Report No. 34942 Pakistan Country Assistance Evaluation February 17, 2006 Country Evaluation and Regional Relations Independent Evaluation Group Document of the World Bank Cont’d NPL Non-performing Loan NRM Natural Resource Management NWFP Northwest Frontier Province OED Operations Evaluation Department* OECF Overseas Economic and Cooperation Fund (now known as Japanese Bank o f International Cooperation (JBIC)) OEG Operations Evaluation Group (now known as IEG-IFC) OEU Operations Evaluation Unit (now known as IEG-MIGA) PER Public Expenditure Review PIA Pakistan International Airlines PIFRA Pakistan Improvement o f Financial Reporting and Auditing Project PIHS Pakistan Integrated Household Survey PPAF Pakistan Poverty Alleviation Fund PPAR Project Performance Assessment Report PRHS Pakistan Rural Household Survey PRSP Poverty Reduction Strategy Paper PSD Private Sector Development PSEDP Private Sector Energy Development Project PTLC Pakistan Telecommunications Corporation, Ltd. QAE Quality at Entry QAG Quality Assurance Group QOS Quality o f Supervision ROA Return on Assets ROE Return on Equity RVP Regional Vice Presidency SAC Structural Adjustment Credit SAL Structural Adjustment Loan SAP Social Action Program SAR South Asia Region SBA Stand-By Agreement SBP State Bank o f Pakistan SOE State Owned Enterprise TA Technical Assistance TCF Trillion cubic feet UNDP United Nations Development Programme USAID United States Agency for International Development WAPDA Water and Power Development Authority WBI World Bank Institute * OED has changed i t s official name to Independent Evaluation Group (IEG). The new designation“IEG” will be inserted in all IEG’s publications, review forms, databases, and web sites. CONTENTS Preface ................................................................................................................................. i Pakistan CAE Ratings. Fy94-03 .................................................................................... ... 111 S U M M A R Y ........................................................................................................................ v 1. Introduction and Country Background ..................................................................... 1 Historical and Political Context ............................................................................... 1 The Situation in the Early 1990s .............................................................................. 3 2 . The Bank’s Program .................................................................................................... 3 The Bank’s Strategy ................................................................................................. 3 Portfolio Management ............................................................................................. 6 Analytical and Advisory Activities (AAA) ............................................................. 7 Bank Partnerships .................................................................................................... 8 3 . Macroeconomic Stabilization ...................................................................................... 8 The Bank’s Strategy ................................................................................................. 9 Macroeconomic Outcomes .................................................................................... 10 Assessing the Bank’s Contribution ........................................................................ 11 4 . Poverty Reduction and Social Sector Development................................................ 12 The Bank’s Strategy ............................................................................................... 12 Poverty Reduction and Social Sector Outcomes ................................................... 14 Assessing the Bank’s Contribution ........................................................................ 16 Lessons and Recommendations ............................................................................. 17 5 . Sustainable Growth ................................................................................................... 18 The Bank’s Strategy ............................................................................................... 18 Energy and Infrastructure ...................................................................................... 19 Finance ................................................................................................................... 24 Agriculture and Natural Resource Management (NRM)....................................... 27 Assessing the Bank’s contribution ......................................................................... 30 Trade. Privatization. and Investment ..................................................................... 31 Summary ................................................................................................................ 32 6 . Governance ................................................................................................................. 32 The Bank’s Strategy and Program ......................................................................... 33 Governance Outcomes ........................................................................................... 34 Assessing the Bank’s Contribution ........................................................................ 35 Lessons and Recommendations ............................................................................. 37 7 . Conclusions, Lessons and Recommendations .......................................................... 37 Overall Assessment ................................................................................................ 37 Lessons ................................................................................................................... 39 Recommendations ................................................................................................. -42 Boxes B o x 1.1. Rebasing o f National Accounts Statistics ........................................................................ 2 B o x 5.1: The Independent Power Projects .................................................................................... 20 Figures Figure 5.1: M 2 and credit in Pakistan ( ?%I o f GDP) ........................................................................ 25 Figure 5.2: Pakistan: Gross fixed capital formation (% o f GDP) 1993/94-2003/04 ...................... 32 Tables Table 2.1 : W o r l d Bank Lending to Pakistan (Sector Allocation) .................................................... 4 Table 2.2: IBRD/IDA Commitments FY94-03 (US$ milions) ........................................................ 5 Table 2.3: IEG Project Evaluation Ratings Approved FY94-03 ..................................................... 7 Table 2.4: Annual Budget Allocations (% o f Total) ........................................................................ 7 Table 3.1 : Structural Adjustment Operations in Pakistan FY04-03 ................................................ 9 Table 3.2: Selected Macroeconomic Indicators ............................................................................. 10 Table 3.3: Remittances (US$ millions) .......................................................................................... 11 Table 4.1: Social sector projects, FY94-03 .................................................................................... 13 Table 4.2: Selected Social Indicators, 1993-2002 ......................................................................... 15 Table 4.3: Government Allocations and Expenditures o n S A P (Percent o f GDP) ........................ 15 Table 5.1 : Infrastructure operations FY94-03 ............................................................................... 19 Table 5.2: Financial Sector Operations in Pakistan FY94-03 ........................................................ 24 Table 5.3: Selected Banking Sector Statistics ................................................................................ 26 Table 5.4: Agriculture and Natural Resource Management Operations in Pakistan FY04-03 ......28 Table 6.1 : WBI Governance Indicators Percentile Rank: Control o f Corruption .......................... 33 Annexes Annex 1: Structural Adjustment Loans and Credits ...................................................................... 43 Annex 2: Statistical Tables ............................................................................................................ 46 Annex 3: List o f People M e t .......................................................................................................... 69 Annex 4: Guide t o IEG’s Country Evaluation Rating Methodology ............................................. 76 Annex 5: Management Action Record .......................................................................................... 80 Annex 6: Comments f r o m the Government .................................................................................. 81 Annex 7: IEG Response................................................................................................................. 87 Annex 8: Chairman’s Summary .................................................................................................... 91 Bibliography .................................................................................................................................. 95 i Preface This report examines World Bank assistance to Pakistan during the period 1994- 2003. I t analyzes the objectives and content o f the Bank’s assistance program during this period, the outcomes in terms o f economic and social development in Pakistan, and the contributions o f the Bank and other development partners to development outcomes. The report i s based on a review o f project files, economic and sector reports, implementation completion reports (ICRs), Project Performance Assessment Reports (PPARs) and other IEG evaluations, Quality Assurance Group (QAG) ratings o f quality at entry for Pakistan projects and economic and sector reports, and interviews with Bank staff. The Country Assistance Evaluation (CAE) also incorporates work on background papers in macro- economic management, financial sector, rural and environment, infrastructure, and poverty and social sectors. The CAE team visited Pakistan in December 2003 and April 2004 for discussions with Pakistani officials, representatives o f other development partners, the private sector, academia, NGOs, and staff o f the World Bank field office. A l i s t o f people met i s provided in Annex 3. Their cooperation and assistance i s gratefully acknowledged. Comments from the Bank’s Regional Management have been incorporated in the report. The report was also sent to the Pakistan authorities whose comments are reflected in the report and attached in Annex 6. IEG’s response to the Government’s comments is attached as Annex 7. A discussion o f the CODE Subcommittee was held o n November 28,2005, and a summary o f this discussion i s attached as Annex 8. This evaluation was prepared by Lily Chu (IEGCR), the Task Manager. Background papers, as inputs to the CAE, were prepared by Adil Kanaan, Geoffrey Fox, James Brown, Richard D. Stem, Manuel Hinds (consultants), and the Center for Global Development (Nancy Birdsall, M i l a n Vishnav, and Adeel Malik). Peer reviewers were Rene Vandendries (IEGCR), Mustapha Nabli (MNSED), and Professor Ayesha Jalal (Tufts University). Brenda Manuel (consultant) provided input on portfolio management and data analysis. Gulmira Karaguishiyeva (consultant) provided research and data support. Roziah Baba provided administrative support. .. 11 111 Pakistan CAE Ratings, FY94-03 I Country Assistance Outcomes of Bank Assistance Program Objectives Ratings for Bank Program Objectives Assistance Program Outcomes Macro Macroeconomic Stability. Deterioration or stagnation in most Moderately Satisfactory indicators has been followed by improvement in some areas, including restructuring debt, rebuilding reserves, and controlling inflation. Fiscal deficits have been reduced but remain high, tax mobilization remains flat, and expenditure reform has not yet had a major impact (paras. 3.6-3.7). Social Poverty Reduction and Social Sector Development. Poverty has Unsatisfactory increased (para 4.7) and the urban-rural poverty gap has widened (para. 4.8). Some social indicators (childhood immunization, fertility, female primary enrollment, literacy) have improved, but s t i l l lag comparator countries (para. 4.7). Growth Sustainable Growth Moderately Satisfactory Infrastructure Infrastructure. Increased power generating capacity, but sector Moderately Unsatisfactory finances have worsened, restructuring has been slow, and allegations o f corruption harmed perceptions o f the investment climate (para. 5.7). Some policy reforms have been initiated, especially in the o i l and gas sectors and highways, although some distortions remain (paras. 5.8- 5.10). Finance Finance. Banking sector has been strengthened, and capital markets Satisfactory have grown, although credit to the private sector has not changed substantially (paras. 5.26-5.27). AgricultureNRM Agriculture/Natural Resource Management. Agriculture Unsatisfactory production has increased, but yields remains l o w and water use i s inefficient. Rural poverty has not improved, as land inequality has increased, and access to other key inputs such as water and credit remains unequal (paras. 5.38-5.40). Trade, Privatization, and Trade, Privatization, and Investment. Tariffs have decreased Moderately Satisfactory Investment substantially, and have been simplified (para. 5.45). Privatization i s proceeding, but uneven pace has led to the continued drain o f Government resources to support SOEs (para. 5.46). Gross capital formation decreased through much o f the review period, and has just begun to rebound. Foreign Direct Investment remains l o w at 1 percent o f GDP (para. 5.47). Governance Control o f corruption has improved, but s t i l l affects the lives o f Unsatisfactory many Pakistanis (paras. 6.7-6.1 1). Governance issues still impede fiscal reforms (paras. 6.12- 6.13). t-- Overall Moderately Unsatisfactory V SUMMARY 1, After outpacing i t s South Asian neighbors in GDP growth over the first four decades o f i t s existence, in the 1990s Pakistan began to trail them, and also lagged on a number o f key social indicators. Political instability, compounded by exogenous factors such as a cotton virus, floods, and economic sanctions, contributed to a deteriorating economic situation. Since 1999, a stable government, consistent reform policies, and greater international support have helped the country move back o n to a better economic track; macroeconomic indicators have improved, although social indicators still lag. 2. This evaluation covers the period FY94-FY03. The World Bank i s an important source of funding for Pakistan, with IBRD and IDA loans representing approximately 28 percent o f outstanding external debt in 2004. Historically, the Bank had supported traditional sectors such as energy, infrastructure and agriculture; the review period saw a shift to a greater emphasis o n public sector reform and the social sectors. Four overriding themes consistently appeared in the country assistance strategies during the period o f this review: (i) macroeconomic stability; (ii) poverty reduction and social sector development; ( iii)sustainable growth; and (iv) governance. Each was highly relevant to the development needs o f the country. 3. Throughout the review period, the Bank provided substantial support, primarily through adjustment loans and AAA, for a broad-based macroeconomic reform agenda. The most critical area o f Bank focus was o n public finance, in particular on tax mobilization and public expenditure management, which were also to be underpinned by a broad range o f longer t e r m structural reforms. 4. At the end o f 1999, a new Government began a consistent reform program, which attracted international support. After September 11, 2001, Pakistan also was the recipient o f increased aid flows, debt rescheduling, and increased remittances. All these factors have combined to improve the country’s fiscal situation. Reserves are at n o w at over six months o f imports. Public debt has fallen but remains high at 69 percent o f GDP (according to the Government’s rebased numbers). Inflation (CPI) fell from 10 percent in 1993 to an estimated annual rate o f 4 percent in 2003/04, but i s estimated to have increased to over 7 percent in 2004/05. The budget deficit has also improved moving from 6-8 percent o f GDP (excluding grants) in 1990s to 4 percent in 2004/05. The Bank’s advice and lending have contributed to the current positive fiscal outlook, helped avoid default o n the public debt, and had a positive impact o n some o f the longer term structural reforms (such as financial sector reform and trade, discussed below under growth), but other areas o f Bank focus, including tax mobilization and improved efficiency in expenditure management, have been slow to have effect. Outcomes in this area were moderately satisfactory. 5. The Bank’s second major area o f focus was poverty reduction and social sector development through focusing on the expansion and quality improvement o f social services. Although some of the Bank’s programs did appear to help assist the Government to make progress in a number of areas (school enrollment, literacy, immunization, fertility, and child mortality), the country s t i l l lags i t s neighbors and countries o f similar income levels, and the Bank recognizes that much needs to be done if Millenium Development Goals (MDGs) are to be achieved. In addition, although poverty may have declined since 2000, it remains above the vi poverty level at the beginning o f the period under review. While the programs were constrained by the intense fiscal problems faced by the governments during this period (with high deficits and soaring debt, successive governments were hard pressed to fund social programs), they also were weakened by a lack of a clear strategy to address the roots o f poverty, and poor program implementation. Outcomes in these important areas were unsatisfactory. 6. Growth was the third major objective o f the Bank’s program. The overall strategy was to support a shift from public sector ownership and management to the private sector. Specific areas o f focus included the financial sector, infrastructure, rural development and natural resource management, trade, privatization, and improving the environment for private sector investment. Although Pakistan’s GDP growth has picked up in the last three years, outcomes were mixed: (i) banking sector reform has been proceeding well; (ii) agricultural production has improved, but rural poverty has not decreased, as unequal access to land, water, credit and other inputs remain unaddressed. In addition, despite Bank support for large irrigation and drainage programs, insufficient attention to natural resource management has meant that water-logging, soil salinity, and inefficient water use may limit agriculture growth in the future; ( iii) the Bank’s support o f trade reform has shown good results as tariffs have decreased and been simplified; (iv) good progress also has been made in privatization, particularly in banking, but the uneven pace in other areas has led to a continued drain o f Government resources to support some major state-owned enterprises; and (v) investment has started to rebound, but gross capital formation and foreign direct investment have not changed appreciably since the beginning o f the review period. 7. Infrastructure support was also a key component o f the strategy to support growth. The capacity o f the power sector was increased, albeit at a price which was likely higher than could have been obtained if contracting and reforms had been handled better. The contracting process for power production also raised allegations o f corruption and caused renegotiation o f contracts with several private companies, which damaged the country’s investment climate reputation. Recent reforms have begun in o i l and gas, highways, and telecom. However, fixed infrastructure s t i l l remains inadequate to support the growth needs o f the country, and hndamental financial problems remain in the power sector, the primary focus o f the Bank’s infrastructure work. Overall, outcomes for the Bank’s program in these areas designed to support growth were moderately satisfactory. 