4.4 3.5 3.5 Memorandumitems: Real effective exchange rate (% change) -2.6 -1.2 6.0 3.7 0.0 0.0 0.0 Real per capita GDP (% change) 0.3 0.7 3.6 2.8 2.7 2.7 2.7 GDP at market prices (billions o f PK rupees) 3,423 3,629 4,018 4,392 4,796 5,242 5,730 Source: Data provided by the Pakistani authorities and World Bank staff. 1/ Includes state-owned enterprises. 2/ Excludes SBP. Annex B Page I O of 20 Table A.5. Pakistan: Low Case: Consolidated Government Budget, 2000/01-2006/07 2000/01 2001102 2002103 2003104 2004105 2005106 2006107 Prov. Act. Proj. Proj. Proj . Proj. , . Dercent o f GDPi (in Revenue and grants 17.3 19.3 20.8 f8.l 18.3 18.0 17.7 Revenue 16.2 17.0 17.9 17.1 16.9 16.7 16.5 Tax revenue 12.9 13.0 13.8 13.5 13.4 13.3 13.2 Federal 12.3 12.5 13.3 12.9 12.8 12.7 12.6 C B R revenue 11.5 11.1 11.5 11.5 11.4 11.3 11.2 Petroleum surcharge 0.5 0.81 1.2 1.o 1.o 1.o 1.o Gas surcharge 0.4 0.49 0.5 0.3 0.3 0.3 0.3 Provincial 0.6 0.5 0.5 0.6 0.6 0.6 0.6 Nontax revenue 3.3 4.0 4.1 3.6 3.5 3.4 3.3 Grants 1.2 2.3 2.8 1.o 1.4 1.4 1.3 Expenditure 21.7 24.0 22.3 22.2 22.4 22.4 22.4 Current expenditure 19.6 19.0 19.7 18.3 18.7 18.8 18.9 Federal 14.8 14.2 14.9 13.6 14.0 14.2 14.3 Interest payments 6.8 6.8 5.2 4.8 4.8 4.6 4.5 Domestic 5.4 5.1 4.0 3.9 3.9 4.0 4.0 Foreign 1.5 1.7 1.2 0.9 0.8 0.7 0.5 Defense 3.8 4.1 4.0 4.1 4.4 4.7 5.0 Running o f the c i v i l govemment 2.1 1.6 1.5 1.6 1.7 1.8 1.9 Pensions for defense and civil govemment 0.9 0.7 0.9 0.9 0.9 1.o 1.o Subsidies 0.6 0.5 1.2 1.5 1.3 1.2 1.1 Power sector 0.9 0.8 0.7 0.6 Other SOEs 0.5 0.4 0.4 0.3 N o n SOEs 0.2 0.1 0.1 0.1 Grants 0.5 0.6 0.6 0.7 0.7 0.7 0.7 Provincial 4.9 4.8 4.8 4.7 4.7 4.7 4.6 Development expenditure and net lending 2.1 3.5 2.7 3.5 3.3 3.1 3.0 Public Sector Development Program 2.6 3.5 3.2 3.3 3.1 2.9 2.8 N e t lending -0.5 0.0 -0.6 0.2 0.2 0.2 0.2 Payment o f contingent liabilities 1.4 0.5 0.5 0.5 0.5 Statistical discrepancy -0.3 -0.4 0.1 0.0 0.0 0.0 0.0 Budget balance (excluding grants) -5.2 -6.6 -4.5 -5.1 -5.6 -5.7 -5.9 Budget balance (including grants) -4.1 -4.4 -1.6 -4.1 -4.1 -4.4 -4.6 Primary balance (including grants) 2.5 2.0 3.6 0.7 0.7 0.3 -0.1 Financing 4.1 4.4 1.7 3.4 4.1 4.4 4.6 Extemal 2.3 1.4 -0.2 1.o -0.1 -0.3 0.0 Domestic 1.7 2.1 1.6 2.1 4.1 4.5 4.5 Privatization proceeds 0.0 0.2 0.3 0.2 0.1 0.1 0.1 Memorandum items: Primary balance (excluding grants) 1.6 0.1 0.7 -0.3 -0.8 -1.1 -1.4 Revenue balance (excluding grants) -3.5 -2.0 -1.8 -1.2 -1.8 -2.1 -2.5 Spending excl. interest, defenses, subsidies, cont. 10.2 10.9 12.0 11.4 11.5 11.4 11.3 liabilities, % GDP Share o f total spending (percent) 47.4 46.0 53.5 51.2 51.1 50.9 50.5 Nominal GDP (in millions o f Pakistani rupees) 3,423 3,629 4,018 4,392 4,796 5,242 5,730 Source: National authorities; and staff projections and calculations Annex B Page I I of 20 G. Balance o f payments (Table A.6): 14. Key assumptions: Stable real exchange rate m Average annual export volume growth: 4 percent. Slower structural reforms result in slower productivity growth. Hence the competitiveness in export and import competing industries i s worse. Average annual import volume growth: 8.3 percent. m Private transfers: as in the base case it i s projected that workers’ remittances in FY03/04 fall back to the FYO1/02 level (about $2.4 billion) - although slightly lower than in the base case reflecting lower confidence - and then growing slowly in line with international inflation. Official grants are projected to decline from 2.4 percent o f GDP in FY 02/03 to 0.8 percent o f GDP in FY 06/07, throughout the period lying 0.1 - 0.3 percent o f GDP lower than in the base case, reflecting that donors are less confident about the govemment’s commitment to continued reforms. Lower program financing (adjustment loans on concessional terms) from the W o r l d Bank and the AsDB throughout the period (on average 0.5 percent o f GDP per year compared to 1.4 percent in the base case) and consequently recourse to larger financing at commercial terms. 15. Results. The trade deficit widens back to the level in F YO O - FYO1, averaging 2.3 percent o f GDP. The current account balance (including official transfers) returns to deficits, reaching 2.3 percent o f GDP by FY07 compared to 1.3 percent in the base case. Given the limited financing at concessional terms the increasing current account deficits are partially financed by a drawdown o f international reserves. They drop from about US$ 8 billion in March 2003 (about 25 weeks o f import cover) to about U S $ 4.4 billion (about 13 weeks o f import cover) by FY06/07. Extemal debt relative to exports o f goods and services declines from 309 percent in F Y 0 2 t o 240 percent in FY07 but remains s t i l l far above the sustainability threshold o f 150 percent. 