8. The fourth major area o f Bank focus was improving governance. However, the Bank had difficulty defining a clear strategy. In early Bank strategy documents, governance was raised as a problem but there were no actions or projects planned. Over time, a more concrete approach began to develop, as projects and AAA focusing o n administrative governance and strengthening institutions emerged. Outcome indicators were not clearly defined. However, in areas where the Bank has focused, such as tightening o f fiscal discipline or improvement in the delivery o f basic services, there were n o appreciable improvements, resulting in unsatisfactory outcomes. 9. In sum, the outcomes o f Bank support for macroeconomic management and growth have achieved success in some areas, especially in the last few years. However, outcomes o f Bank assistance were unsatisfactory in poverty reduction and social sector development, governance, agriculture and natural resource management, fixed infrastructure, and revenue mobilization and vii expenditure management. Therefore, overall outcomes o f the Bank’s assistance program are rated moderately unsatisfactory. 10. W o r l d Bank project performance has been uneven. The Bank portfolio improved greatly after a portfolio clean-up in 1999. Bank documents show that problems o f commitment and institutional capacity appeared consistently as major impediments to project implementation and sustainability; yet, project design did not do enough to take these ongoing problems into account. 11. The Bank’s analytical work was generally o f good quality. However, i t s relevance and timeliness could be improved through strategic work in areas such as rural development, the social sectors, and power. In addition, in the early part o f the review period, insufficient resources were spent o n ESW. The Bank i s now devoting more resources to analytical and strategic work, which should improve project performance and overall outcomes. 12. Going forward, the Bank would need to focus o n the following priorities (as discussed further in Chapter 7): (i) continuation o f i t s strong support o f analytical work, augmented by translating the analysis into implementable and prioritized actions and programs, and taking into . account political economy constraints. Priorities are in poverty, rural development, and governance; (ii) greater focus o n building sustainable institutional capacity; ( iii) narrowing the scope o f projects, and scaling them to fit the capacity o f the implementing agencies; (iv) improving partnerships. Vinod Thomas Director-General Independent Evaluation Group ... Vlll 1 1. Introduction and Country Background 1.1 Pakistan has a population o f about 148 million, and, based on new Government estimates, GNI per capita was estimated at US$638 in 2003/04. After outpacing i t s South Asian neighbors in GDP growth over the first four decades o f i t s existence, Pakistan began to trail them, and has not fared as well on a number o f key social indicators. Pakistan achieved growth rates o f over 6 percent in the 1980s, but the growth rate declined to about 4 percent per year during the 1 9 9 0 representing ~~ average per capita GDP growth o f only 1.4 percent. Export growth fell from over 10 percent per year in the 1970s and 1 9 8 0 to ~~under 3 percent per year in the 1990s. The economic situation has improved in the last few years. GDP growth rebounded to 5.1 percent in 2002/03, and 6.4 percent in 2003/04. Total government debt has fallen from a high o f 108 percent o f GDP in 2000/01 to 84 percent in 2003/04, government deficits (excluding grants) have declined from 8.9 percent in 1993 to 4 percent in 2003/04, and gross reserves increased from less than one week o f imports in 1998 to more than six months o f imports in 2003/04. 1.2 This evaluation covers the Bank program during FY94-FY03, which represents a difficult period in the economic path o f Pakistan. This report will review the Bank’s strategy during this period. In particular, it addresses the following questions: (i)did the Bank correctly assess the problems Pakistan faced; (ii) was the Bank’s strategy appropriate for meeting Pakistan’s development needs; (iii) h o w effective was Bank assistance in implementing those strategies; (iv) what were the outcomes o f the assistance; (v) to what extent did contributions to outcomes involve the Bank, other development partners, the Government, as w e l l as exogenous forces? 1.3 The review i s structured as follows: this section will provide the country context in which the Bank strategy was designed. Section 2 will discuss the overall Bank program. Sections 3-6 will each discuss one o f the major country assistance strategy (CAS) themes (Macroeconomic Stability, Poverty Reduction and Social Sector Development, Sustainable Growth, and Governance). Each o f the thematic sections will: (i)assess the relevance of the objective; (ii)summarize the Bank strategy assistance related to the objective; (iii)summarize outcomes, as well as achievement o f some o f the project specific objectives; (iv) assess the Bank contribution to those outcomes; and (v) draw lessons from those findings.’ Section 7 will present overall assessments, lessons and recommendations. Historical and Political Context 1.4 The history of modem Pakistan has strongly affected its economic growth path. Due to regional tensions, military expenditures in Pakistan have consistently absorbed a ’ Each o f the country assistance strategy (CAS) themes i s intrinsically related to the other themes. For example, the fiscal situation affected the Government’s ability to invest in the social sectors or in gross capital formation; governance problems led to l o w revenue mobilization, poor use o f funds in the social sectors and real sectors; slowing growth deepened the government’s fiscal problems. However, in the interests of brevity and structural clarity, certain issues, which could be addressed in multiple parts o f the paper, have only been addressed in one area. 2 significant proportion o f the budget (about one-quarter to one-third o f total revenue; see Annex Table 3). Combined with interest expense which amounted to another 30-45 percent o f revenue; there remained very little fiscal space for basic government expenditures, or development expenditures. Box 1.1: Rebasing of National Accounts Statistics In the spring o f 2004, the Federal Bureau o f Statistics completed a revision o f Pakistan’s national accounts statistics. The objective was to bring Pakistan’s national accounts closer in line w i t h the 1993 UN Systems o f National Accounts. As part o f the exercise, the base year was moved from 1980181 to 199912000. The new statistics capture better the changes that have occurred in the Pakistani economy in the past 20 years, as several areas o f economic activity had been either seriously underestimated or not captured at all. The rebasing exercise resulted i n GDP estimates for 199912000 and following years that were almost 20 percent higher compared to the old base estimates. This, however, leaves the fundamental policy assessment largely unchanged. Per capita income, while somewhat higher, remains low. The debt-to-GDP ratio i s lower, but i s still relatively high. Revenue collection and social spending now appear even lower in relation to the new GDP, underscoring the need for continued policy efforts. For the purposes o f this evaluation, the previous statistics are generally used (rebased numbers w i l l be noted when used), as the analysis and decisions made by the Bank during the review period were based on the previous statistics, and rebased numbers are not available for the full period. Pakistan: Effects o f the Re-Basing o f national Accounts 2002103 2003/04 2004105 estimated projected Old New Old New Old New (In percent of GDP; unless indicated otherwise) Consolidated Public Finances Revenue (including grants) 20.8 17.3 18.5 15.1 18.3 14.7 Expenditure 22.4 18.7 20.7 16.9 21.2 17.0 Budget balance (including grants) -1.7 -1.4 -2.2 -1.8 -2.9 -2.3 Budget balance (excluding grants) -4.5 -3.7 -2.9 -2.4 -4.0 -3.2 Total government debt 89.2 74.3 84.2 68.7 79.1 63.6 External sector Current account (including official transfers) 6.1 5.1 2.4 2.0 -0.9 -0.7 Current account (excluding official transfers) 4.6 3.8 1.7 1.4 -1.3 -1.1 GDP at market prices (billions of Pakistani rupees) 4,018 4,821 4,455 5,458 4,960 6,164 GDP at market prices (billions o f US$) 68.8 82.6 77.5 94.9 84.1 104.5 Per capita GDP (in US$) 471 566 521 638 555 690 Source: IMF Staff Report for the 2004 Article IV, November 16,2004. Additional text by World Bank staff. 1.5 In addition, Pakistan’s international relations, particularly with the West and primarily the United States, have had a very large impact o n its aid flows, which in turn have affected i t s macroeconomic performance. During the AfghadSoviet war in the 3 ~~ 1 9 8 0 Pakistan received large inflows o f concessional aid. However, aid declined after the war ended and later, when Pakistan conducted nuclear tests in 1998, aid was cut sharply and sanctions were imposed that influenced the programs o f multilateral institutions, including the Bank. Following the events o f September 11, 2001, Pakistan’s cooperation with the West, especially with the U.S., led to large amounts o f aid from the U.S. and other bilateral donors. 1.6 Pakistan’s economic situation has also been affected by the complexities o f domestic politics and weak management in fiscal expenditures. Major factors included periods o f domestic political instability, with shifts in government,* accompanied by reversals in major, system-wide economic policies by the different governments; widespread corruption; weak and uneven human resources development; law and order and other stability issues; and persistently l o w levels o f domestic savings. The Situation in the Early 1990s 1.7 Despite the relatively high growth rates leading up to the 1990s, poverty was s t i l l widespread, and social indicators were poor (see Annex Table 2). Pakistan’s population growth rate o f over 3 percent per year increased pressure o n the environment and o n investment needs for future growth. Due to the l o w savings rate and productivity (poor quality o f human resources), growth relied heavily on both domestic and foreign borrowing. Budget deficits o f about 6 percent o f GDP and dependence on foreign aid were the norm. As long as domestic and international interest rates were relatively low, the servicing o f this debt was manageable; but as these rates increased, the economy became hostage to a tenacious debt and financing gap spiral and plunged into foreign exchange crises in 1993 and again in 1997/98. 2. The Bank’s Program 2.1 Pakistan became a member o f the W o r l d Bank in 1950, and began borrowing from the IBRD in 1952, and from IDA in 1962 when IDA was formed. A s o f June 30, 2003, the Bank had made total commitments to Pakistan o f US$18.2 billion for 203 projects, which makes it one o f the Bank’s ten largest borrowers. From FY94 to FY03, commitments amounted to US$4.7 billion for 35 projects. The Bank i s a major source o f funding for Pakistan, with IBRD and IDA loans representing approximately 28 percent o f outstanding debt in 2004, and between 1993 through 2003, the Bank represented about 22 percent o f total donor aid and multilateral loan flows (see Annex Table 11). The Bank’s Strategy 2.2 Bank Programs during the last decade were guided by country assistance strategies in FY92, FY94, FY95, and FY02 as well as country assistance strategy progress reports in FY97, FY99, and FYOl. While there was some variance among the documents, the dominant and consistent themes were: (i)Macroeconomic Stabilization; * Annex Table 4 lists the major government changes from 1978-present. 4 (ii)Poverty Reduction and Social Sector Development; ( iii) Sustainable Growth; and (iv) Governance. In the early years o f the review period, the Bank also identified Environment as a major area where assistance was needed, including water management, natural resource conservation, and developing solutions to urban pollution problems. However, the focus on environmental work faded over the course o f the review period. Given the problems facing Pakistan, each o f the four primary themes was relevant and will be discussed in more detail in later sections. Environment, which i s also a highly relevant theme, was gradually subsumed into the overall growth agenda. 2.3 These themes were reflected in a shift in the lending portfolio from agriculture, energy and infrastructure prior to the 1990s toward programs which would bring direct benefits to the poor (see Table 2.1). In addition, public sector management and structural adjustment showed large increases, with adjustment loans in the last three fiscal years accounting for 70 percent o f the lending. Lending also shifted from a blend o f 70/30 IBRD/IDA in the early 1990s to 100 percent IDA in FYO1-03, as most o f Pakistan’s creditworthiness indicators were too l o w for IBRD lending. Table 2.1: World Bank Lending to Pakistan (Sectoral Allocation) FY52-93 FY94-03 Maior Sector Total As a % of Total As a % of Total Commitment Total Commitments’ Commitment s (US$ Commitment (US$ million) million) Agriculture and Natural 2,461 18% 429 9Yo Resources Energy & Infrastructure 5,091 38% 915 19% Finance 1,698 13% 793 17% Social Industry and Trade 1,724 926 13% 7yo 907 --- ___ 19% Public Sector 1,041 8% 1,664 35% Other 412 3% Overall Result 13,534 100% 4,712 100% Source: Business Warehouse. Loans categorized by primary sector. ’Excludes guarantees and grants. 2.4 In general, lending scenarios were optimistic. The 1992 strategy projected a strong performance by Pakistan in carrying out a reform program that would promote private sector activity and efficiency while improving government fiscal performance. The 1994 strategy did not change appreciably; while it acknowledged that the 1993 crisis reflected a weak fiscal situation, i t expected a rebound in perfomance. 2.5 At the time o f the 1995 strategy, the IMF program had gone o f f track. The strategy stated that given the “intense public jockeying between the current and former ruling parties, lower GDP, and higher inflation”, the Government had “chosen to slow the pace o f reform.” However, the strategy argued that since the Government had not “back- tracked,” an abrupt decline in lending was not appropriate. Despite the weak reform environment, the strategy outlined four lending scenarios ranging from the l o w case o f US$200 million in annual IDA commitments to a high case o f up to U S $ l billion, with likely lending scenarios in the US$500 to US$750 m i l l i o n range. Actual commitments began to fall o f f in late FY96 and FY97, reflecting Pakistan’s poor performance. The 1997 strategy noted that “due to the weak policy performance and deteriorating 5 indicators” (most notably the macro and credit worthiness indicators), lending was expected to b e in the “low-base case” o f about US$300-US$400 million per year. Table 2.2: IBRDDDA Commitments FY94-03 (US% millions) FY94 FY95 FY96 FY97 FY98 FY99 FYOO FYOl FY02 FY03 Planned Base Base Base Base Low High High High High Base Scenario Base Base Base Base Base Expected 600- 600- 500- 500- 300- 890 805 765 750 400 Lending 700 700 750 750 400 Actual Commitments IBRD 380 466 385 250 350 IDA 362 240 75 85 558 90 - 374 800 297 Total 742 706 460 85 808 440 - 374 800 297 % 34% 0% 0% 0% 31% 68% - 93% 63% 69% adjustment lending Source: Business Warehouse, country assistance strategy documents - planned scenarios reflect most recent preceding country assistance strategy or progress report. 2.6 Following Pakistan’s nuclear tests in M a y 1998, economic sanctions were placed o n the country; new bilateral and multilateral lending for non-basic human needs was ~uspended;~ and investment flows as well as aid decreased sharply, leading to a severe foreign exchange crisis. The 1998 strategy progress report acknowledged large estimated financing gaps, and relied on assumptions o f financing support f i o m multilaterals, as well as rescheduling o f bilateral debt and commercial debt, but proposed that the volume o f Bank commitments be linked to: (i) the strength o f Pakistan’s reform effort; (ii) the maintenance of macroeconomic stability; (iii) progress in the social sectors; and (iv) improvements in the country’s creditworthiness. The 1998 strategy progress report argued that the new Government (elected March 1997) was truly committed to reform although the new Prime Minister had not sustained reforms during his previous administrations, and macro performance under his current government was mixed. Despite the past history o f failing to follow-through with reforms, weak macro environment, severely deteriorating creditworthiness, and minimal progress in the social sector, the Bank deemed Pakistan to be in the “high-base case” scenario, qualifying it for about US$750 million in annual lending. 2.7 An immediate crisis was averted. However, by spring 1999, “there were concerns that the government’s commitment to the reform program was faltering. In particular, resource mobilization and governance improvements remained weak.”4 I n September 1999, the IMF formally suspended the Enhanced Structural Adjustment Facility (ESAF) program. In 1998/99, total debt to GDP was 92 percent; by 2000/01, it had increased to 108 p e r ~ e n t . ~ World Bank disbursements for existing projects continued. Also, the definition of “non-basic human need” was not clearly stated, which allowed a number of loans to proceed. World Bank. 2001. Pakistan: Countiy Assistance Strategy Progress Report (Report no. 22219-PAK). IMF estimates. 6 2.8 The Musharraf Government took power in October 1999. It inherited a very vulnerable macroeconomic situation, and, given the weak reform performance by previous governments, difficult relationships with lenders and donors. With the deteriorated economic situation, Bank lending dropped to zero in FYOO. However, in November 2000, the Government reached a new Stand-by Arrangement with the IMF, and negotiated another Paris Debt Restructuring (US$1.8 billion) in January 2001. The Bank restarted lending in FYOl, 2.9 The Bank Country Assistance Strategy Progress Report (CAS PR) in M a y 2001 and the 2002 Country Assistance Strategy (CAS) had many o f the same themes as previous CASs (fiscal improvement, governance, removing distortions that impede growth, improving service delivery). However, the 2002 CAS, perhaps because i t was the f i r s t full C A S in seven years, made more use o f lessons from past experiences to influence the program. The Bank would focus more on policy dialogue with the Government, and o n the need for better implementation capacity, not just at the federal, but at the provincial and district levels. Portfolio Management 2.10 The Region initiated Country Portfolio Performance Reviews (CPPRs) in FY90 to manage the portfolio and monitor i t s performance. In spite o f efforts to improve project performance, problem projects increased from 7 percent o f the portfolio in FY94 to 34 percent in FY98. In FY98, Pakistan was ranked as one o f the 25 worst performers Bank- wide with more than 50 percent o f projects and commitments “at risk.” 2.1 1 In December 1997, the Bank, together with the federal and provincial governments, initiated an “aggressive portfolio improvement strategy” and a new portfolio and risk management process based o n “outcomes, ownership and good governance.” Quarterly national portfolio review meetings and periodic provincial portfolio reviews were established and continue to be held. Problem projects are further reviewed at monthly portfolio meetings. Portfolio management functions were increasingly decentralized when the Country Management Team, including the Country Director, was moved to the field in 1997. 2.12 T w o portfolio improvement plans were developed in FY99 and FYOO. As a result o f the portfolio improvement reviews, the number o f projects in the portfolio declined from 44 in FY93 to 13 in FY03. Two actions, in particular, resulted in the closing o f 32 projects and in few entrants to the portfolio, namely: (i) a “no-extension” policy outside offorce majeure; and (ii) enhanced selectivity and “quality at entry” criteria.6 At the end o f FY03, projects at risk were 8 percent o f the portfolio, compared to 18 and 15 percent for the region and Bank-wide, respectively. The Bank dropped from the portfolio or pipeline all operations which were not driven by “strong” client ownership and/or those which did not have clearly defined development impact/outcomes. 