16. Figures 2.1 and 2.2 present the results o f a sensitivity analysis for the public debt burden and the external debt burden, respectively, under the baseline scenario. The robustness o f the projected public debt burden (as measured in percent o f fiscal revenue) and the external debt burden (measured against exports of goods and nonfactor services) i s tested with regard to a wide variety o f exogenous shocks and assumptions about the macroeconomic stance over the medium term. 17. The main conclusion from the stress tests o n the public debt burden (Figure 2.1 and Table A.7) i s that almost none o f the shocks identified here lead to a worsening debt burden over the medium term. Most scenarios show sustained reduction in the debt-to-revenue and the debt-to-GDP ratios and remain on a path toward debt sustainability within the 10-year horizon identified in the Fiscal Sustainability and Debt Limitation Ordinance. However, one scenario-reversal o f macroeconomic policies and environment to the average o f 1992/93 -1998/99--stands out as unsustainable. In the other cases o f severe shocks the debt-to-revenue and debt-to-GDP ratios generally s t i l l declines although very slowly. T w o sets o f stress tests have been run: one where we test the extent o f fiscal policy reversal that could make the public debt situation unsustainable and one where we test for shocks to the growth rate o f real GDP, the interest rate, and the GDP deflator within a magnitude o f 2 standard deviations o f the average over the last 10 years. Annex B Page 12 of 20 Table A.6: Low Case: Pakistan: Balance of Payments, 2001/01-2006/07 Proj . Proj. Proj. Proj. Proj. 2000101 2001102 2002/03 2003104 2004105 2005106 2006107 (in millions of US Dollars) Current account (excluding official transfers) -167 96 3,029 -1,210 -1,487 -2,159 -2,747 Current account balance (including official transfers) 9 1,591 4,029 -450 -351 -1,020 -1,604 Trade balance -92 -292 -536 -1,099 -1,609 -2,165 -2,640 Exports f.0.b. 2,351 9,140 10,889 11,704 12,364 12,999 13,676 Imports f.0.b. -2,443 -9,432 -1 1,425 -12,803 -13,973 -15,164 -16,3 16 Services (net) -675 -2,617 -2,172 -3,230 -3,028 -3,172 -3,337 Ofwhich: interest payments -326 -1,579 -1,272 -1,115 -1,050 -891 -782 Private transfers (net) 600 3,005 5,737 3,119 3,151 3,178 3,230 Official transfers (net) 176 1,495 1,000 760 1,136 1,139 1,142 Capital account 440 -2,322 -1,529 -2,547 -871 -281 299 Public medium- and long-term capital 272 -1,613 -1,497 -1,650 -624 -311 159 Project and nonproject loans 249 -983 -1,696 -1,072 -549 -280 0 Disbursements 669 531 723 728 735 787 844 Amortization ” -420 -1,514 -2,419 -1,800 1,284 -1,068 -844 Commercial banks and IDB (net) 16 -224 -158 -33 284 284 284 Other 7 -407 357 -545 -359 -315 -125 Public sector short-term (net) -11 -1,064 -233 -564 -117 -100 -100 Private medium- and long-term 152 -80 221 218 400 545 569 Private short-term (including errors & omissions) 27 435 -19 -551 -530 -415 -329 Overall balance (before debt relief) 449 -731 2,501 -2,997 -1,223 -1,300 -1,306 Financing -449 731 -2,501 2,997 1,223 1,300 1,306 Reserve assets (increase -) -883 -3,082 -5,267 1,392 475 711 691 Fund repurchases -40 -194 -418 -526 -362 -135 -111 N e t exceptional financing 474 4,007 3,185 2,131 1,110 725 725 Arrears (increase +) -8 0 0 0 0 0 0 Rescheduling 31 211 1,210 1,009 100 100 100 100 Rollover o f foreign deposits with banking system 271 1,314 900 1,200 500 500 500 Program financing f r o m IFIs 0 1,367 1,090 731 400 0 0 Privatization receipts 0 117 186 100 110 125 125 Memorandum items: (in percent of GDP) Current account (excluding official transfers) 0.0 2.7 5.9 -0.6 -0.4 -1.2 -1.7 Current account (including official transfers) -0.3 0.2 4.4 -1.6 -1.8 -2.5 -2.9 Interest payments/GDP -2.7 -1.9 -1.5 -1.3 -1 .o -0.8 Extemal primary balance (incl. off. transfers)/GDP 5.4 7.7 0.9 0.9 -0.1 -0.9 Total foreign financing (net)/GDP Official grants/GDP 2.5 1.5 1.o 1.4 1.3 1.2 Total public- and publicly-guaranteed extemal debt 40.9 58.0 48.9 40.9 37.6 34.9 32.7 01w Public- and publicly-guaranteed extemal debt excl. SBP 38.1 52.7 45.8 38.1 35.0 32.5 30.4 Total foreign financing (net) incl.resched./GDP 4.6 2.0 0.7 1.3 1.1 1.3 Trade balance -0.2 -0.5 -0.8 -1.4 -2.0 -2.5 -2.8 Exports f.0.b. 4.0 15.4 15.9 15.4 15.2 14.9 14.7 Imports f.0.b. -4.2 -15.9 -16.7 -16.9 -17.2 -17.4 -17.5 End-period gross official reserves 1,679 4,330 9,486 7,920 7,024 6,000 4,979 (in weeks o f imports o f goods and nonfactor services) 7.5 15.9 30.3 23.4 19.2 15.2 13.7 Sources: State Bank o f Pakistan: Ministry o f Finance; and World Bank staff estimates. 1/ Assumes early debt retirement o f $1 b i l l i o n in 29003/04, and $500 m i l l i o n each in 2004105 and 2005106. Annex B Page 13 of 20 Figure 2.1: Sensitivity Analysis for Public DebffRevenues and Grants I - ---A B - -_-- ~ .c D G E H _ ---F I -Baseline Projection A = If interest rate, real GDP growth rate, primary balance and non-debt flows (in percent o f GDP) are at average o f past 10 years B = Ifinterest rate in 2003104 and 2004105 i s average plus two standard deviations, others at baseline C = Ifreal GDP growth rate in 2003104 and 2004105 i s average minus two standard deviations, others at baseline D = If primary balance (in percent o f GDP) in 2003104 and 2004105 is average minus two standard deviations, others at baseline E = Combination o f 2-4 using one standard deviation shocks F = Annual additional contingent liability payments o f 5 percent o f GDP in 2003104 and 2004/05, others at baseline G = Same as A, but averages and standard deviations calculated up to 1998199 only. H = Same as D, but shock extending through 2006107 I= Same as E, but shocks extending through 2004105 Baseline Projection = Public debthevenues and grants Annex B Page 14 of 20 Figure 2.2: Sensitivity Analysisfor External Public Debt in percent of Exports of Goods and Nonfactor Services 400 350 300 $ 250 200 I I I I I I I 150 I I I I I I I 100 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 200304 2004/05 2005/06 2006/0 -Baseline Projection - . - .- A - 0 C D .._ ~ -E -_ ____F-G Baseline Projection = Ratio Extemal public debt to Exports o f goods and NF services A = If interest rate, real GDP growth rate, U S $ GDP deflator growth, non-interest current account, and non-debt flows (in percent of GDP) over the medium t e r m are at average o f past 10 years B = If interest rate on extemal debt in 2003104 and 2004105 i s IO-year average plus two standard deviations, others at baseline C = If real GDP growth rate in 2003/04 and 2004/05 i s average minus two standard deviations, others at baseline D = If US$GDP deflator in 2003/04 i s average minus two standard deviations, others at baseline E = I f non-interest current account (in percent o f GDP) in 2003104 and 2004105 i s average minus two standard deviations, others at baseline F = Combination o f 2-5 using one standard deviation shocks G = 20 percent depreciation in 2003104 and in 2004/05, others at baseline Annex B Page I 5 of 20 18. In test 1 w e look at the impact o f a fiscal policy reversal back to the average primary balance (in percent o f GDP) over the last 10 years as well as a reversal to a similar macroeconomic environment with the interest rate, real GDP growth rate, and non-debt flows (in percent o f GDP) all assumed equal to the 10-year average. (This period includes the last four years with reforms and progress in fiscal consolidation and macroeconomic stabilization.) A policy reversal within this magnitude would not end up in explosive debt dynamics but would rather be characterized by severe stagnation: in the debt burden, economic growth and poverty reduction. The debt burden would decline only very slowly, economic growth would be slow and progress on poverty reduction, social and growth-related spending would be minimal. 19. In the second stress test we check the sensitivity o f returning to the policies and macroeconomic ffamework in the seven years preceding the initiation o f reforms, i.e., we use the averages up to 1998/99. The conclusion i s clear: such a policy regime i s totally unsustainable and with substantially worse debt indicators than in test 1. 20. A third set o f shocks (3. - 7.) tests the debt burden sensitivity o f temporary shocks to the following macroeconomic variables: interest rates (test 3), real GDP growth rate (test 4), primary balance (test 5) and a combination o f tests 2. - 4 covering a set o f macroeconomic variables (test 6)70but with a smaller shock (using averages plus 1 standard deviation). Test 7 analyzes the impact o f subsidies t o public enterprises and fiscal payments o f contingent liabilities o f 5 percent o f GDP in addition to the baseline amounts annually in 2003/04 and 2004/05. The general conclusion i s that temporary shocks o f these magnitudes only slow down the process o f debt burden reduction but a sustainable debt position would s t i l l be reached by 201 1/12. Hence, the baseline scenario i s robust to temporary macroeconomic shocks. 21. A fourth set o f shocks (8. - 9.) tests the debt burden sensitivity o f shocks persisting throughout the projection horizon o f the same magnitude as above. The general conclusion i s that these permanent shocks lead to stagnation or very slow decline in the debt burden even over the whole 10-year horizon and the debt position remains unsustainable. 