7 Portfolio Performance 2.13 Of 24 closed projects approved from FY94 to FY03 (for a total value o f US$3.4 billion in commitments), IEG rated 84 Table 2.3: IEG Project Evaluation Ratings percent o f commitments satisfactory Approved FY94-03 for outcome, slightly above South Satisfactory Likely Substantial Outcome Sustainability Inst. Dev. Impact Asia’s rating o f 83 percent, and 80 YO % % percent Bank-wide.’ Sixty-eight Pakistan 84 68 8 percent o f the Pakistan projects were South Asia 83 81 38 judged as likely to be sustainable, and Bank-wide 80 79 48 8 percent were rated as having substantial institutional development Source: Business Warehouse impact, both significantly below regional and Bank-wide averages. Supervision, implementation completion reports, and CPPRS have also frequently raised issues o f lack o f commitment and weak institutional capacity. Additional attention should have been given to assessing the incentives for the counterparts, and designing projects that take into account the weak institutional capacity, as well as providing assistance to improve that capacity. Other systemic issues included governance, financial management, procurement, project management and counterpart financing. 2.14 During the C A E period, QAG reviewed three Pakistan projects for Quality at Entry (QAE) and rated all three as satisfactory. QAG’s review o f 22 projects for Quality o f Supervision (QOS) judged the supervision o f 56 percent o f those projects as satisfactory compared to 80 percent for both South Asia and the Bank as a whole. I t is, however, difficult to derive conclusions about Pakistan’s portfolio based solely on QAG assessments because the sample o f projects rated for QAE was quite small and the most recent QOS were three projects reviewed in FYOO. Analytical and Advisory Activities ( A M ) 2.15 As the Bank reduced i t s portfolio, resources were freed up for AAA. Table 2.4 shows the shift; in FY94, only 17.7 percent o f Bank resources were used for AAA; by FY03, 34.3 percent was allocated for AAA. Table 2.4: Annual Budget Allocations (YOof Total) FY94 FY95 FY96 FY97 FY98 FY99 FYOO FYOI FY02 FY03 Supervision 39.5 38.2 43.4 52.2 50.6 56.7 60.8 36.6 27.7 36.4 Lending 42.9 40.3 37.3 29.8 31.2 29.8 13.6 16.4 40.4 29.3 Preparation m , 17.7 21.5 19.3 18.0 18.2 13.5 25.6 47.0 31.9 34.3 Total 100 100 100 100 100 100 100 100 100 100 Source: Business Warehouse Subsequent to the initial release o f this report to the Bank’s Executive Directors, audits on seven loans in the portfolio were carried out. Three adjustment loans, accounting for US$700 m i l l i o n in commitments, were downgraded from satisfactory to unsatisfactory. As a result, the percentage o f commitments which were rated as satisfactory during the exit period covered by the CAE was 67 percent, or substantially below Bank or regional averages. See Annex 1 for a more detailed discussion o f adjustment loans. 8 2.16 QAG reviewed 12 o f 3 1 AAA tasks completed in the FYO1-03 period. Two tasks were rated highly satisfactory, nine were satisfactory, and one was unsatisfactory. IEG concurs with the positive assessment o f recent AAA. However, in the early years o f the review period, although the quality o f AAA was good, its coverage and relevance were weak. As discussed below in sector specific chapters, there were notable gaps in AAA, including a lack o f strategic work in rural development, the social sectors, power, and governance. Sometimes, as with infrastructure, much o f the needed work was done in the context o f lending preparation and supervision. 2.17 Candor o f E M One issue that has been raised by both Bank staff working on Pakistan and others outside the Bank, including academics and journalists, i s the trade-off between partnership with the Government and the objectivity and candor o f Bank’s analytic work. In reviewing and discussing ESW with Government officials, the Bank needs to weigh the trade-offs involved between openness and candor o n the one hand, and client partnership on the other. Bank Partnerships 2.18 Donor agencies and many NGOs respect the World Bank, and felt that the World Bank played an important leadership role. There was also praise for a number o f individual task managers or individual projects. The Bank has worked hard to develop complementary programs with other donors (e.g., focusing i t s provincial work on Sindh and NWFP, while ADB focuses on Punjab and Baluchistan). However, a number o f donors felt that the respect was not always reciprocated, and the Bank was not always easy to work with. Among issues raised were failure to consult with other donors before making key decisions; lack of continuity o f Bank staff interacting with the donors; intermittent participation o f senior staff at donor meetings; and poor information flow to donors and NGOs. In addition, the Bank i s perceived by many donors, NGOs, and members o f civil society to be a poor communicator o f its policy directions and recommendations and that i t does not adequately involve stakeholders outside government in p r o g r d p r o j e c t design, monitoring and evaluation. For example, several NGOs commented that they were presented with a fait accompli, rather than involved in substantive discussions. Some NGOs have instructed their staff not to attend Bank workshops, since the NGOs would be listed as participants, whether or not they felt their views had been adequately considered. 3. Macroeconomic Stabilization 3.1 Throughout the review period, macroeconomic stability was a k e y issue. In 1993/94, debt outstanding was approximately 93 percent o f GDP, debt service represented more than 23 percent o f exports, fiscal deficits (excluding grants) had reached over 6 percent o f GDP, and gross official reserves had declined to about two weeks worth o f imports. GDP growth had slowed to 1.8 percent in 1993 (or -0.8 percent GDP per capita growth) due in part to massive floods in 1992 and a cotton virus. However, after these crises, during the rest o f the decade average growth stayed under 4 9 percent, Pakistan seemed to be entering a vicious cycle, as increasing debt service and budget deficits limited the government’s ability to maintain infrastructure, invest in new capital formation or social development, all o f which were needed to move back onto a higher growth track. The Bank’s Strategy 3.2 The Bank’s country strategies reflected the country’s strong need for fiscal and structural reform. The Bank consistently pressed for tax reform, especially widening the tax base and eliminating exemptions, and adequate cost recovery for public services. The cost recovery components were meshed with a push toward greater private sector participation in the delivery o f services (both through privatization and through inviting private investment in new projects), especially in the energy and infrastructure area. Public administration, expenditure management, and financial sector were also priority areas for reforms. 3.3 During the early 1990s, the Bank’s main instrument for fiscal reform and structural adjustment was through dialogue and AAA. The Bank had been preparing an adjustment loan since 1990, but did not proceed because o f a weak macroeconomic framework. On September 23, 1993, to take advantage o f the reforms supported by an interim Government, a Public Sector Adjustment Loadcredit (PSAL/C; U S 2 5 0 million)8 was approved.’ Macro performance continued to decline, however, and the Bank refrained from economy-wide adjustment operations for several years,” returning to E S W and working with the IMF as the main instruments for support in this area (see Annex Table 5 for l i s t o f IMF programs). 3.4 In the aftermath o f Pakistan’s nuclear tests, a severe foreign exchange crisis developed in 1998/99. Despite the poor macro-reform progress, the Bank approved a US$350 m i l l i o n adjustment loan as part o f a comprehensive financing program with the IMF, ADB, and the Paris Club. Although the immediate crisis was averted, reforms were not sustained, and by September 1999, the IMF formally suspended the Enhanced Structural Adjustment Facility ( E S A F ) program. Table 3.1 : Structural Adjustment Operations in Pakistan FY94-03 Project Name FY Project Cost (in US$ million) Public Sector Adjustment Loan (PSAL) 1994 250 Bank Structural Adjustment Loan (BSAL) 1997 250 Structural Adjustment Loan (SAL) 1999 350 Structural Adjustment Credit (SAC) 2001 350 Second Structural Adjustment Credit (SAC II) 2002 500 Sindh Structural Adjustment Credit (Sindh SAC) 2003 100 Northwest Frontier Province Structural Adjustment Credit (NWFP SAC) 2003 90 Note: The BSAL i s discussed in Section 5. T h e Japanese Overseas Economic Cooperation Fund (OECF) co-financed $150 million. T h e IMF later put in a ESAFiEFF for $1.4 billion. This loan had a problematic gestation; five previous appraisal missions had been aborted before the loan was finally appraised and negotiated in 1993. The loan was originally conceived as a low conditionality emergency loan to help Pakistan with the effects o f the Gulf crisis. When the crisis ended, the Bank felt i t could no longer move ahead with a low conditionality loan, which caused lingering problems with the Government. lo T h e Banking Sector Adjustment Loan (BSAL) did, however, go forward in FY97. 10 3.5 About a year after the Musharraf Government took power, the Government reached a new Stand-by Arrangement with the IMF (November 2000), and negotiated another Paris Debt Restructuring in January 2001. The Bank restarted lending in 2001. Macroeconomic Outcomes 3.6 Table 3.2 shows deterioration in macroeconomic indicators at the beginning o f the review period. However, since FYOO, the Government has been making steady progress o n a number o f key issues, including controlling inflation, rebuilding reserves, and restructuring debt. The combination o f improved finances, a good reform agenda, and international support allowed Pakistan to access the private bond markets with a US$500 m i l l i o n five-year Eurobond, at a rate o f 6.75 percent or only 370 bps above the U.S. treasuries. The bond was rated B by Standard & Poor’s and B 2 by Moody’s. Table 3.2: Selected Macroeconomic Indicators (YOof GDP, except where noted) 1993194 1994195 1995196 1996197 1997198 1998199 19991 20001 20011 20021 20031 2000 Ol(a) 02(a) 03(a) 04(a) Total revenue and 17.2 17.0 17.5 16.1 15.8 16.3 16.6 17.3 19.5 20.8 18.3 grants Tax revenue 13.3 13.8 15.0 13.4 13.0 13.3 12.8 12.9 13.2 13.8 13.5 Non-tax revenue 4.0 3.2 2.5 2.7 2.8 3.0 3.7 3.3 4.0 4.1 4.0 Grants 0.1 1.2 2.3 2.8 0.7 Expenditures 23.5 23.7 25.3 22.9 23.5 22.4 22.9 21.4 22.8 22.4 21.6 Budget Balance -6.3 -6.7 -7.8 -6.8 -7.7 -6.1 -6.3 -4.1 -3.0 -1.7 -3.3 Budget balance -6.3 -6.7 -7.8 -6.8 -7.7 -6.1 -6.4 -5.2 -5.2 -4.5 -4.0 (excluding grants) Exports 12.7 13.0 13.1 12.9 13.5 12.9 13.3 15.2 15.4 15.2 15.7 Imports 16.6 16.8 18.9 17.9 16.5 16.5 15.6 17.4 15.9 16.5 16.6 Current Account -3.8 -4.5 -7.6 -6.1 -3.1 -4.1 -3.4 -3.3 0.2 4.6 2.1 (excluding official transfers) Total Public Debt 93.2 86.8 86.3 87.5 89.4 91.9 91.6 108.0 96.7 89.2 84.1 olw External Public 49.3 45.4 42.7 43.9 43.4 46.0 45.7 55.4 48.3 42.0 39.4 Debt Total Reserves (mos. 2.7 2.2 1.7 1.1 1.o 1.7 0.9 1.7 3.7 6.5 6.0 of imports) Inflation 10.0 12.4 12.3 10.4 11.4 6.2 4.1 4.4 2.5 3.1 4.1 (CPI, annual X) Reer exchange rate 97.6 91.4 87.9 94.5 84.4 85.5 85.2 85.3 84.1 79.4 76.1 index (end of year) Source: IMF Country Reports for Pakistan (various years); Global Development Finance. (a) Provisional Actual. The Government has noted that the statistics from the Economic Survey o f Pakistan indicate that reserves are at 11 and 12 months o f imports for 2002103 and 2003/04 respectively; the 6.5 and 6.0 months cited above are from the IMF PGRF Ninth Review. 3.7 However, reforms are still fragile. Although fiscal deficits have been reduced, excluding grants they remain at 4 percent o f GDP. Tax revenue mobilization has stayed 11 flat, and although expenditures as a percent o f GDP have fallen slightly, expenditure reforms have not yet had a major impact. Public enterprises still drain the budget." Assessing the Bank's Contribution 3.8 The Bank's overall advice (increase privatization, simplify tax structures, increase tax bases, streamline Government administration) was sound, consistent, and remained valid over the course o f the review period; over the years these messages contributed to the eventual reforms. AAA, particularly in the last few years, has been o f good quality, and has influenced Government policy. The environment in which policy advice was delivered, however, changed considerably during the period under review. During the 1990s, the instability o f governments and the resulting changes in economic policy made i t difficult to implement any serious reform program. In the last five years, stability o f the government and i t s current policies have allowed a Table 3.3: Remittances (US$ millions) number o f reforms to take root. 2000-01 1,086 200 1-02 2.389 2002-03 4,237 3.9 Pakistan's economic management was also 2003-04 3.872 affected by exogenous factors. The cotton virus and Source: State Bank of Pakistan floods in the early 1990s and the economic sanctions in 1998 had major negative effects on outcomes. In contrast, after September 11, 2001, the country benefited from significant direct donor support (the U S alone contributed US$600 million), the external financing package supported by the IMF for US$1.25 billion over FY02-04, and US$3.5 billion in debt relief from the Paris Club. Remittances also increased (Table 3.3) sharply." Debt service pressures were also eased by the drop in international interest rates. 3.10 There were, however, weaknesses in the Bank's performance. The PSAL/C and the SAL were approved despite a history o f poor reform. In both cases, the approval o f the adjustment loan and overall lending levels signaled that the country was considered to be at the high end o f the base case. While the PSAL/C was strongly supported by an interim Prime Minister, the knowledge that his likely successor would be one o f two candidates, each o f whom had previously served as Prime Minister with a weak reform performance should have given pause to those familiar with Pakistan. Lending levels in FY94-95 were at the high end o f base case levels, which was inconsistent with policy performance. In the case o f the SAL, the accompanying 1998 strategy progress report made it clear that none o f the major conditions for moving to high base case had been " The Government notes that Pakistan Steel, PIA and Pakistan Railways have been profitable for the last 2-3 years and therefore disagrees that public enterprises are a drain on the budget (see Annex 6). However, KESC and WAPDA still require infusions accounting for about 1 percent o f GDP. In addition, although IEG notes that Pakistan Steel, PIA, and Pakistan Railways are recording profits, PIA and Pakistan Railways still depend on the Pakistani Government for financial support. Data from Pakistan Railways show that from 1994195 to 2003/04, subsidies increased from 1.6 billion rupees to 6.6 billion rupees, and capital transfers from the government increased from 1.8 billion rupees to 4.6 billion rupees. W h i l e PIA has not received any subsidies, it has received cancellation o f interest of 2.6 billion rupees and 5.3 billion rupees to replace aircraft, in exchange for equity, increasing the Government's ownership of PIA from 57.7 percent in 2001 to 75.9 percent in 2002. The sustainability o f these flows i s uncertain. Initially, it appeared that the increase in remittances might be temporary, as Pakistanis were concerned that their funds abroad might be frozen or seized. Other explanations include (i) as international money-laundering rules were tightened, more people chose to send money back through the formal financial sector; and (ii) the improved economic reform environment i s attracting more capital. If the latter explanations are the driving forces, the recent increased remittance levels may continue. 12 satisfied. The PSAL and the S A L were also poorly designed and had little sustainable impact o n structural reform, although they did help the country avoid defaultI3 (see Annex 1 for discussion on these adjustment loans). PSAL, SAC, and SAC I1lacked prioritization and selectivity, as their coverage was quite broad, possibly in order to justify the large commitment amounts. Some prior actions were listed in both SAC I and 11, some were continuations o f ongoing programs, and some were being supported by other Bank loans. 3.1 1 Most o f the progress in the adjustment loans was in the areas directly under control o f the federal government, including civil service reform, public accounting and auditing, trade reform, and budget allocation^.'^ Reforms in areas such as power, gas pricing, i.e., where the implementation fell under agencies or ministries not receiving funding from the loan, have not proceeded as well. In a country as large and complex as Pakistan, i t m a y be more appropriate to address sector reforms with more focused adjustment loans, rather than through economy-wide loans. The Bank has started to do this to some extent, making some provincial level one-tranche adjustment loans. While the provincial loans are s t i l l somewhat broad, the one-tranche loans allowed the Bank to limit follow-up lending if the provincial government did not follow-through with continued actions. 4. Poverty Reduction and Social Sector Development 4.1 Government statistics estimated poverty at 26.8 percent o f households in 1992/93 .I5The 1992 country strategy noted that the high poverty levels were due in large part to the limited access o f the poor to productive assets and to inadequate public services. The high population growth rate and the neglect o f human resources development resulted in some o f the poorest social development indicators in the world: infant mortality o f 9 1 per thousand, an adult illiteracy rate o f 65 percent,I6 and high malnutrition among children. Moreover, largely because o f rapid population growth, landlessness increased, with the average size o f farms operated by the poor also declining. The Bank noted that weak social development would affect the country’s long-term growth and human development. The Bank’s Strategy 4.2 Given the needs in poverty reduction and social sector development, it was appropriate that the Bank made improvement in these areas primary objectives. During the review period, the Bank’s country assistance strategies consistently listed Poverty Reduction as a priority. The 1992 strategy discussed a three pronged approach: (i) l3 The experience of these loans raises the question of whether the Bank should provide emergency liquidity to countries to avoid default in the absence o f a strong reform program. T h i s issue goes beyond the scope o f the CAE. l4Progress in areas where there has been extensive Bank support through other loans (e.g., the banking sector), or where the CAS documentationlists accomplishments under ongoing programs that were planned independently of these loans (such as the Lady Health Workers or the establishment o f the Micro-Credit Bank) should not be considered accomplishments o f these projects. l5Government of Pakistan, Poverty Reduction Strategy Paper, December 2003, p. 12. l6 World Bank. 1995. Pakistan Poverty Assessment (Report no. 14397). 13 support for the Government’s Social Action Program ( S A P ) ; ( ii) support for projects in agriculture and industry which expanded opportunities for small farmers and small to micro-sized enterprises, and (iii) “a focus on poverty reduction goals as part o f the Bank’s macro-economic and policy dialogue,” including those factors which would affect the economy’s ability to grow and create new jobs, and fiscal issues which affected the Government’s ability to meet infrastructure and human development needs. The primary vehicles for implementing the first prong o f the strategy were to be the Social Action Program and ESW work on poverty, human resource development, and execution o f an Integrated Household Survey.17 4.3 Over the next few years, the strategy shifted to a focus on service delivery, mostly under the umbrella of the S A P , supplemented by focused service delivery projects (Table 4.1). The 1998 strategy progress report kept the S A P as the centerpiece o f the Bank’s strategy, but also introduced a discussion on targeting programs to help provide social protection, as well as the establishment o f the Pakistan Poverty Alleviation Fund (PPAF; FY99) to help provide micro-credit (especially for women) and small-scale community infrastructure. 