22. The stress tests o n the external debt burden (Figure 2.2 and Table A.8) show that a return to historic macroeconomic perfonnance i s unsustainable. This emerges from three tests: Test 1 assumes macroeconomic performance throughout the medium term to be the same as the average over the last 10 years. Test 6 and 7 assume a temporary reversal to historic macroeconomic performance in the f i r s t two years o f the projection period but with an additional worsening o f key macroeconomic variables by 1-2 standard deviations. Such reversals will trigger a worsening o f the debt burden measured both in terms o f exports o f goods and nonfactor services (all three tests) and in terms o f GDP (tests 6 and 7). In particular the first test shows clearly an unsustainable path. These three tests once again underline the importance for the Government to maintain the reform momentum, pursue sound macroeconomic policies and avoid reverting to the macroeconomic policies o f the 19990s. ' O The shocks are modeled as follows: in the first two years - 2003104 and 2004105 -the projected variables are set equal to their historic averages plus a worsening by two standard deviations. Subsequent years use the baseline scenario data. 9 9 9 9 9 9 9 9 9 2 2 2 2 2 2 2 2 2 I A I A L A I A I A ~ I A I A I A 09090909090909090909 h 2 x b P 8 Y 8 O h -$b a %:E x - Annex B Page 18 of 20 23. However, the external debt sustainability o f the baseline scenario i s fairly robust t o most o f the other stress tests. None o f the singular shocks identified here (separate tests for temporary shocks t o interest rates (Test 2), real GDP growth rate (Test 3), U S $ GDP deflator (Test 4), and the non-interest current account balance (Test 5)) leads t o a worsening debt burden over the medium term. However, if a l l these temporary shocks were t o come at the same time (Test 6), the combined massive shock to the economy w o u l d induce such a sharp, initial rise in the debt burden that i t w o u l d s t i l l be substantially heavier after five years. After the initial peak the debt burden w o u l d steadily b e reduced, reflecting the underlying strength o f the baseline scenario and would be o n the path towards sustainability. However, debt sustainability would only b e reached beyond the 10-year horizon given the substantial size and comprehensiveness o f the initial shock. An almost identical conclusion applies t o the impact o n debt sustainability o f 20 percent depreciation in each o f the first two years (Test 7). 2 N p! t- N .- * 0 2 I ei . r- 0 u3 0 0 N e u3 m 2 g .= 0 N V o m 'E 4B 8 0 N d e W 0 0 N W w F s .- c e Y Annex C Page 1 of 5 ON DYNAMIC SIMULATIONS BASED BENEFITINCIDENCE OF PUBLIC EDUCATION EXPENDITURES IN PAKISTAN I. Static Benefit Incidence Analysis and i t s Limitation 1. In Palustan, the social gaps (urban-rural, male-female, and rich-poor) in educational attainment are prominent. According to the Static Benefit Incidence analysis, in order to narrow these gaps, public expenditure for primary education i s particularly important, since such expenditure i s pro-poor, and especially benefits the rural poor children (see B o x 3.1). The Palustan governments have appropriately allocated more public funds to primary education than other educational sectors, although the total expenditure for primary education has recently stagnated. 2. Although these results are useful, they may be debatable in the following sense. The static benefit incidence measures discussed above are static in that they do not take into account the dynamic effects o f spending an additional dollar on the use o f public services. In particular, it i s not clear whether spending an additional dollar benefits the rural poor even though the static benefit incidence analysis suggests it does. 11. The impact of expanding public expenditure for education: Dynamic benefit incidence analysis 3. This section evaluates the incidence o f expanding public expenditure for education. The primary objective i s to measure the impact o f expanding public expenditure for primary education on the gross enrollment rate o f the poor, while controlling for expansion in private schooling. To do so, it i s appropriate to use the dynamic benefit incidence analysis proposed by Lanjouw and Ravallion (1999) (see B o x 1 for a brief description o f the methodology). 4. The analysis utilizes the variation among districts in the gross enrollment rate (GER) o f different income/consumption groups, to infer the impact o f expanding public expenditure o n the GER o f each o f these groups. To do so, one must estimate the marginal odds ratio o f enrollment (Lanjouw & Ravallion), which represents the change in the enrollment o f a given consumption group when there i s an increase in aggregate enrollment. As long as the size o f aggregate enrollment i s proportional to that o f public expenditure - an assumption that i s not perfect, but most “neutral” under the circumstances - one can infer from the marginal odds ratio o f enrollment h o w an increase in public expenditure on education will affect school enrollment o f each income group. I n addition, in order to obtain the true relationship between expanding public expenditure and enrollment o f the different groups, certain heterogeneous characteristics deemed important - namely private school availability, literacy rate and rural/urban - are controlled for. 5. The first three columns o f Table 5 show the estimated marginal odds ratio o f primary enrollment for different consumption groups, which can also be interpreted as the estimated impact o f 1 percent increase o f public expenditure on the GER o f different consumption groups (under the above-stated assumption that the size of aggregate enrollment i s proportional t o that o f public expenditure). These results indicate that the lowest three consumption groups receive the largest benefit from public expenditure expansion. 1 percent increase in total public expenditure increase enrollment rate by 0.9 percent for the bottom quintile, 1.1 percent for the second quintile, 0.9 percent for the third quintile and 0.6 percent for the fourth quintile. O n the other hand, expanding public expenditure does not have any significant impact on the public primary enrollment rate o f the richest group. Annex C Page 2 of 5 Box 1: A Standard Methodology for Dynamic Benefit Incidence Lanjouw and Ravallion (1999)71have proposed a method t o calculate the marginal odds ratio o f enrollment based o n a single cross-sectional survey including data o n school enrollment across districts within provinces. The procedure is as follows. First, the average school enrollment rate for each quintile and each district i s calculated. Second, an overall average school enrollment rate i s computed for each province. T o estimate the marginal odds o f enrollment by expenditure quintile, one has t o regress the quintile-specific enrollment rates across districts o n the relevant province’s average enrollment rate separately for each quintile. Furthermore, to address the correlation between a quintile-specific enrollment rate and an overall province-specific average rate, the leave-out mean i s used as a n instrument for the province-specific rate (see L&R, 1999 for details). There are several caveats. First, the size of aggregate enrollment i s interpreted as the size o f public expenditure if the subsidy rate for educational service i s constant not only across provinces and income groups,72 but also between before and after expansion of public expenditure. But this m a y not be the case. Second, it is assumed that a l l factors other than province-level average enrollment rate are identical across provinces.73 This assumption might be too strong. T o partly address t h i s problem, two province-level variables are included as explanatory variables inall regressions: literacy rate and private school enrollment rate. Province average literacy rate i s included because school enrollment i s positively correlated with parents’ education in Pakistan.74 The effect o f private school enrollment rate i s included because the impact o f expanding public expenditure might fade out as the private sector becomes more accessible. Third, the gender gap and the urban-rural gap i n education outcomes for Pakistan suggest there are some structural differences between male and female as w e l l as between the urban and the rural. T o address the urban-rural gap, a n urban dummy i s included as an explanatory variable in a l l regression^.^' To deal with the gender gaps, the marginal odds ratio of school enrollment rate i s regressed separately for each gender i n some 76 regressions. 6. Although expanding p u b l i c expenditure increases both m a l e and f e m a l e e n r o l l m e n t for the poorest a n d t h e second poorest i n c o m e groups, there i s a gender difference: t h e i m p a c t o f public spending on female enrollment appears to b e significantly l o w e r t h a n that o f m a l e e n r o l l m e n t (Table 1). 1 percent increase in p u b l i c expenditure for primary education raises the gender gap in t h e p o o r e s t i n c o m e group by 0.3 percent, s i m p l y by i m p a c t i n g b o y s ’ e n r o l l m e n t m o r e than girls’. 7. To some extent, t h e differential impact of p u b l i c primary spending on b o y s and g i r l s may b e explained by the fact that parental decisions to send their c h i l d r e n to school may b e affected by sociocultural factors, o v e r and above the availability o f schools ( w h i c h i s the factor that i s d i r e c t l y affected by p u b l i c spending), m o r e in the case of g i r l s and for boys. M o r e o v e r , t h e decisions to send g i r l s to school may also depend on t h e availability of certain facilities in schools, l i k e separate latrines and female teachers, m o r e than it i s the case for b o y s for o b v i o u s reasons - facilities that may b e q u i t e independent o f the amount of public spending. For p o l i c y m a k e r s in Pakistan, t h i s indicates an i m p o r t a n t issue to address - namely, to explore how to enhance t h e i m p a c t o f public spending on girls’ enrollments. To overcome instances of low incentives to send girl c h i l d r e n to school due to sociocultural barriers, targeted scholarships or subsidies can b e i n s t i t u t e d - p r o g r a m s l i k e these h a v e b e e n operational in countries l i k e Bangladesh and N e p a l with some success. 