4.4 The 2001 Country Assistance Strategy Progress Report stated that the considerable support to S A P had not delivered commensurate results, that outcomes fell “far short o f the SAP’S targets” and that “overall results in this area remain disappointing despite significant investments from the government and the international donor community including IDA-financed interventions. Primary education i s by far the worst- perfoming, but the very poor governance (weak financial management, poorly motivated teachers with lax accountability for results) lack o f leadership and resistance to mobilizing NGO’s help undermined efforts in health, family planning, and community infrastructure as well.” Table 4.1 : Social sector projects, FY94-FY03 Project Project Social Protection & Poverty cost Health & Population cost (in US$ (in US$ million) million) - 1994 Social Action Program 200.0 1995 - Population Welfare Program 65.1 1998 - Social Action Program I1 250.0 1996 - Northern Health 26.7 1999 - Poverty Alleviation Fund 90.0 2003 - Partnership for Polio Eradication 20.0 - 2003 HIV/AIDS Prevention Project 37.1 Education Community Infrastructure 1995 - NWFP Primary Education Project 150.0 1996 - NWFP Community Infrastructure 21.5 1998 - Northern Education 22.8 2003 - AJK Community Infrastructure and 20.0 2003 - National Education Assessment 3.6 Services System Source: Internal Bank data. 4.5 The government’s new strategy was to increase devolution and accountability at the local levels, in order to improve implementation o f service delivery. The Bank supported these initiatives with the Sindh and NWFP adjustment loans, a community infrastructure loan, and smaller, more focused health and education projects. ” Several other projects were in the pipeline, including the NWFP Primary Educationproject (FY95), the Population Welfare Project (FY95) and Northern Health (FY96), but they were not discussed in the strategy. 14 4.6 Gender issues were consistently raised in the country assistance strategies and other documents. The lags in female literacy, school enrollment, and health indicators led the Bank and Government to design projects with a greater gender focus in service delivery. The Bank support o f the Government’s Lady Health Workers program, for example, sought to improve access o f women to health care, particularly in the areas o f family planning, pre-natal care, and birth delivery assistance. Extra incentives were given to encourage girls to attend school, such as providing scholarships and free textbooks to schoolgirls. Extra resources were also devoted to improving girls’ school facilities and recruiting and training more female teachers. Poverty Reduction a n d Social Sector Outcomes 4.7 The Bank, as well as many knowledgeable Pakistanis, refer to the 1990s as the “lost decade” for the poor, as little progress was made in the Bank’s objectives o f reducing poverty and improving service delivery. The 2001 household survey (the latest available data) indicates that the percent o f Pakistani’s households below the poverty line increased from 26.8 percent in 1992/93 to 32.1 percent in 2000/01. Although strong growth since then may have caused a recent decrease in poverty, preliminary data from a 2004 household survey suggests that poverty i s s t i l l higher than it was at the beginning o f the review period.” The Government also has noted that quality o f services i s a problem, stating, “The ability o f the state to effectively deliver quality services to the citizens i s very limited. For example, rural hospitals and dispensaries lack staff and facilities and do not have effective systems o f supervision over the dispensation o f publicly provided medicines, The hygienic conditions o f even some o f the best hospitals are inadequate. The staff i s poorly motivated and badly managed. Neglect and malpractices are commonplace .. .Conditions in education, social welfare, environmental conservation, and population welfare programs are broadly ~ i m i l a r . ” ’ ~ 4.8 Although there has been marked progress in important areas such as childhood immunization, fertility, infant mortality, and female primary enrollment rates, the 2004 CAS PR noted that “Human Development indicators continue to lag behind those o f countries with similar per capita incomes and despite encouraging progress in some areas, more rapid progress will be needed to achieve the MDGs (Millenium Development Goals) by 2015.” (See Table 4.2 and Annex Tables 8-10.) Pakistan ranks 138th out o f 177 countries on the UNDP’s Human Development Index (HDI).20 Infant mortality rate of 76 per thousand i s well above Bangladesh’s rate o f 48 per thousand, and more than four times greater than Sri Lanka’s rate o f 16 per thousand. Its total fertility rate (as o f 2002) o f 4.5 i s substantially higher than the rates o f India and Bangladesh (2.9 and 2.95 respectively). Girls between ages 1 and 4 had a 66 percent higher death rate than boys in the 199Osy2’contributing to Pakistan having a lower female proportion o f population than See Government comments on this in Annex 6. World Bank. 2001. Pakistan: Joint StaffAssessment o f The Interim Poverty Reduction Strategy Paper, pg 35 (Report no. 23 189-PAK). 2o United Nations Development Program. 2004. Human Development Report (HDR), Oxford University Press, New York, reports a ranking of 142 out o f 177, based on combined measures o f l i f e expectancy, school enrollment, literacy, and income. However, the Pakistani Government has received a recent addendum to the HDR which upgraded the ranking o f Pakistan to 138 out o f 177. *’ Easterly cites Tinker, 1998. 15 Female school enrollment and literacy continue to lag men. There are also differences between rural and urban areas, and between provinces.23 While urban poverty fell between 1992/93 and 2002/03 (28.3 percent to 22.4 percent), rural poverty increased from 24.6 percent to 38.7 percent, a substantial widening o f the urban-rural gap. Table 4.2: Selected So( 11 Indicators, 1993-2002 Nepal Sri Lanka South Asia 1993 2002 1993 2002 1993 2002 Immunization (% of 51 72 90 98 58 70 children 12-23 months) Fertility rate (births per 5.3 4.2 2.5 2.1 NA NA woman) Infant mortality rate (per 81 62 19 16 78 68 1,000 live births) School enrollment, female 88 113 NA NA 83 88 primary (% of gross)’ School enrollment, male 128 130 NA NA 106 103 primary (%) of gross)‘ Female Literacy (age 15 17 26 86 90 36 44 + )* Male Literacy (age 15+)* 52 53 1 64 68 I 46 50 51 62 93 95 66 73 Source: WDI, Government o f Pakistan Poverty Reduction Strategy Paper. ’ L a s t available data for Bangladesh, Nepal, and Pakistan (PRSP) i s 2001; India 2000. Last available data for India i s 2000; Pakistan, 1998. Note that the Pakistan PRSP cites more recent figures o f 37 percent female literacy and 61 percent male literacy based on recent surveys, but does not define the age range. Hence, for comparative purposes, the W D I data i s used. 4.9 In addition to poor social outcomes, the Bank’s program failed to attain its primary intermediate output: an increase in Government expenditures for the social sectors. By 2002, the Government’s contribution to S A P as a percent o f GDP had actually dropped (see Table 4.3).24 Table 4.3: Government Allocations and Expenditures o n S A P (Percent of GDP) SAP Allocations and Expenditures (as % of GDP) 1993194 1994/95 1995196 1996197 1997198 1998/99 1999/00 2000/01 2001/02 Allocations 1.6 1.7 1.8 1.8 2.3 2.3 2.1 1.9 1.8 Expenditure 1.6 1.6 1.7 1.6 1.7 1.6 1.4 1.4 1.5 Source: The World Bank, Social Action Program ICRs and supervision reports (various years). 4.10 Several elements of the non-SAP health and education projects have had positive outcomes, including teacher training, family planning and immunization, although a strategy for scaling up these improvements i s s t i l l needed. In addition, the Pakistan 22 Easterly, 2003, Table 4. 23 A recent National Human Development Report documented marked and persistent differences in the Human Development Index (HDI) across provinces as well. See UNDP, Pakistan National Human Development Report 2003 (UNDP: Islamabad, 2003). These differences are also noted in the World Bank’s 2002 Poverty Assessment (Report no. 24296-PAK). 24 Bank staff commented that, given the increasingly difficult fiscal situation, i t was a positive achievement that the Government maintained its social sector spending at roughly comparable levels. I t i s likely that without Bank pressure, spending levels would have decreased further. Also, the existence o f the SAP kept the importance o f social spending in the forefront o f major policy discussions. 16 Poverty Alleviation Fund (PPAF) is doing well, having extended loans to 122,000 clients with an estimated recovery rate o f about 95 percent; the results were encouraging enough that a second PPAF o f US$238 million was approved by the Board in 2003. Assessing the Bank’s Contribution 4.1 1 Although the environment for improving social services was hampered by the extremely limited fiscal space to increase expenditures aimed at the poor, the Bank’s strategy for poverty reduction and social sector development had its o w n weaknesses as well. First, the strategy was focused on service delivery, rather than o n the root causes o f poverty reduction and on promoting income generating activities for the poor, particularly in rural areas. Outside o f the social sectors, there was limited attention paid in the lending program to the poverty impact o f Bank interventions. Bank support for rural development, for example, was focused on irrigation and drainage, to the neglect o f rainfed areas and livestock; within irrigation, little attention was paid to the access o f the poor to water supplies, although Bank analytic work had identified land inequality and therefore unequal access to irrigation as a k e y issue affecting income inequality. As another example, Bank support in infrastructure was concentrated o n large power projects, with no support for rural electrification during the period under review.25 4.12 Second, although the Bank had identified over the years and in a number o f analytic and strategic documents political economy issues and power relations that were relevant to development efforts,26these were largely ignored in lending or in project design. Thus, issues o f female access to labor and factor markets and land inequality, for example, were not taken into account in the design o f Bank programs. 4.13 Third, within service delivery, the Bank’s lending focus was mostly on S A P , and within S A P , on expanding the quantity rather than the quality o f the services. Even as S A P wound down, most o f the new social sector projects also focused o n service delivery. Research in Pakistan had found that the demand for private education was growing, for example, indicating a willingness to pay for perceived quality o f services. Nevertheless, S A P aimed at increasing expenditures o n social services, with little attention to outputs or impact o f increased spending. In addition, S A P suffered from a number o f major problems: complex project design, and large numbers o f government units and donors involved, which in turn meant large administrative costs to appraise and supervise, and inadequate fiduciary safeguards on the use o f finds. Finally, i t i s not clear whether sufficient Bank effort was put into the critical buy-in to the S A P by provincial 25 A Rural Electrification Project was approved in FY90, but this delivered little in terms o f its original targets for rural electrification or controls on tubewells. 26 The 1995 Pakistan Poverty Assessment (Report no. 4397-PAK) noted, for example, “Women in Pakistan have been discriminated against in seeking access to labor and credit markets, and to such government services as education and agricultural extension (pp. x-xi)” and “The evidence suggests that there are gaping inequalities in ownership o f land across gender, and the law on inheritance.. .has failed at the implementation level in Pakistan.” (p.14), and the 2002 Poverty Assessment (Report no. 24296-PAK) found that, “The deepest and most pervasive poverty in the country i s rural and it i s worst in areas that have traditionally been considered ‘feudal’. . ..Rural elites have exceptional influence in Pakistan.. .theyhave had relatively little interest in enhancing their constituents’ access to education or insuring that the poor could obtain the protection o f the l a w without elite intervention.” 17 ministers, and thus the program was not owned by the sub-national governments who were expected to implement its2’ 4.14 Beyond problems in the overall strategic thrust o f the Bank’s program, the Bank failed to design a program that was realistic. The S A P projects were the most extreme examples o f this failure, but the problems can be seen in other projects. Among the design issues were: (i) inadequate mechanisms to measure intermediate results or the effectiveness of spending; (ii) overly complex project design. For example, although implementation completion reports stressed the need for focused projects with a limited geographical and administrative span, SAP involved twenty-seven government units and numerous donors with missions teams o f more than twenty members; and ( iii) financial management requirements which did not reflect the capabilities o f the counterparts.28 Weaknesses in financial and project management were common. Audit reports by the Auditor General (AG) contained a large number o f observations which pointed to major irregularities. 4.15 Supervision reports and ICRs suggest that Bank staff and the other donors were aware o f implementation problems, but had difficulty addressing them. For example, following o n the experiences o f SAP-I, the second phase o f SAP laid a greater emphasis o n issues o f governance, quality, and community/NGO participation, but there was s t i l l an absence o f indicators defining acceptable performance. 4.16 Delivery o f the program was also quite expensive. SAP I cost U S $ l .l million to prepare, and US$1.2 million to supervise. S A P I1cost US$1.2 m i l l i o n to prepare, and US$2.2 m i l l i o n to supervise. In addition, the Bank spent another US$1.3 million o n other SAP associated costs (dialogue and other related tasks). Altogether, delivery o f the S A P program alone cost close to US$7 million.29 Lessons and Recommendations 4.17 Overall strategy. Government and Bank policy papers all recognize the underlying issues that affect poverty, such as access to land, credit and other inputs; ability to improve livelihoods, need for social protection, etc. Yet, these analyses have not worked their way into the lending program, with the exception o f the PPAFs. Poverty issues should translate into the selection o f projects and into project design. More effort must be made to have staff from across different sectors develop a holistic approach to lending for poverty alleviation. 4.18 Institutional capacity. Projects need to have a stronger emphasis o n building institutional capacity, not just in a project management unit, but in the government and community at large. There should be greater emphasis o n technical assistance (TA) in ’’IEG will be carrying out a PPAR on SAP Iand I1 in the near future. *’The ICR for SAP I1noted, “the manual tracking and documenting of several million manual transactions to IDA standards was humanly impossible.” Supervision reports and other documents noted problems with the submission o f SOEs. 29 Bank budget only. Excludes trust funds or other donor costs. Average preparation costs for projects in Pakistan from FY94-03 were US$356K;averages for S A R and the Bank were US$433Kand US$357K,respectively. 18 the program, and more realism about the start-up time for projects as local capacity must be developed. Projects should also be scaled back as needed to meet the counterparts’ capabilities. 4.19 Political economy and governance. While the Bank i s limited in i t s ability to break through bottlenecks caused by vested interests, Bank project must take these issues into account. This could mean focusing more on aligning incentives, and improving monitoring and evaluation systems. 5. Sustainable Growth 5.1 Inthe early 199Os, the state controlled most infrastructure, a number o f large industrial companies, and most o f the financial sector. Foreign Direct Investment (FDI) was less than 1 percent o f GDP, and gross fixed investment (both public and private) was about 17 percent o f GDP. 5.2 Pakistan had a limited production and export base, with agriculture contributing 26 percent o f GDP, but employing h a l f the labor force and providing 70 percent o f exports, including agriculture-based manufactured goods. Industry was concentrated in cotton processing, textiles, petroleum refining and food processing, and suffered from poor product quality, outdated technology, and an untrained labor force. 5.3 In infrastructure, power outages were increasingly common, telecom density was less than 1 per 100 people (one o f the lowest in the world), and the inadequate transport system (roads, railways, ports) delivered poor service levels at high costs. The Bank’s Strategy 5.4 The Bank’s overall strategy was to support a shift from public sector ownership and management to the private sector. This included trade reform, expanding and modernizing the financial sector, accelerating the Government’s privatization program, improving inadequate infrastructure and a poorly trained workforce, industrial deregulation, deregulation o f administered prices, and a more flexible exchange rate. Law-and-order issues were also highlighted as an impediment for attracting new investment. 