71 Lanjouw, P and M. Ravallion (1999), “Benefit Incidence, Public Spending Reforms, and the Timing o f Program Capture,” The World Bank Economic Review 13: 257-73. 72 Lanjouw and Ravallion (1999) shows that the differences are relatively small 73 Or all factors other than province-level average enrollment rate vary in ways that they are uncorrelated with province-level average enrollment rate. 74 Lanjouw et.al. (2001) also includes the literacy rate of province in the analysis for Indonesia’s public expenditure reform. Lanjouw, P., P. Menno, F Saadah, H. Sayed, and R. Sparrow (2001), “Poverty, Education, and Health in Indonesia:Who Benefits from Public Spending?” Policy Research Working Paper 2739, the World Bank: Washington, DC. 75 Furthermore, all province-level variables are computed separately for urban and rural districts within a province in all regressions. 76 In these regressions, all variables are computed separately for each gender as well. Annex C Page 3 of 5 8. The first three columns o f Table 2 report how 1 percent increase in province average private school enrollment rate affects district average public school enrollment rate. It i s worth noting that the public primary school enrollment rate for the poorest income group significantly grows by 0.6 percent as the private school enrollment rate grows by 1 percent. On the other hand, the public primary school enrollment rates for the richest and the second richest groups decline by 1 percent and 0.3 percent, respectively. Thus for the lower income groups, a larger private sector in primary schooling boosts public school enrollment; whereas, for higher income groups, the effect i s the opposite. 9. Finally, after the availability o f private schooling and literacy rate has been controlled for, the urban-rural gap does not seem to be significant. 10. Loolung at the marginal odds ratio o f secondary school enrollment, the third poorest consumption group receives the largest benefit fkom expanding public expenditure for secondary education. The gender gap o f secondary education i s larger than that o f primary education. For the poorest group, expanding public expenditure for secondary education does not increase female enrollment rate significantly (Table 1). As Table 2 indicates, availability o f private schooling does not have any significant impact o n public secondary school enrollment. Table 1. Marginal Odds of Primary and Secondary School Enrollment in Pakistan Public Primary School Public Secondaiy School Quintile Boys Girls Both Boys Girls Both 1 (poorest) 1.052** 0.773' 0.925** 0.902** 0.055 0.747* 2 1.097** 1.092** 1.121** 0.848" 0.475* 0.727** 3 0.898' 0.907 0.912* 1.176** 0.886* 1.295** 4 0.156 0.773 0.597* 0.096 0.486 0.054 5 0.096 0.388 0.145 -0.152 0.676 0.154 Note: n e table gives the instrumental variables estimates o f the regression coefficients of the quintile-specific primary (and secondary) school enrollment rates across districts on the average rate by province. The leave-out mean enrollment rate is the instrument for the actual mean. The numbers in parentheses are t-ratios. + refers to significant at 10%; + significant at 5%; *+ significant at 1%. Source: Authors' calculations based on the 1998-99 PIHS. Table 2. The effect of province average private school enrollment rate on district average P public school enrollment rate Public P r i m a v School Public Secondary School Quintile Boys Girls Both Boys Girls Both 1 (poorest) 0.599' 0.698 0.613' 0.530 -0.365 0.535 2 0.410 1.039' 0.560 0.169 0.7 15 0.210 3 0.177 11.982 0.238 1.370 0.063 0.986 4 -0.971** -0.167 -0.71 1* 0.681 -0.779 -0.810 5 -0.366 -0.675 -0.517' -1.995 -0.689 -1.259 _ (and Note: The table gives the instrumental variables estimates o f the regression coefficients o f the quintile-specific primary . secondary) school enrollment rates across districts on the average rate by province. The leave-out mean enrollment rate i s the instrument for the actual mean. The numbers in parentheses are t-ratios. + refers to significant at 10%; * significant at 5%; ** significant at 1%. Source: Authors' calculations based on the 1998-99 PIHS. 111. Simulations to Relate Targets with Public Expenditures 11. Using the analysis presented above, this section addresses the following question: what are the rough and preliminary estimates o f public expenditure required t o achieve a target GER (fixed at 80 percent o f gross primary enrollment rates) for the bottom two consumption expenditure quintiles by the fiscal year 2007/08? An important assumption underlying this exercise i s that the impact o f overall public expenditures on education, on the enrollment o f different incomehonsumption groups will Annex C Page 4 of 5 continue to be the same - over time and as public expenditures increase - as that revealed by the data o f 1998-99 (and estimated in the previous section). 