5.5 The 1992 country strategy identified specific areas to support growth: (i) diversifying and expanding the productive agricultural and industrial base; (ii) improving the management o f the country’s natural resources, particularly water and domestic energy resources, with increased attention to environmental concerns; and ( iii) overcoming severe infrastructure bottlenecks, as manifested by energy shortages and inadequate transportation infrastructure. Over the course o f the review period, this translated into programs which focused on: (i) energy and infrastructure (18 percent o f total lending), with an emphasis o n power generation; (ii) finance (17 percent); and ( iii) agriculture and natural resources management (9 percent), with an emphasis o n irrigation and drainage. In addition, while there were no specific loans for issues such as trade, 19 privatization, and investment climate, reforms were supported through AAA and conditionality in structural adjustment loans. This section discusses each o f these areas below. Energy and Infrastructure 5.6 The Bank’s infrastructure strategy, in Pakistan and worldwide, shifted away from Government-owned and operated infrastructure, towards greater involvement o f the private sector. The Bank concentrated i t s initial efforts in Pakistan o n the power sector, supporting an expansion o f the role o f the private sector (through the use o f guarantees and Other products), Table 5.1 : Infrastructure operations FY94-03 financing public sector investments, and promoting FY Project Amount in US$ million improvements in the 1994 Power Sector Development Loan 230.0 Operating efficiency of public 1994 Sindh Special Development Credit 46.8 1995 Second Private Sector Energy Development Project 250.0 enterprises. However, 1995 Uch Power Guarantee 75.0 investment lending for 1995 Hub Power Guarantee 240.0 1996 Ghazi Barotha Hydro Loan 350.0 energy disappeared in the 1996 Telecommunications Regulatory and Privatization 35.0 late 1990’s as a result of poor 2000 Policy Reforms in the Petroleum Sector IDF Grant 0.50 2001 Trade and Transport Credit 3.0 performance in power sector Total 1230.0 restructuring, poor progress ~~ in policy reform in other infrastructure subsectors, and the need to focus o n other borrowing priorities. The Bank did maintain an active dialogue, supplemented with AAA and targeted TA. Infrastructure Outcomes 5.7 Power Sector. The most notable progress in the power sector has been in terms o f physical expansion. M o r e than 5,000 MW o f new private power generation capacity was ~ ~ one-third o f current capacity, which helped ease the chronic installed in the 1 9 9 0 over shortages in generating capacity which Pakistan had experienced in the preceding decade.30 One generation company was privatized, the transmission system was upgraded to service the new generation capacity, and chronic supply shortages were reduced. The sale o f 73 percent o f Karachi Electric Supply Company (KESC) was recently announced. On the other hand, the price o f new generation capacity developed under the private power program was high3’,serious allegations o f corruption were raised related to awarding and negotiating contracts with private participants (Box 5. l), and the process o f restructuring and privatization has been slow. Finally, the sector’s financial Performance has worsened dramatically, so that in recent years losses at the Water and 30 In addition to private power, the Bank also supported the Government run Ghazi Barotha hydropower project, which had a low generation cost o f US$l.O7 cents per kwh. 3 1 The cost per K W for the Hub, Uch, Rousch and Southem Electric power projects funded under the two power sector development loans ranged from US$1,205 to US$1,395. Price comparisons must be interpreted carefully because world generation prices fell as technology and competition advanced. However, although it may be that the cost per K W was reasonable for the IPP projects at that time, the failure to move to competitive bidding at an earlier time and to phase in the new contracts meant that the system had excess capacity for a number of years, and could not take advantage of the lower prices that accompanied technology advances and optimized fuel mix. 20 Power Development Authority (WAPDA) and KESC have been some $500 million per year, amounting to 1 percent o f GDP. Box 5.1: The Independent Private Power Program: Issues’ The Bank approved a series o f loans and guarantees to assist Palustan in attracting private sector resources. In 1995, afier more than six years, the 1292 MW, US$1.6 b i l l i o n Hub Power Project reached financial closure. I t was hailed as a landmark in infrastructure financing, and a model for the viability o f private financing in infrastructure in a developing country. The Government subsequently issued Letters o f Support for a n additional 34 projects totaling 9,000 MW, o f w h i c h 19 IPPs, representing 3400 M W o f additional capacity, reached financial closure. Four o f those projects were n o t completed. The Bank provided partial risk guarantees for two projects, the IFC provided loans a n d o r equity for five projects, MIGA extended guarantee coverage t o three projects, and the Government o f Pakistan provided subordinated loans t o six projects through the IBRD-financed Private Sector Energy Development Projects. See Annex 2, Table 16 for details. In 1994, p r i o r t o tariff agreements reached under the IPPs, W A P D A ’ s average tariff charged t o the consumer was 4.5 cents per kWh. However, in order t o attract potential investors, a n indicative t a r i f f o f 6.5 cents per kWh was set for power sold t o W A P D A ; this increased WAPDA’s costs considerably. In addition, the power sector suffered f r o m distribution and system losses, and uncollected customer payments. The rupee depreciated by 45 percent f r o m 1994 t o 1998, when most o f the power came online, making it yet more difficult to pass o n increased costs to consumers. In the end, W A P D A and K E S C experienced heavy financial losses, requiring Government support. Given the high level o f investor interest, a competitive bidding system should have been set up, which n a lower tariff, Furthermore, the B a n k had projected that, after Hub, only 2000 l i k e l y w o u l d have resulted i additional MW were needed at the time, and yet supported projects with total capacity in excess o f that projection. Over-capacity was compounded by weaker-than-projected demand for power because economic g r o w t h weakened. Widening cross-subsidies f r o m industrial and commercial consumers to other consumers l e d m a n y o f the industrial and commercial consumers t o install their own captive capacity, m h e r weakening effective demand for power f r o m WAPDA. Finally, transmission and distribution capacity failed t o keep pace with the expansion o f power generation, so that even if demand had materialized, the system was n o t able to deliver power to the end users. In 1998, the Government moved t o cancel seven IPP contracts o n grounds o f corruption and two on technical grounds. This represented about two-thirds o f the private power capacity contracted. Several government committees were established t o renegotiate tariffs and investigate corruption charges (allegations were also made against Bank staff). After a great deal o f frustration o n a l l sides, several IPPs agreed to tariff reductions, and, in at least one case, a court-ordered reduction in tariffs resulted. W h i l e nc accusations were ever proven in court, these allegations and court cases made the IPP program highlq politicized, resulted in a loss o f investor confidence, and severely damaged perceptions o f Pakistan’: investment climate. ’ T h i s section draws heavily on World Bank, Energy and Mining Sector Board Discussion Paper No. 14 “Lessons from thc Independent Private Power Experience in Pakistan,” May 2005. 5.8 Oil and Gas. Pakistan has limited known reserves o f crude oil, but has exploitable gas reserves o f about 27 TCF (equivalent to about 25 years o f production at current levels). Although the government’s stated policies were to emphasize private sector participation, through most o f the decade the government tightly controlled the o i l and gas industries. Starting in 1998, with support from the Bank, the Government developed a new policy framework. Policy capacity was built up in the Ministry o f Petroleum and Natural Resources. Fuel o i l and LPG markets were liberalized in 2000. A formula-based fortnightly adjustment o f petroleum product prices, a new consumer gas 21 price framework, and a new wellhead gas price framework were introduced in 2001-02. These reforms have helped catalyze more than US$1.3 billion in investment in the sector and a 50 percent increase in gas production in the last three years. Distortions in pricing remain, however, as shown by the huge discrepancies in gas pricing, with the fertilizer industry paying 37 Rs. per thousand cubic feet (mcg, and commercial customers paying 205 Rs per m c f (see Annex Table 14). 5.9 Telecom. Although five mobile telecom licensees have increased competition and access, physical line density still remains l o w at 2.84 lines per 100 inhabitants, and remains a constraint on economic development. However, regulation has improved, as the Pakistan Telecommunications Authority (PTA) has become more independent. The spectrum allocatiodsite clearance process has been reduced from over four months to seven days. Installation o f a national frequency management and monitoring system has improved those functions considerably. Pakistan Telecommunications Corporation Ltd (PTCL) has lost i t s monopoly status, and bids have been invited for additional fixed l i n e providers, The government has twice tried to sell a 26 percent stake in PTCL, but has not yet succeeded, which may reflect systemic problems such as overstaffing, restrictive labor practices, and regulatory and policy constraints. 5.10 Transport. Pakistan’s transport sector has been characterized by poorly targeted investments, neglect o f essential maintenance, inefficient labor and non-commercial practices resulting in severe bottlenecks, high transport costs, poor safety standards, and l o w levels o f service. In the roads sub-sector, maintenance spending covered less than 15 percent o f stable network needs; meanwhile, major public expenditures were made to support expensive highway projects with no obvious economic return. Deferred and inadequate road maintenance leads to Rs 16 billion o f road assets being lost each year. Recently, however, there has been a movement in policy reform. An integrated transport policy has been drafted, the National Highway Authority has been strengthened and downsized, the priority o f maintenance over new construction has been re-established and major progress has been made in mobilizing resources for sustainable maintenance. Assessing the Bank’s contribution 5.11 The Bank’s basic strategy o f shifting the burden o f providing infrastructure from public to private investment was reasonable. While there were many design and implementation issues (discussed below), global experience with private infrastructure provision was very limited in the early 1990’s and much has been learned in the intervening period. Pakistan pioneered the Bank’s new energy policies (enunciated in 1992) which emphasized unbundling, competition and privatization, but, as a result, had to face the consequences o f adopting an untested set o f reforms. Indeed, the early “private energy” experience in Pakistan helped catalyze this change in Bank policy. The quality at entry o f the Bank’s interventions was judged superior at the time and both the strategy and the individual products designed to deliver it were regarded as highly innovative and responsive to the problems facing the country’s energy sector. 5.12 The Bank also displayed considerable flexibility in using a variety o f lending, guarantee and non-lending instruments as w e l l as conditionality in implementing the 22 strategy. When progress in restructuring stalled in the mid 1990’s (power and ports) or investment priorities became distorted (highways), the Bank appropriately refrained from investment lending and maintained the dialogue through adjustment lending (power sector), supervision (power, ports, highways) and AAA (oil and gas, power). While formal sector work was limited, significant advisory services were delivered via supervision and project preparatiodappraisal. 5.13 Despite the overall coherence o f the Bank’s approach, a number o f problems emerged in the implementation o f the infrastructure agenda, most notably in the power sector. Some of these were specific to the situation in Pakistan, while others related to the teething issues and problems associated with implementing the Bank’s global infrastructure initiatives which were enunciated in the early 1990’s. While such issues are easy to see with hindsight, some could have been avoided, and are discussed below, Lessons 5.14 Inadequate attention to thefinancial aspects o f reform. The problems o f the inadequacy o f the level and structure o f power tariffs, theft and collection (particularly in key provinces and cities) were well known; project documents consistently highlighted the importance o f continued tariff reform as well as steps to improve collection performance. The financial crisis that followed the commissioning o f the IPP projects was the result o f insufficient attention to this issue. Pakistan faces immense social and political constraints in tackling these problems; a more direct focus o n the tariff, cross- subsidies, the level and nature o f the losses and the sector’s underlying financial viability, coupled with innovative prescriptions designed to help the government tackle these problems should be pursued. 5.15 Failure to design projects that reflect the political economy and governance climate of the client. Given concerns about governance, a transparent competitive bidding system should have been used for the IPPs. The first project (Hub) was an unsolicited offer, and the prices for the second IPP round were set before investment interest was gauged. The Bank then recommended a switch to competitive pricing, but this did not happen. Given the subsequent rapid reduction in the global costs o f generation which occurred in the mid-1990s (as a consequence o f both technological advances and increased competition), and given the significant opportunities for corruption in the awards of such contracts, particularly in a weak governance environment, such transparency was particularly important. 5.16 Overemphasis on newproducts. While the Bank’s approach was innovative, a more measured and pilot type approach would have been more appropriate. Specifically, fewer contracts could have been awarded in the second phase, and then subsequent contracts could have followed a competitive procurement process. A lower cost and more optimal set o f generation investments which were better synchronized with the restructuring of the gas sector might have resulted. The availability o f gas, associated with the restructuring and price reform, would have allowed the installation o f more efficient combined cycle generation in the second IPP round. 23 5.17 Bank Management must better supervise new approaches. Bank management failed to recognize the considerable risks associated with the IPP program in Pakistan, particularly the excessive fiscal and external account risks placed on the country as a result o f the guarantees extended by the government to the private power producers. Bank management also failed to recognize the credit risk o f the Bank itself, in the event of a default by the government on i t s guarantees (which almost happened). These failures m a y have resulted in part from the decision to have staff working on the IPP transactions bypass normal regional management process and report directly to senior management. Establishing specialist groups to drive a new agenda item i s a tried and tested management technique; however, at the Bank, such an approach requires close super~ision.~~ 5.18 Government commitment topartial reform is not enough. There was demonstrable government support and “buy-in” for private generation transactions but no appreciation of the need for and the implications o f the broader reform agenda. While successive governments committed themselves to the reform process and took some important actions to implement the agenda, key officials remained uncommitted. For example, deeply entrenched vested interests successfully delayed the restructuring o f WAPDA, which was central to the ultimate success o f the unbundling and privatization process. I t i s now completed in form, but not yet in substance. Furthermore, after years of dialogue, the overall structure and level o f tariffs are still inconsistent with a viable and competitive energy sector. The FY90 rural electrification project failed primarily because o f inconsistencies in the tariff coupled with the unwillingness to pay for losses. M a n y o f the new distribution companies (public or private) will not be solvent unless there i s significant progress o n tariff reform and collection. 5.19 In contrast, the Bank’s refusal to lend for highways for most o f the period being reviewed by the CAE i s a good example o f the value o f insisting o n substantive policy reform. Lending only resumed in late 2003 when the government had clearly demonstrated a reordering o f priorities away from large road projects to highway maintenance. Similarly, while the Bank refrained from making any new investment loans to the o i l and gas sector during the last decade, it extended TA and AAA support when a clear commitment to reform was exhibited by i t s government counterparts. 5.20 Building government policy capacity should be a focus for Bank Assistance. Despite the existence o f high level professionals in Pakistan, the policy capacity o f the infrastructure ministries i s weak. For example, the Water and Power Secretariat has limited capacity to analyze issues facing the power sector and to develop relevant policy options, and instead relies on AAA from the World Bank and WAPDA, an operational entity with vested interests. 5.21 The effectiveness o f ESW. While important policy dialogues were carried out through project preparation, it might have been efficient to have spent resources on ESW 32 T h i s point i s also made in “World Bank. 2003. Power for Development: A Review of the World Bank Group’s Experience with Private Participation in the Electricity Sector,” (Report no. 28042) a j o i n t review by IEG-WB, IEG- IFC and IEG-MIGA. 24 and other policy dialogue instruments. The recent energy strategy and water strategy papers have been well-received by the Pakistani government, showing that policy dialogue can be effective, even without lending. Finance 5.22 At the beginning o f the 1990s the financial sector was repressed: (i) the government controlled interest rates; ( ii) the State Bank o f Pakistan (SBP) regulated tightly the allocation o f resources, giving priority to the financing o f the government and i t s directed credit programs; (iii) the government owned most o f the financial system; and (iv) the Government controlled the sale and deposit o f foreign exchange. In addition, all the risks o f the financial system were centralized in the government as it effectively guaranteed most financial obligations, directly through the political decisions that led to lending and indirectly through i t s ownership o f financial institutions. In this environment, financial institutions and their regulators and supervisors were not conscious o f the risks involved in financial operations and failed to develop the skills necessary to manage them. Regulations regarding the quality o f the portfolios were extremely lenient and supervision formalistic. Banks reported loans as performing that should have been written o f f and kept accruing interest on them, obscuring their true financial situation. The system’s deficiencies in terms o f risk assessment and management became obvious during the 1990s, when arrears escalated in both the Nationalized Commercial Banks (NCBs) and Development Finance Institutions (DFIs) to the point o f rendering them technically bankrupt. The Bank’s strategy 5.23 An FY93 financial sector report proposed a strategy based o n three pillars: continued reform o f macroeconomic and financial policies; institutional strengthening; and increasing access to resources to permit new and reformed institutions to meet the needs o f their Table 5.2: Financial Sector Operations in Pakistan FY94-03 Project Name FY Project Cost clients. Specific (in US$ million) measures included Financial Sector Deepening and Intermediation Project (FSDI) 1994 $216 the strengthening of Bank Structural Adjustment Loan (BSAL) 1997 $250 Banking Sector Restructuring and Privatization Project (BRSP) 2001 $300 SBP’s regulatory Banking Sector Technical Assistance Loan (BSTAL) 2002 $26.5 and supervisory functions and the acceleration o f the privatization o f the NCBs and DFIs. The strategy was supported by four operations (Table 5.2). 5.24 Despite the analytical work, and the poor performance o f a series o f previous financial intermediation loans (FILS), the first loan after the financial sector report was a FIL with a small TA component. The US$200 m i l l i o n l i n e o f credit was undisbursed and cancelled, while the small TA component was implemented. The loan was unsatisfactorily designed since the financial system was not ready to absorb productively an intermediation loan, but was considered satisfactory because the TA component achieved institution building, which was one o f the k e y loan objectives. 25 5.25 In early 1997, the SBP and the Ministry o f Finance put together a reform program, and approached the Bank for support. The new strategy was to prepare the financial institutions for privatization by installing independent managers, under new legislation that protected the process from political interference. These new teams were required to implement action programs, approved by SBP, aimed at reversing the flow o f losses and downsizing the banks. The Bank supported the project with the Bank Structural Adjustment Loan (BSAL) and followed with the Banking Sector Restructuring and Privatization Project (BSRP). Both were fast-disbursing loans designed to support the costs o f restructuring the banks, including the cost eliminating redundancies, which eventually separated 29 percent o f the staff at the institutions, at a cost o f US$350 million. Financial Sector Outcomes 5.26 The Pakistani financial system has improved significantly. Among the positive achievements are: (i) losses in the banking system have been stemmed; ( ii) five NCBs have been restructured and capitalized; o f which four have been partially privatized and the profitability and efficiency o f these banks have improved in the process; (iii) the SBP has become a modem central bank, focused on monetary policy and the regulation and supervision of the banking system, and dropping non-core functions; (iv) Bank regulation and supervision improved markedly; (v) directed credit schemes have been substantially reduced; and (vi) commercial banks are now free to allocate their resources in accordance with market signals. In addition, the capital markets have become much more buoyant; market capitalization o f publicly traded companies has increased from Rs289 billion in 1998-99 to Rs1.4 trillion (US$24 billion) in 2003-04. Figure 5.1: M2 and credit in Pakistan (% o f GDP) 60% I I - 50% 40% Comesic Credit Claims on General Go*. (Mt) 30% Claims on Private Sector 20% 10% 0% I Source: International Financial Statistics. 5.27 There are still, however, many things to do. Credit to the private sector as a percent o f GDP has not changed substantially in the last decade (see Figure 5.1). Privatization i s not yet complete, and the privatization o f the most important o f all the NCBs, the National Bank o f Pakistan (NBP), will not occur for several years. 26 Regulation and supervision, while improved, still need to be applied in a more consistent way, particularly in the non-bank financial institutions. Substantial losses s t i l l need to be absorbed and operational costs reduced. With non-provisioned non-performing loans around 10 percent o f the claims on the private sector, the banks s t i l l have a long way to go to clean their balance sheets. Housing credit remains underdeveloped and delinquent loans in that sector are too high. The new legal collection system i s much more efficient than the one it replaced, but the backlog o f collections i s s t i l l substantial. It i s the direction o f change; however, more than the attained objectives, that marks the difference between 1993 and 2003. While there i s s t i l l a long way to go, a critical mass o f reforms has been achieved, making reversion to the old state-managed system unlikely. Table 5.3 : Selected Banking Sector Statistics 1993 ’ 2004 Total Deposits of the Banking System (Rs Billions) 620.4 2143 NPLs as O h of Gross loans (All Banp) 19.0% 14.5% Capital as % of Assets (All Banks) 2.8% 5.1% ROE (All Banks) 19.4% 18.3 % ROA (All Banks) 0.6% 1.O% Average interest rate spread 4.8 % 4.1 % Sources: State Bank o f Pakistan; IMF; World Bank. ’ Deposit data in f i r s t column are for 1992; 2004 data are as of June, 2004, except deposit data, which are year-end 2004. Regulatory Capital to Risk Weighted Assets would be the better measure. However, risk weighted assets were not properly tracked in earlier years. The SBP reports an average Regulatory Capital to Risk Weighted Assets of 9.8 percent as of June 2004. Note that interpreting NPLs, ROA, and ROE i s difficult, since loan provisioning was not consistently applied. Assessing the Bank’s Contribution 5.28 In IEG interviews, government officials and private sector participants expressed the view that the participation o f the Bank was crucial in the successful reform o f the financial system. The strategy was appropriate and, while i t evolved over time to respond to new developments, i t remained essentially the same throughout the period o f review. I t correctly identified the key dimensions o f reforms that were required to transform the financial system into a more efficient mechanism o f development. 5.29 With the exception o f the FSDI, the Bank loans were well-designed. The BSAL and BSRP provided critical funding and technical assistance to a solid program o f bank restructuring. The loans also strengthened the asset recovery departments o f NCBs, loan recovery procedures, and DFIs. These measures l e d to a recovery o f 25 percent o f the loans within one year o f launching the programs. 5.30 The loans also gave attention to longer run problems, including strengthening regulations and banking supervision-with measures such as increasing and enforcing capital requirements, stopping income accrual o n delinquent loans beyond 90 days, tightening regulations o n loan concentration, and improving bank examination and disclosure-as well as promoting the integration o f the financial markets and improving legal and judicial enforcement o f loan contracts. The recent B S T A L i s too new to assess. 27 5.3 1 The program was effective at the end o f the decade, even if i t was a failure during the early part o f the period, mainly due to the increased government ownership o f the strategy, which became strong only after 1997. Lessons 5.32 Adjustment lending should support government ownership and reform. One reason why the B S A L performed better than the other adjustment loans i s that the reforms were truly “home-grown.” The program was designed by the State Bank o f Pakistan and Ministry o f Finance. The counterparts had a clear vision o f what they wanted to achieve and the costs o f implementing those reforms, and looked to the Bank and other donors to help them with financial and technical assistance. This stands in sharp contrast to the PSAL and SAL, which were quickly put together to create a justification to provide needed balance o f payment support; not surprisingly, as soon the funds were disbursed, commitment waned and reform disappeared. 5.33 Develop an overall strategv, then focus on achievable steps. The Government and the Bank had an overall strategy for the sector, but focused initially o n the banking sector; within banking, focused primarily on the state-owned banks; and within the state- owned banks, focused on a specific set o f restructuring actions for the initial loans. While there i s s t i l l a large agenda for reform, these initial reforms are likely to be deeper and more sustainable than if an overly broad reform agenda had been immediately embarked upon. 5.34 Institutional development is important at all levels to support meaningful and sustainable changes. The counterparts and Bank staff undertook a program which not only built capacity at the Central Bank, but which also instituted reforms at the large commercial banks in the program. This included not only recruiting new top management, but also attracting strong independent Board members, and reducing staff with a buy-out program. The counterparts understood that institution building needed to cover more than a small cadre o f staff in a ministry or implementation unit. The Bank and other donors followed up with TA hnding to support these changes. Agriculture and Natural Resource Management (NRM) 5.35 Agriculture accounts for h a l f the labor force and one-quarter o f GDP. However, productivity o f Pakistan’s agriculture i s very low. The prevailing arid and semi-arid climate in most o f the country has meant that water resources available for irrigation and the expansion o f irrigated land are critical to increases in productivity. Agriculture uses an estimated 95 percent o f available water resources, and irrigated land produces 80 percent o f agricultural production. The total irrigated area in Pakistan i s about 46 million acres, or about 82 percent o f irrigable area. Crop production accounts for 60 percent and livestock about 38 percent o f agricultural GDP. 28 The Bank’s Strategy 5.36 The Bank’s FY94 strategy outlined the Bank’s objectives in agriculture. “In agriculture, the focus o f our policy dialogue i s o n reshaping public and private sector roles. For the public sector, the main objectives are to improve capacity to efficiently carry out the large investment programs in irrigation and drainage, strengthen support services in research and extension, and improve management o f watersheds and water resources.. .. With respect to the private sector, recently completed economic and sector work on irrigation and drainage recommends a fundamental shift in the management o f the irrigation and drainage system to reduce the need for public financing and give the private sector a greater stake in the efficient management o f the system.” The FY95 strategy and FY97 strategy progress report echo similar approaches, except that Environment, which had been listed as a key area o f reform in the 1992 and 1994 strategy documents, was by the 1995 strategy subsumed along with agriculture, as part o f the growth reform agenda. 5.37 While the Bank Table 5.4: Agriculture and Natural Resource Management Operations in Pakistan FYO4-03 strategies projected 23-25 Project Name FY Project Cost percent o f lending for (in US$ million) Agricu It ure agriculture and NRM, actual commitments in these sectors Punjab Forest Sector Development 1995 24.9 Baluchistan Community Irrigation and Agriculture 1996 26.7 were only 9 percent, o f Punjab Private Sector Groundwater Development 1997 56.0 which irrigation and drainage National Drainage Program 1998 285.0 NWFP On-Farm Water Management 2001 21.4 accounted for 91 percent.33 A major role for irrigation Natural Resource Management and drainage in the rural Baluchistan Natural Resource Management 1994 14.7 portfolio i s consistent with GEF-Protected Areas Management 2001 10.1 the importance Of irrigated Source: Business Warehouse. agriculture in Pakistan and the cost and complexity o f the infrastructure needed to support that resource. However, the limited scope and size o f operations in other sub-sectors could imply either that the Government and Bank believed the issues were not critical or that they agreed the Bank had no comparative advantage in addressing them. There i s n o record that such a dialogue took place. Agriculture and Natural Resource Outcomes . ~ ~ subsectors 5.38 The sector grew at an annual rate o f 4.5 percent in the 1 9 9 0 ~ Some did very well; from 1990/91 to 2002/03, milk production grew by 80 percent, poultry by 135 percent, and egg production by 145 percent. Notwithstanding, rural poverty in 33 Including the Northern Resource Management Project approved one month before the review period as well as the value o f the GEF project would result in figures o f 83 percent irrigation and drainage, 12 percent NFW and 5 percent forestry. 34 The World Bank Pakistan: Rural Factor Markets (November 19,2004; Report no. 30381-PAK) states that agricultural growth i s overstated due to a change in the base for cattle production. The study estimates that the true agricultural growth rate i s 3.1 percent. 29 Pakistan remained more or less stagnant. A complex set o f interactions i s responsible. Some o f the major issues are: The 2002 Poverty Assessment states that the Gini Coefficient o f land concentration increased from 0.65 in 1990 to 0.78 in 2002. 2 percent o f households o w n more than 40 acres o f land and control 44 percent o f total land area. Almost one-half o f rural households o w n no land. The Poverty Assessment also indicates that frictions in land purchase and leasing markets have contributed to inequality in operated area per household. Due to collateral requirements, land-poor households are mostly excluded from the formal credit market.35 Adoption o f “green revolution” technology supported agricultural growth over the last 40 years. However, much o f the gains from the introduction o f new technologies for major crops had been realized by the early 1990s. Total area cultivated has increased by only 0.18 percent since 1990. Although yields have continued to increase, land productivity could be improved. 36 Over the past decade, partial liberalization o f agriculture input markets, the reduced dominance o f the public sector in agricultural marketing, reduced tariffs, trade liberalization, introduction o f pest resistant varieties o f cotton and improved crop management have contributed to improvements in sector growth. However, SOEs and powerful private marketing associations continue to intervene in the cotton and rice markets. Under-investment in research, development and extension, the poor quality o f inputs and the unreliability o f rural services, particularly for irrigation, also contribute to l o w agricultural yields and inefficient water use. Inadequate rural infrastructure and education and health services and the lack o f effective incentives to encourage more efficient water use similarly constrain production. Diversification into new, higher yielding crops and the use o f efficient irrigation technology have occurred in pockets, but programs to promote and foster an expansion or scaling up o f these trends have not emerged. Water prices are not high enough to cover costs, and recovery rates for assessed tariffs are low. In Punjab, revenues cover less than one-third o f expenditures. 5.39 Irrigation patterns also affect productivity and poverty significantly. Patterns o f unequal land ownership as well as governance problems result in unequal access to water. For example, the World Bank 2002 Poverty Assessment reports that a survey o f six 35 The Government has commented that it does not agree that land-poor households are excluded from formal credit markets, as both the volume o f disbursements to the agriculture sector and the number of borrowers have increased (see Annex 6). However, as noted in the Bank’s 2004 PakistanRural Factor Markets Study, while the volume o f lending for agriculture has increased, the volume of rural lending i s still small, representing only 3.4 percent o f formal lending as of 2002. In addition, only 11 percent o f farmers received formal loans (14 percent of land-owners, but only 2 percent of non-landowningfarmers). Less than 1 percent of formal credit to the sector goes to tenants, with most non- landowning farmers relying on informal credit. 36 For example, wheat production in India’s Punjab province was about 15.5 million tons from 3.4 million hectares in 2000-01, while similar growing conditions in Pakistan’s Punjab province produced 15.4 million tons from 6.2 million hectares. 30 sample villages indicates that i t was routine to bribe irrigation officials to ensure a supply o f water. 5 -40 Water-logging and increased salinity will further limit agricultural growth. Since drainage i s inadequate, increasing amounts o f salt are circulating in the system, progressively poisoning the soil. Water-logging and salinity have depressed major crop yields by an estimated 25-30 percent, and as much as 40-60 percent in Sindh province. Assessing the Bank’s contribution 5.41 The Bank’s rural assistance program, with few exceptions, was poorly managed, designed and executed. It lacked vision, flexibility and responsiveness to the political volatility and institutional weakness that characterized much o f the review period. Given the breadth and seriousness o f sector issues and constraints to agricultural growth and poverty reduction, the lack o f clear strategies and implementation plans was a serious shortcoming. Despite many references in Bank documents to the importance o f increasing agriculture productivity and exports, assuring food security, reducing rural poverty and strengthening natural resource management, each country assistance strategy focused o n a limited Bank agenda: policy dialogue on removing distortions in pricing, trade, and taxes; and work on reforming the irrigation system. While the National Drainage Program (NDP) (1997) had an ambitious reform agenda, the dominance o f irrigation and drainage in the Bank’s program crowded out important policy and lending support for programs and projects that would have addressed the core problems o f rural poverty and environmental degradation. 5.42 The lack o f strategic focus and the poor integration o f agriculture in the country assistance strategies may have resulted from the failure o f economic and sector work to reach consensus on a critical path to address k e y constraints to agriculture growth. For example, the Bank’s 2001 agriculture sector strategy did not attempt to evaluate the Bank’s past agriculture lending programs, nor did it provide a rationale for the projects selected for future agriculture lending. The strategy followed the Bank’s global approach to rural development through improvements in the enabling environment, but issues concerned with improving rural productivity for the poor, generating value-added production activities, SME development and non-farm employment, and achieving reforms in wateddrainage management were lightly treated or absent in the strategy. 5.43 As discussed in paragraph 4.12, the Bank failed to address land inequality, and overlooked the capture o f the benefits derived from Bank support to the sector by rural elites, local and provincial administrations, and de facto agribusiness and market monop01ies.~’ Increasingly severe drainage problems were belatedly addressed under the NDP, but over-ambitious design and the lack o f comprehensive ownership o f this project by the provinces severely limited i t s impact. In the NDP and other projects, the Bank generally overlooked shortcomings in ownership, as well as over-estimated the financial, institutional and absorptive capacities for its activities. The geographical split o f donor interests also resulted in poor coordination when a multi-stakeholder approach was 37 The Bank was aware o f the issues, but did not adjust i t s program design accordingly. 31 critical to the timely achievement o f policy reforms and effective resource and knowledge transfers. 5.44 Support for Natural Resource Management was even more limited. Although the 1992 and 1994 strategies identified environment as a major objective, only two small projects went forward in this area. A planned environmental strategy paper has been repeatedly postponed. Although a number o f projects, such as the NDP, have important environmental effects, there seems to be little support for environmental work, either from the Government, or in the Bank program. Trade, Privatization, and Investment 5.45 The Bank has supported these areas primarily through AAA and conditionality in structural adjustment loans. Trade reform was steady over the period; maximum tariffs have been reduced substantially, from 95 percent in the early 1990s to 25 percent in 2004 (with an average rate o f 14 percent), and the number o f slabs has been reduced to four. Exports as a percent o f GDP have also seen an increase, from about 13 percent during most o f the decade to 15-16 percent in the last two years; one concern, however, i s that at the end o f 2004, the global system o f bilateral quotas in textiles expired. Pakistan i s vulnerable to such shifts because the share o f clothes and textiles in total exports exceeds 70 p e r ~ e n t . ~ ’ 5.46 Pakistan has made progress in privatization, with 144 transactions for Rs 142 billion, from 1991 through November 2004,39 and the recently announced privatization o f KESC was an important step. The Government still holds a number o f large industrial companies (steel, chemicals, fertilizer, textiles, etc.), however, and the sale o f infrastructure companies i s moving slowly. As discussed in the fiscal and infrastructure sectors, these state-owned companies continue to drain fiscal resources.4o 5.47 While annual GDP growth has rebounded to over 5 percent in the last two years, gross capital formation fell through most o f the period, and has only rebounded in the last year, while FDI remains l o w at 1 percent o f GDP.41Thus the sustainability o f growth i s o f some concern. 38 The Government has also expressed concern about the robustness o f Pakistan’s exports, noting “Pakistan’s export base i s not diversified and i s concentrated in a relatively few low value-added products.” (World Bank. 2001, Pakistan; Joint StaffAssessment o f The Interim Poverty Reduction Strategy Paper, pg 22. (Report no. 23 189-PAK). 39 The 1994 PSAL notes that the Government had already completed 84 sales by October 1994, implying that the pace has slowed considerably. 40 See Footnote 12. 41 Figure 5.2 has been calculated using rebased GDP figures. 32 Figure 5.2: Pakistan: Gross fixed capital formation (YOo f GDP) 1993/94-2003/04 120 10 0 .““.“-m I . . ---.I,, , 80 --.. +Pn”ats Fixed c a ,,SI z I I 20 I .. 1994i95 1995148 1996197 1997198 1 9 9 8 ~ 9 iss9moo 2000i01 2001i02 mo2m 200~04 Source: IMF and Government o f Pakistan statistics, and IEG calculations. Summary 5.48 On balance, outcomes for the objective o f sustainable growth were moderately satisfactory. While there was progress in a number o f key areas (banking, trade) and partial and promising reforms in others (power, o i l and gas, highways), the l o w level o f investment, slow progress in privatization and regulatory reform, and the lack o f a rural strategy which addresses hndamental problems, emphasize the fragility o f progress. 