12. Table 3 illustrates the results from this exercise. Using the coefficients from Tables 1 and 2, it i s estimated that overall public spending o n primary education should be at least Rs. 42.2 billion (at 1999/00 prices) to achieve 80 percent o f primary GER for the poorest quintile in 2007/08 FY, while Rs. 49.3 billion (at 1999/00 prices) i s needed to achieve the same for girls in the poorest quintile in 2007/08 FY. I t should be noted that these’ spending figures will be enough to ensure GERs o f 80 percent or above for corresponding groups among higher consumption quintiles. 13. Before examining the numbers in Table 3 in detail, it i s worth describing in some detail the restrictive assumptions that underlie these calculations. First, as mentioned in the previous section, the subsidy rate for primary education i s assumed to be constant in real term, which implies that public expenditure i s proportional to a province level primary enrollment rate. Second, the structure estimated above does not change from 1998/99 (when PIHS data was gathered) to 2007/2008. In other words, the impact o f expanding public expenditure (or private enrollment rate) in 2007/08 i s assumed to be the same as the number reported in Table 1 (or Table 2). If the estimate o f coefficient i s not significant, i t i s treated as zero. Third, the private enrollment rate o f a particular quintile i s assumed to grow from 1998/99 to 2007/08 at the same rate as between 1990/91 and 1998/99. Under this assumption, the expected private enrollment rate o f each quintile in 2007/08 i s computed. Fourth, the annual growth rate o f school-age population i s assumed to be 1.025, which i s the average annual growth rate o f population between 1991 and 2001. Given these assumptions, this exercise should be considered indicative, which does not take into account structural shifts, like a change in the demand conditions for schooling or a major change in the pattern o f private schooling - any o f which factors can lead to very different scenarios in terms o f what the public sector has to achieve. 14. Under these assumptions, Table 3 shows that Pakistan needs to increase the overall public expenditure for primary education by 73 percent in real terms (or Rs. 17.9 billion at 1999/00 prices) by the fiscal year 2007/08, in order to achieve 80 percent o f primary school enrollment rate for the poorest quintile. T h i s implies that Palustan’s government will need t o spend 0.96 percent o f GDP for primary education (assuming an annual growth rate o f GDP o f 4.5 percent) in 2007/08, while they spent 0.8 percent o f GDP for primary education in 1998/99. If the share o f primary education in total public expenditure o n education i s fixed at the current level (47 percent), Palustan’s government will need to increase total education expenditure from 1.68 percent o f GDP in 1998/99 to 2 percent o f GDP by 2007/08. A similar GER target for the second poorest quintile will necessitate a 39 percent increase in public primary education spending. Combining these two groups - i.e., in order to achieve a primary GER o f 80 percent for the poorest 40 percent o f the population - public primary education spending would need to be 55 percent higher in real terms than the current level, amounting to 0.86 percent o f GDP in 2007/08. If the share o f primary education in total public expenditure on education i s fixed at the 1998-99 level, totalpublic expenditure on education has to increase to 1.8 percent o f GDP by 2007/08. 15. The gender gap remains serious in these projections too. T o achieve the same goal for girls in the poorest income group, Pakistan’s governments need to double the overall public expenditure o n primary education in 2007/08 - amounting to 1.1 percent o f real GDP. Achieving the same goal for girls in the bottom 2 quintiles combined - the poorest 40 percent o f the population - will require a spending increase o f 85 percent in real terms by 2007/08, which will amount t o 1.02 percent o f GDP. Thus, if the share of primary education in total education expenditure i s fixed at the 1998-99 level, totalpublic expenditure on education has to increase from 1.68 percent o f GDP in 1998/99 to 2.1 percent o f GDP in 2007/08. 