6. Governance 6.1 The Bank, as well the Government and other stakeholders, have long noted that governance issues in Pakistan have been impediments to development throughout the period. While governance i s difficult to measure, there i s evidence o f serious problems: In the 1990s, two democratically elected governments were dismissed o n the basis o f corruption charges. As o f 1996, when the World Bank Institute (WBI) began collecting survey data, Pakistan was ranked in the 15th percentile o n surveys o n control o f corruption, the lowest in the South Asia region (see Table 6.1).42 An inquiry in 1996 by the education department o f Sindh discovered that 2,932 schools in rural areas existed only on paper. M a n y o f the buildings had been converted into guesthouses, stables, or storage facilities. The Punjab government discovered similar abuses; it found about 1,600 “ghost schools” which had not operated for years, but the teachers continued to get paid.43 42 Similar results are found in other surveys. In 1994, the International Country Risk Guide (ICRG) gave Pakistan a rating o f 2 out of 6 on corruption (with a high score being less corrupt). In 2004 the rating dropped to 1.5. 43 Operations Evaluation Department (OED). Mainstreaming Anti-Corruption Activities in World Bank Assistance: A Review o f Progress Since 1997 (Report no. 29620). Similar results were found in other Bank reports. The World Bank 2002 Pakistan Poverty Assessment (Report no. 24296-PAK) noted, “In surprise visits, Gazdar (2000) found that one- quarter o f the schools surveyed were not open, there were no teachers present at all in 19 percent of them, and only one teacher was present in 35 percent. Only 38 percent of the schools were classified as “functional”, only a quarter of the schools had electricity, and only half had a latrine.” 33 0 Governance problems complicated Pakistan’s efforts to improve i t s fiscal state. Tax revenues were stalled at about 13.7 percent o f GDP in 1993-94. Fewer than 1.5 m i l l i o n people and entities filed tax returns. 6.2 Governance problems were clearly highly relevant to development in Pakistan; hence the Bank focus on this issue was appropriate. Table 6.1 : WBI Governance Indicators Percentile Rank: Control of Corruption Country 1996 1998 2000 2002 2004 Afghanistan N.A. N.A. 1.I 3.1 3.9 Bangladesh 36.7 43.7 30.4 7.7 10.3 Bhutan N.A. 75.4 86.4 80.9 75.4 India 43.3 60.1 52.7 49.5 47.3 Maldives N.A. 33.9 39.8 56.1 60.6 NeDal 48.0 30.1 40.2 46.9 63.5 Sri ‘Lanka 50.0 57.4 58.7 54.6 52.2 Pakistan 15.3 20.2 27.2 29.9 20.2 South Asia (average) 37.9 45.7 39.4 40.1 37.9 Source: WBI data. The Bank’s Strategy and Program 6.3 In Pakistan, as in many client countries, the Bank has been struggling to develop a governance strategy. In early strategy documents, governance was raised as a problem, but there were no actions or projects planned to address the issue. Governance became a core priority in the FY97 strategy progress report. It did not, however, l a y out any focused lending or AAA for governance; rather, i t described the Bank’s approach as “assisting the effort to improve governance with actions under individual projects, and through sector program agreements such as S A P , Bank-financed projects are promoting transparent procurement processes, including third-party monitoring, and merit-based personnel practices.” There was one project in this area, the Pakistan Improvement o f Financial Reporting and Auditing Project (PIFRA, FY97). 6.4 The 1999 strategy progress report laid out a more comprehensive strategy: (i) a SAL would focus o n improving financial governance, by collecting overdue loans, tax arrears, unpaid electricity and gas bills, while putting in place k e y reforms; (ii)improved governance through project implementation, ( iii) strengthened institutions through projects such as PIFRA and planned tax administration and (iv) policy dialogue and ESW, including the Public Expenditure Review and the C i v i l Service Reform Study. 6.5 The 2002 Country Assistance Strategy laid out a Bank program which would support governance reforms, through supporting the Government’s devolution program, primarily through “analytical work, technical assistance, and policy dialogue to support the implementation o f the devolution strategy, particularly the strengthening o f district 44 The 1998 strategy progress report shows two planned Tax Administration projects. A US$5 million LIL Tax Administrationproject planned for FY99 and a US$40 million Tax Administration I1project planned for FYOl were dropped. Bank staff said this was due to the departure o f the “champion” o f tax administration reform from the Government. 34 level capacity to manage public service delivery”, using the Sindh SAC and a Northwest Frontier Province (NWFP) SAC45as instruments. Governance Outcomes 6.6 Defining progress or designing indicators for governance i s difficult. The early strategy papers did not set any indicators or performance measures. The later strategies and strategy progress reports list either general qualitative outcomes (such as “Improvement in government effectiveness,” “Improvement in overall quality o f financial management”) which are difficult to measure, or intermediate outputs (“Promulgation o f a Procurement Ordinance,” “National Procurement Authority Established”), which do not reflect outcomes. Based on a review o f the strategies, the poverty reduction strategy paper, and other documents produced by the region and the government however, there are some indicative outcomes which can be considered. Perceptions of corruption 6.7 The Government has taken a number o f initiatives, including civil service reform, designed to depoliticize recruitment and promotions. Work o n financial management, financial reporting, accounting, auditing and procurement i s underway. 6.8 One o f the most visible actions o f the Government has been the strengthening o f the National Accountability Bureau (NAB),which has been mandated to investigate and prosecute cases o f corruption. A number o f government officials, politicians, and senior military officers have been sentenced to prison terms, received heavy fines, and been banned from holding public office. Major loan and tax defaulters were also forced to repay their overdue loans and taxes. About US$500 million has been recovered so far. 6.9 Surveys o n Pakistan show an improvement (Table 6.1) on the control o f corruption; but the country lags behind most o f i t s South Asian neighbors and Pakistan is only in the 20thpercentile o f 195 countries covered by WBI. In addition, o f the other five indicators on governance measured by WBI, Pakistan declined between 1996 and 2004, and lags the South Asia region o n the other indicators. (Annex Table 15). 6.10 Trade. Significant progress was made in reforming trade policy (paragraph 3.1 1) which improved the environment for potential foreign investors and trade partners. However, as the 2002 Poverty Assessment noted, through much o f the period “the tariff regime sent mixed signals to producers and exporters. The complexity and non- transparency o f the tariff regime was further exacerbated by ad-hoc exemptions and concessions, which allowed considerable scope for discretion and corruption in customs administration.” 6.1 1 Administrative Corruption. Evidence suggests that governance issues s t i l l have negative effects o n the lives o f most Pakistanis, including service delivery. As cited in the 2004 Pakistan Public Expenditure Management (PEM) report, o f the 57 percent o f 45 The NWFP was planned only in the high-case lending scenario. 35 3,000 survey respondents in Pakistan who recalled using the services o f one or more government agencies (education, health, power, land administration, taxation, police, or the judiciary), almost all reported corruption as part o f the interaction. Fiscal governance 6.12 The Government has taken a number o f actions to improve fiscal governance, including conducting risk-based audits, reducing tax exemption categories, restructuring the Central Board o f Revenue, and improving the technology for revenue collection. However, problems such as tax evasion and unpaid electricity and gas bills still continue. A s discussed in Section 3, tax revenue as a percent o f GDP has not changed appreciably during this period (Table 3.2), nor has the tax base expanded, with only about 1-1.5 m i l l i o n income tax filers.46 Also, although progress has been made in decreasing operating losses o f a number o f public enterprises, losses from the power sector in 2002- 03 continued to be close to one percent o f GDPY4’ and some SOEs continue to drain resources from the Government budget. 6.13 The Bank has acknowledged that governance problems s t i l l are major impediments to fiscal reform. The Bank’s 2004 PEM report noted that “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 the widespread collusion (corruption) between taxpayers and tax officials, a situation that leads to tax evasion and lack o f tax c o m p l i a n ~ e . ” ~ ~ 6.14 O n balance, although progress has been made in governance, the overall outcomes remain unsatisfactory. Assessing the Bank’s Contribution 6.15 The Bank appeared to have difficulty in devising a strategy to support this objective. In the first h a l f o f the review period, the Bank relied primarily o n policy dialogue. Later, the strategy included elements such as: (i) using Bank projects as an example, demonstrating the importance o f governance through the introduction o f good practice procurement, financial management, and other tools; and (ii) pushing for governance through the 1999 SAL, which was focused o n fiscal management. Near the end o f the period, a clearer strategy was developed o f supporting devolution and improving public administration through national and provincial loans, as well as through PIFRA, and providing AAA including the CFAA, advice o n procurement, civil service reform, and tax administration. 6.16 This i s not to say that the Bank did not have some effect. The Bank raised governance issues in every major ESW piece written in the period, highlighting to the 46 World Bank. 2004. Pakistan Public Expenditure Management: Strategy Issues and Reform Agenda (Report no. 25665-PK). 4’ Ibid, p. v. 48 Ibid, p. 25. 36 Government the importance that the international community (including the investor community) placed on this issue. The Government has embraced this view, and has raised i t in i t s o w n documents such as the PRSP. 6.17 Dialogue on fiscal issues such as trade and tax reform has helped, as the trade regime has been greatly simplified, and tax exemptions have been cut back. The Bank’s administrative reform support, through components o f adjustment loans as well as AAA, has helped improve public administration capabilities. 6.18 Fiscal reform, particularly under the SAL, was not effective. IEG’s evaluation o f the SAL noted that the loan had resulted in “only minor sustainable improvements in public sector governance” and that institutional development impact was negligible. Indeed, the fiscal figures discussed in Table 3.2 support the earlier IEG evaluation, in that the S A L does not seem to have made a significant impact o n improving governance. 6.19 One aspect o f the Bank’s program was the visible signs o f corruption identified with some o f the Bank’s largest projects. Although the Bank wanted i t s projects to demonstrate good governance, the perception was at times quite different. Bank documents for S A P I and 11, as well as reviews by the Department Auditor General o f Pakistan (see Annex Table 17), repeatedly cite governance problems in recruitment, site selection, absenteeism, and procurement. The Bank did try to deal with the problems in some projects, such as suspending and then canceling part o f the Baluchistan Primary Education Project; and withholding disbursements in other loans where there were irregularities. However, governance remained a problem in education. 49 6.20 The experience with the IPPs (Box 5.1) adversely affected perceptions in the international business community. When allegations o f corruption in the IPP contracts emerged, the government seriously considered defaulting o n several o f the IPP contracts; eventually the government renegotiated several contracts. The non-competitive nature o f the contract award process, the high price for power at a time when industry costs were falling rapidly (a result o f competition and technical change) and the high development fees included in the contracts should have warranted extra diligence by the Bank. The perceptions o f these problems together with the “near default” and the re-negotiation o f several o f the IPP contracts contributed to a deterioration in Pakistan’s investment climate in the eyes o f the international financial community. 6.21 Other projects also had problems. Political patronage in the selection o f sites or inappropriate use o f funds was cited as a problem in number o f other projects. 49 Bank documents on the SAP report other governance issues such as “vehicles are working out o f the program and should be recovered”, “delivery o f educational materials i s essentially supply driven and does not reflect the needs o f each school as determined by teachers and head teachers.” Other Bank projects had similar problems. One reported “75 percent o f the teachers recruited were unqualified and had been appointed directly by political entities, teacher training programs were cancelled due to lack o f vehicles; the monitoring and quality o f c i v i l works was unacceptably poor and procurement issues persisted.” Another project had the collapse o f part o f a school under construction, which required increased review o f other construction sites; another noted that millions had been spent o n the purchase o f supplemental reading materials, even though n o textbooks were delivered and that science demonstration benches had been delivered, while only h a l f o f the schools received the equipment to be used with the benches. 37 6.22 In summary, the lack o f a clear Bank strategy combined with poor implementation resulted in an ineffective outcome and entailed reputational risk to the Bank. Lessons and Recommendations 6.23 There is a need for a clear strategy, and a dejkedprogram to support it. The Bank has recently been working with the Government on supporting devolution to increase accountability at lower levels o f government. This, coupled with administrative reform, i s expected to increase the efficiency and effectiveness o f Government services. However, since many o f the patronage problems have occurred at the local level, and the Bank i s primarily working with federal and provincial counterparts, the Bank will need to ensure that the strategy and project design address problems at the right level. 6.24 The Bank must spend more time on project design and developing institutional capacity. Since the Bank i s aware that corruption i s an issue, project design must take this into account. Bank projects must ensure that incentives are properly aligned, and must include or develop financial and monitoring systems that are appropriate to the institutional capacity o f the counterpart. This does not mean that the Bank should abrogate responsibility by relying more heavily o n quick-disbursing loans that require minimal documentation o f financial expenditures. Rather, i t means that the Bank will need to spend more time o n helping the Government with institutional development and monitoring systems, including more TA components and close supervision. I t may also mean that the Bank may need to scale back i t s projects, to match the capacity o f i t s counterparts. 7. Conclusions, Lessons and Recommendations Overall Assessment 7.1 During the review period, the Bank embarked o n a series o f strategies to support important and relevant challenges faced by Pakistan. Outcomes were mixed, as changes in governments made it hard to find a committed reform partner who would see a sequence o f reforms through. Development outcomes were also affected by the severe fiscal problems the governments faced, due to poor policies and implementation, as well as exogenous shocks. 7.2 The macroeconomic situation worsened through most o f the review period. Although two near crises were averted, in part with Bank assistance, sustained reforms supported by the Bank were not implemented until the end o f the review period; indeed, assistance packages appeared to do little but increase the country’s debt burden. The macroeconomic situation has greatly improved since 2000, both due to a consistent Government reform program and favorable exogenous events. Debt i s decreasing, reserves have increased to six months, and inflation has dropped to 4.1 percent. However, the situation i s fragile. Total debt i s still high (69 percent o f GDP according to rebased numbers), deficits excluding grants are s t i l l over 4 percent, and many reforms 38 have not yet borne h i t , most noticeably revenue mobilization and improvements in expenditure management. Outcomes in this area are moderately satisfactory. 7.3 The outcomes o f the Bank’s programs inpoverty reduction and social sector development have been unsatisfactory. The PRSP and other documents have recognized that although improvements in some areas have been made, poverty has increased, and the country still lags i t s neighbors and other low-income countries in k e y indicators. The social sectors suffered from a squeeze in resources as the fiscal situation deteriorated during most of the review period, a lack o f a strong strategy to address the roots o f poverty, and poor program implementation. While increased fiscal space and devolution o f service delivery may improve social services, these initiatives have not yet made an impact. Outcomes are judged unsatisfactory. 7.4 The last two years have seen a rebound in growth. This may be due to increased international support, or new inflows o f funds spurred by investor confidence in the Government. Where the Bank focused i t s support, there has been progress in important areas such as trade and banking. Recently, reforms have started in a number o f infrastructure areas, including o i l and gas, highways, and telecom. The Government continues to privatize companies, and has opened up competition in airlines, telecom and the financial sector. However, these sectors are s t i l l in the initial phase o f reforms. Infrastructure has shown limited physical improvements, primarily in the power sector, where investment was handled in a way which damaged perceptions o f the country’s investment climate. Financial sector reform, while strengthening and improving the SBP, improving regulation, and restructuring and partially privatizing four NCBs, s t i l l faces a large reform agenda, including completion o f privatization o f the NCBs and DFIs, improving credit standards and banking performance, and developing a deeper financial market. 7.5 There was little progress in agriculture and natural resource management. While agricultural growth was solid, the effects o n rural poverty were minimal, as the key issues of land inequality; unequal access to water, credit, and other resources; and generating rural employment and increasing rural incomes were not addressed. In addition, part o f the sustainability aspect o f growth i s natural resource management. Environment was initially considered an important area for reform, but this area was generally neglected in the Bank’s program. Given the large dependence o f the economy, particularly the poor, o n agriculture, the effects o f poor water management and increased soil salinity, if unaddressed, may have severe effects on future growth and poverty. Other areas such as industrial and urban pollution have also been ignored. Overall outcomes for the sustainable growth objective are moderately satisfactory. 7.6 The Bank has worked to support the Government in improving governance. However, corruption i s s t i l l endemic in the delivery o f basic services, and the tightening of fiscal discipline has not resulted in improved tax compliance or increased tax revenue mobilization, and SOE fiscal reforms still have not taken hold. Outcomes in this area are judged unsatisfactory. 39 7.7 Although improvements have been made in some areas o f macroeconomic management and growth, reforms in those areas are s t i l l fragile. W h e n taken together with disappointing outcomes in social sector and poverty reduction, governance, agriculture and natural resource management, fixed infrastructure, and revenue mobilization and expenditure management, overall outcomes are moderately unsatisfactory. Lessons 7.8 Throughout the period, the Bank generally did a good j o b o f assessing development needs; objectives were relevant. In many sectors, the Bank’s policy advice was sound, consistent and remained valid over the course o f the review period. The Bank has managed to influence policy over the years, even in areas where lending was curtailed (oil and gas, transport) or where lending programs failed (social sector). 