16. Achieving the same targets for boys would require far less. In particular, Pakistan’s governments will achieve the goal for boys in the bottom 40 percent o f the population by increasing real public Annex C Page 5 of 5 spending o n primary education by only 23 percent, an allocation that will amount to 0.68 percent o f the country's GDP, which i s less than the current share o f primary education in GDP (0.8 percent o f GDP). Table 3. Public Expenditure for Primary Education in 2007/08 to achieve 80 YOprimary GER for a given quintile Public ExDenditure Primary Primary All Education All Education Ratioa education Education 9819gd 2007/08e 1998/9gb 2007108' All 1.73 24.3 (0.80%) 42.2 (0.96%) 50.8 (1.68%) 87.9 (2.0%) The poorest quintile Girls 2.03 24.3 (0.80%) 49.3 (1.12%) 50.8 (1.68%) 103.1 (2.3%) The second poorest All 1.39 24.3 (0.80%) 33.8 (0.77%) 50.8 (1.68%) 70.6 (1.6%) quintile Girls 1.68 24.3 (0.80%) 40.9 (0.93%) 50.8 (1.68%) 85.3 (1.9%) All 1.55 24.3 (0.80%) 37.6 (0.86%) 50.8 (1.68%) 78.7 (1.8%) The bottom two quintiles combined B O Y S 1.23 24.3 (0.80%) 29.8 (0.68%) 50.8 (1.68%) 62.5 (1.4%) Girls 1.85 24.3 (0.80%) 45.0 (1.02%) 50.8 (1.68%) 94.0 (2.1%) Note: All public expenditure figures are billion Rs at 1999/00 prices. Parentheses refer to the share of real GDP (at 1999/00 prices) for the corresponding year. a. Ratio refers to the ratio of the public expenditure for primary education in 1998/99 FY to that in 2007108 FY. b. Public Expenditurefor primary education in 1998/99. c. Public Expenditurefor primary education in 2007/08 to achieve 80 YO gross enrollment rate for a given group. d. Public Expenditurefor all education in 1998/99. e. Public Expenditurefor all education in 2007108 to achieve 80 % primary GER for a particular group assuming the share o f primary education in total educational expenditure i s the same as in 1998-99. 17. This analysis reveals that the real challenge o f Palustan's governments i s h o w t o increase primary GER for children (especially girls) o f the poorest income group. Achieving such an increase in such a short time i s likely to be a tall order for any government, especially given the trends o f social spending in Palustan in recent years evident f i o m graphs presented in the previous section. 18. The analysis done so far to gauge the impact o f public spending o n education outcomes like enrollment assumes that the current conditions - rather, conditions prevailing in 1998-99 - will continue to prevail. Given that public spending has been highly inefficient in Palustan till date, which has stymied improvements in outcomes, i t i s thus natural that the current projections show that very large - and therefore very hard to achieve - increases in public expenditure would be needed to reach the targets in outcomes. On the other hand, i t can be argued that the same targets can be reached with less increase in the amount o f public spending, if the eficiency o f public spending were to increase - particularly in terms of targeting such spending to the lower income groups. STATISTICAL ANNEXES G s Lo m d 0 '2 0 2- v1 o w .- w % 2 .o .- e E 2 .- .- S E d 6 .= 04 .- 'cl .- P m' 4 i, i (I I m 4 .- 0 5 a0 .- .- D (/I m 2 m .- F .r 5 4 I 4 4 4 M r P 5 M E P N -001 O o 0 l 0 0 0N 0 0 0 0 - 0 - o o tN 0 0 m *" o o N w 0 0 0 oolN n 0 0 0 N 0 0 0 N o m - m m o tN o o t- N o o rN - E O 0 m m o N m Z O O 4 N O N * - w o m o moz E O 0 d - o o w 0 0 2 0 0 2 N O O N o o 0 zocr&o w w m 0" t d I n N - 0 g o 2 o m o - O N 0 t- m m 2 0 - " 1 0- 0 aN 0 - t - N O LNO O N. vi N z w - o " r- m o c o z W w r N r -- 0 " " 2 - m o mo 0 B N m o m b t - mt N O t N - O 9 r-" - N o w o m a w 0010 10 O N 0 lo - 0 - o m 0 N r- N 0 0 0 "" 0 d0 - N 0 0 0 o m 0 m o\ 0 0 0 m m o lo 0 0 0 N o r - N lo10 0 0 0 O N t - r-10 oor- m 0 0 2 d O O r- N m o o m W ; ' f ooz m o o m 10 O O N w N 0 0 % c) m o N 010 m o b 0 0 - m o Z 0 d N N m o m - t - 0 2 2 m m N m - N mW N 0 r- 0 - m m N b m - m ~ - w - m 0 m 0 g "" ~ O r-m w m t- 0 0 0 0 0 0 0 I 0 0 m 0 0 t - m ~ m m ~ O r - " 0 m w 0 0 0 0 0 0 0 0 b m w mi3-Z;" cor- N, G% I - N O m w N " b 0 ~ w 0 m ~ 0 0 0 0 0 o o m b N b t - b r - N m r - - m z w z? 3 I b m - N o o " " w 0 w P'N ?"In" N. w N m 0 0 N 0 0 0 0 - - s b m m O m m 0 b b r - 0 0 r - 0 b m 0 0 0 0 - r- m a - r - r - m m m - m m o m m t-- - 10 OP - d Y h VI in 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 - 0 0 0 0 0 - 0 0 0 0 0 2 0 0 0 00 m o 0 m 0 0 - 0" f 4 8 0000 0 0 - G 2 'c 0 0 0 0 0 0 0 0 - 0 0 - 0 0 - P 4 : : r 4 3 e U m - v i m v i - o o o o b m o t - c m - t - 3 " w m o o 0" v" i O 3" N * i o t - O ~ * m m O m 0 0 0 m 0 0 C m N m o w m v. 3. m N v i vr 02 N - . . 3 3 3 . - . c: i cri m 4 0 3 U 4 & M E P 2: V I V I O O O O ~ N 0 0 0 0 0 0 m 0 ~ 0 0 m o ~m N m t - 0 0 N O m 0 0 0 2 s -r- d o o o - o y ? m N ON O 0 O 0 0 I d N O O O O m d O O 2 2 - I N r - m 0 0 0 0 W r - 0 0 ~m m N N I I" 4 G M t P m k; m c P E 2 .- .- E VI b VI -c 0 e L M 2: o m m - m m * * m N N e m s. 0 m ~ 0 N Q l d N O w 0 0 ~ 0 ~ - w r - 0 t - o N o o m o o o o oo m -".e L o t - w m m m m 0 TIN" 0 d 0 0 0 0 0 0 d r- d 0 0 - 0 0 0 4 1 N N 0 0 0 0 0 0 - z i O O O O O O W w rm o m - 0 0 - o o o m 2" * o o - o o w ~ N 0 O a O m 0 ~ 0 0 0 - Z" r- m r- 0 0 - o o o mm 4 o o w o o o - m W o o momo o o ~ o - o m Z" * P N O O .r O O O O ; r- $1 m z- o * o " w w w m m o N o ~ o r - - 0 0 0 m o o m - o o d o m o - r - N r - m d m 20'; m m r - 0 m m m m r - 2. N wDN 5 d N 4 m o o m o m o - w m m v l m o o m o w o o Z G 8 hl .- c: .- 2 e c