7.9 The Bank also appropriately slowed or stopped lending during some periods o f poor country performance (FY97, FYOO). There were also sectors where the Bank pulled out appropriately, dropping work in areas o f weak reform. 7.10 The Bank demonstrated that i t could leam from its mistakes. The clean-up o f the portfolio and tightening o f quality standards has not only improved performance along some measures (problem projects, project outcome ratings), but it also freed up resources for AAA. 7.1 1 W h i l e the overall analysis o f issues was good, and visions for the program were valid, the Bank did not translate those visions into implementable strategies. The Bank failed to develop realistic strategies in k e y areas such as poverty, rural development, power, and governance. Although the Bank knew that commitment, sustainability, and institutional capacity were limited and that vested interests often overruled good policy, project design failed to take those factors into account. The Bank also failed to stick to i t s o w n plans (for example, moving to high-base case lending in FY94-95 and FY99 when conditions did not justify it), was slow to acknowledge mistakes (following up with S A P I1when S A P clearly was not working) and slow to address repeated problems (e.g., the lack o f commitment and institutional capacity). 7.12 The following are lessons which can be distilled from the earlier sections: Invest in ESK especially before designing loans. The banking sector benefited greatly from having a strong sector review, which set out priorities and recommendations. Poverty, social service delivery, rural development, and governance all failed to have a coherent strategy, and it showed in program design. The power sector assistance would have been helped greatly if a good analysis o f demand, sequencing o f reforms, and fiscal sustainability had been carried out before the lending proceeded. The PSAL was not well-designed, in part because o f the lack o f ESW in some areas. The country team has started spending more o n ESW, and has done some excellent work, especially in recent years (oil and gas strategy, PER, Development Policy Review). 40 Simplijj projects, and reduce scale to match implementation capacity. The program has been characterized by big projects, designed to catalyze sweeping reforms. However, in a country as large and complex as Pakistan, with issues o f commitment and institutional capacity, large projects have proven difficult to implement. These large or broad projects were sometimes less effective than some o f the smaller, more targeted projects. S A P , NDP, and economy-wide adjustment projects all suffered from having uneven levels o f commitment, and differing abilities o f counterparts to deliver. ICRs frequently emphasized the need for simpler project design, covering fewer geographical regions and including fewer implementing agencies. Another example i s that while power projects did deliver the physical components, a phased or pilot approach might have been more cost effective and have averted many o f the problems that arose. This does not mean that the Government and the Bank should not have a sweeping reform program with a long-term vision. I t does means that such a vision should be supported by smaller or more focused projects. This means that reform may take a long time, and may need to be carried out through more projects, but it will be more effective and efficient in the long-run. As discussed in Section 5, the banking sector i s a good example o f focusing on a specific set o f restructuring actions, with the intention o f having follow-up work supported by future loans. D o n 't assume one champion is suficient to ensure commitment. Although many o f the projects may indeed have had a strong champion, or clear commitment from one or more counterparts, almost all projects had multiple counterparts. S A P dealt with 27 different government entities; National Drainage dealt with federal and provincial counterparts; the structural adjustment operations had counterparts in different ministries; the provincial operations (as well as S A P and NDP) rely o n district level staff to carry out implementation. I t was unrealistic to think that commitment would be strong for most, let alone all, o f the parties, especially since there were no incentives that outweighed incentives to engage in patronage or other undesirable actions. Even within smaller projects, commitment at the level o f implementation has been a problem. As part o f gaining commitment, project design should ensure that incentives are aligned. For example, adjustment operations often covered multiple sectors, and not all the affected parties benefited. The Ministry o f Finance may have been strongly committed to the broad reform agenda, but that Ministry was actively involved in negotiations, and received the funding. In contrast, WAPDA, which was supposed to improve i t s tariff structures, increase collections, decrease costs, and dismantle itself, received no funds from the loans. Power tariff reforms are still not complete. Institutional Capacity and Institutional Development matter. Institutional capacity i s almost universally lacking. Even though this issue has been raised continuously in supervision and ICR reports, most projects have little or no institutional development beyond training o f the P M U (as can be seen from the l o w institutional development ratings for the Pakistan portfolio). Even then, P M U staff often leave for better jobs or 41 are rotated.” Also, in many projects, institutional development takes the form o f short two-week or six-week training programs, with little follow-up or on-site experience to bolster the training. The financial sector serves as a good counter- example; Bank projects supported institutional development in the Central Bank and NCBs. Projects will need to have a stronger emphasis o n building institution capacity, not just in a project management unit, but in the government and community at large, Le.: (i)there must be more TA in the program and projects; (ii) the Bank must be realistic about the start-up time for projects as capacity building takes time; (iii) projects may need to be scaled back in size to match the country’s capacity, or phased in on a pilot basis until the capacity i s developed. Institutional capacity also must go beyond project implementation concerns. Many key ministries or agencies do not have adequate resources to do their own analytical or strategic work (an example i s the Ministry o f Water and Power). This i s a problem that will not just be resolved by training, but will need to be addressed by civil service and public administration reform. Adjustment lending should support government ownership and reform. One reason why the B S A L performed better than any o f the other adjustment loans, i s that the reforms were truly “home-grown.” In contrast, the S A L was quickly put together to provide needed balance o f payment support. N o t surprisingly, as soon the funds were disbursed, commitment waned and reforms were not sustained. The experience in Pakistan has also shown that lending i s not always needed to gain “a seat at the table;” the absence o f lending in the o i l and gas sectors and highway sectors did not mean that all dialogue ended. In addition, prior actions for adjustment operations should reflect the reform program being supported by the new operations. Some actions show up as prior actions in more that one project, or reflect progress o f programs which were ongoing independently o f the adjustment loans. Breaking through sector silos. The Bank must be more creative in reaching across sectors to design strategies and programs. The focus in the social sectors on service delivery, with minimal integration o f rural development, infrastructure, and the other sectors needs to be more balanced. Reciprocally, the focus in agriculture on water and irrigation also speaks to a need to broaden staff involvement from other sectors. The Bank must set an example inJiduciary transparency. The Bank must be ready to suspend disbursements and cancel projects if there are problems o f financial accountability. Also, much more work must be done o n project design, including ensuring that incentives are aligned, and that financial management, monitoring and evaluation systems appropriate to the counterpart’s institution capacity are put in place. 50T h i s i s a good indicator o f commitment and sustainability. Some counterparts claimed that staff rotation was to be expected, as i t was part o f the c i v i l service “generalist” policy. However, other Bank and government staff made i t quite clear that when the project was deemed important, the assigned staff would stay in their positions as long as needed. 42 The Bank needs to reactJirmly to macroeconomic slippages and implementation dijjficulties. The information about the inappropriateness o f the S A L and o f the move to “high-base case” lending was in the official documents; the problems o f the S A P were discussed in supervision reports and other internal memorandum; the potential conflicts inherent to the power project should have been clear; and the continued problem o f weak commitment and poor institution capacity i s well documented. Yet, despite many levels o f review, the Bank failed to insist on appropriate actions, programs, and project design. Recommendations 7.13 The Bank should continue with i t s strong support o f analytical work, but it should take i t a step further and translate the analysis into implementable and prioritized actions and programs, taking into account political economy constraints. Priorities are in poverty, rural development, and governance. 7.14 Bank interventions (analytical work, technical assistance, and projects) should have a greater focus on building sustainable institutional capacity. All future projects should have clear institutional development elements or components. In the case o f fast- disbursing loans, institutional development TA or AAA should precede or accompany the loan. 7.15 Projects should be more focused and scaled to fit the capacity o f the implementing agencies. A shift from very broad adjustment loans to more focused loans designed to address a limited number o f sectors or issues would be more appropriate for the current implementing capacity, and would likely lead to more effective loans. Innovative approaches should be initiated by pilot projects before scaling up. 7.16 Improved donor relations, including early consultation o n project design and policy recommendations, and better communications with donors and NGOs, would make the Bank’s program more effective. 43 Annex 1: Structural Adjustment Loans and Credits 1. This annex briefly reviews the national and provincial adjustment loans (the Banking Sector Adjustment Loan i s discussed in the Finance section). These projects have been audited by IEG; the Project Performance Assessment Report has been issued in F Y 0 6 under report number 34 10 1 dated December 19,2005. The Public Sector Adjustment Loadcredit @SALK). On September 23, 1993, t o take advantage o f the reforms supported by the interim Qureshi Government, a Public Sector Adjustment LoanKredit (PSAL/C; US$150 million IBRD/US$lOO m i l l i o n IDA)’ was approved.’ The PSAL/C had 32 agreed actions covering macro, privatization, public finance, social sector, trade, irrigation, and gas. The ICR concluded that the PSAL/C’s had an unfocused “Christmas tree” approach o f vaguely specified, “soff’ and unprioritized measures, and country ownership o f the associated reforms was too narrow. The ICR also criticized the Bank for not having done the necessary analysis and ESW: “One example.. .was a conditionality that recommendations o f the Economy Commission o n public administration reform be implemented, when the Bank did not k n o w precisely what these recommendations were at the time o f negotiations. N o r did the Bank have any preceding E S W on c i v i l service reform issues in Pakistan to draw upon, and n o experts in the field have ever participated in any o f the P S A L preparation/assistance missions. Hence, the Bank was ill-prepared to undertake a dialogue or to assist in the implementation o f public administration measures targeted in the PSAL.” Performance o n agreed policy measures and on actual outcome was mixed, even allowing for the fact that the output performance was hindered by floods and a cotton virus. In the final analysis, the fiscal performance was inadequate, leading to the abandonment o f the ESAF/EFF. I t i s n o t clear why, given the weak performance that the second tranche o f the PSAL/C was released, nor i s the justification for the moderately satisfactory performance rating clear. Given the history o f policy reversals in the country, and given that the loan was being supported by an interim Government which was due to be replaced shortly (the n e w government took office less than a month later) it seems overoptimistic t o have proceeded with this loan, and it also seems that sustainability seemed unlikely (and indeed, the reforms were not sustained). The institutional development impact was rated negligible. The ICR raised the point that unlike previous adjustment loans, n o technical assistance loan was prepared t o accompany this loan. The Structural Adjustment Loan (SAL). In M a y 1998, Pakistan carried out two nuclear tests, As a result, economic sanctions were placed o n Pakistan; investment flows as well as aid decreased sharply, leading to a severe foreign exchange crisis. Reserves fell to US$450 m i l l i o n in early December 1998 (about two weeks worth o f imports). The Government introduced strong foreign exchange controls and other extraordinary measure t o try t o raise revenue and curtail expenditures. The Bank, Fund, and ADB began working with the Government o n a macroeconomic and structural reform program. The C A S noted that financing gaps were estimated t o be US$5.6 billion in FY99, and US$5.1 billion in FYOO, and relied on assumptions ’ The Japanese Overseas Economic Cooperation Fund (OECF) cofinanced $150 million. The IMF later put in a ESAF/EFF for $1.4 billion. The history of this loan was problematic, five previous appraisal missions had been aborted, before the loan was finally appraised and negotiated in 1993. The loan was originally conceived as a low conditionality emergency loan to help Pakistan with the effects of the Gulf crisis. When the crisis ended, the Bank felt i t could no longer move ahead with a low conditionality loan, which caused lingering problems with the Government. 44 o f financing support from the IMF (US$1.5 billion), the W o r l d Bank ( U S $ l billion) and the ADB (US$700), as w e l l as rescheduling o f US$3 billion o f bilateral debt and $1.2 billion o f commercial debt, and voluntary rollover o f institutional deposits in US$1.4 billion and US$1.2 billion f r o m commercial banks and the Central Bank respectively. The 1998 C A S PR reiterated that the volume o f Bank commitments would be linked to: (i) the strength o f Pakistan’s reform effort; (ii) the maintenance o f macroeconomic stability; ( iii) progress in the social sector; and (iv) improvements in the country’s creditworthiness. The P R argued that the new Government (elected March 1997) was truly committed to reform, even though the n e w Prime Minister had not sustained a reform agenda during his previous administrations. So, despite the past history o f policy reversals, weak macro environment, severely deteriorating creditworthiness, and minimal progress in the social sector, the Bank deemed Pakistan to be in the “high-base case” scenario, qualifying i t for about US$750 m i l l i o n in lending.3 Part o f the Bank’s strategy was t o prepare a series o f one-tranche adjustment loans, as part o f a “comprehensive response strategy” in coordination with the IMF and ADB. The Bank began working o n the Structural Adjustment Loan ( S A L ; FY99). The SAL’S primary objectives were to improve fiscal governance, by collecting overdue loans, tax arrears, unpaid electricity and gas bills, while putting in key reforms. The ICR noted the S A L “helped Pakistan stabilize i t s macroeconomics position and strengthen i t s external position.. . overall, the objectives o f the S A L were achieved and all prior actions were achieved before the loan was presented to the Board.” IEG’s evaluation o f this loan noted that it had resulted in “only minor sustainable improvements in public sector governance”, it also observed that compliance o n components dealing with institutional reforms was partial. On that basis i t rated the overall outcome as only moderately satisfactory, the institutional development impact as negligible, but sustainability as likely. Interms o f overall objectives, an immediate crisis was averted, as the IMF also approved US$575 m i l l i o n under the ESAF/EFF arrangement, the ADB lent US$125 million, the Paris Club rescheduled about US3.5 billion, and the London Club rescheduled about US$877 in commercial debt. This support allowed Pakistan to clear i t s external debt arrears and start to rebuild i t s foreign exchange reserves. However, by spring 1999, “there were concerns that the government’s commitment to the reform program was faltering. In particular, resource mobilization and governance improvements (ed. note: i.e, the exact areas o f reform that were targeted by the SAL) remained weak.”4 By September 1999, the government’s failure t o meet two conditions for IMF tranche release under the ESAF, l e d the IMF to formally suspend the ESAF program. I n 1998/99 total debt to GDP was 93 percent; by 2000/0 1, it had increased t o 109.2 percent of GDP.’ Structural Adjustment Credit Iand I1 (SAC and SAC 11). The Musharraf Government took power in October 1999. I t inherited a very vulnerable macroeconomic situation, and, given the weak reform performance by previous governments, difficult relationships with lenders and donors. But by November 2000, the Government reached a new Stand-by Arrangement with the IMF, and negotiated another Paris Debt Restructuring (US$1.8 billion) in January 2001. Although growth was lower than planned (3.1 percent versus a planned 4.5 percent), overall performance under the Stand-By was satisfactory. The Government also was in discussions with the Bank o n starting a new lending program, and in M a y 2001, the Bank approved the US$350 In fact, IBRD and IDA exceeded that planned amount; commitments in FY98 totaled US$808 million. World Bank. 2001. Pakistan Country Assistance Strategy Progress Report (Report no. 22219-PAK). IMF estimates. 45 m i l l i o n SAC, which covered a similar range o f reforms addressed by the previous adjustment loans, including revenue mobilization, improved expenditure management, banking reform, power reform, gas reform. The SAC, like the SAL, was a one-tranche operation. It was designed to be one in a series of programmatic loans, and the design reflects not only prior actions, but lays out anticipated follow-up actions In June 2002, a full C A S was issued. By this time, circumstances had changed. The new Government had established a fairly good reform record over the previous two years. But an exogenous factor also had a major impact. The events o f September 11,2001 l e d to a dramatic change in Pakistan’s finances. As described by the CAS, in addition to significant direct donor support (the U.S.alone contributed U S 6 0 0 million), the external financing package supported by the IMF included a PGRF for US$1.25 billion over FY 02-04, and US$3.5 billion in debt relief from the Paris Club. In addition, remittances increased sharply from US$l -1billion in 2000/0 1 to US$4.2 billion in 2002/03. Debt service pressures were also eased by the drop in international interest rates. While pressing o n the many o f the same general themes and policy issues as other adjustment loans (fiscal improvement, governance, removing distortions that impede growth, improving service delivery), SAC I1had a large focus o n improving government administration. S A C I and SAC I1s t i l l had an extremely broad reform agenda, perhaps, as was noted with the S A L , due to a desire to bulk up the l i s t o f prior actions. Indeed, some prior actions were listed in both S A C I and 11, some were continuations o f ongoing programs, and some reflected w o r k that was being supported by other Bank loans. Looking at the sequence o f reforms covered by the two credits, the most progress seem t o be in some o f the areas most directly under control o f the federal government, including c i v i l service reform, public accounting and auditing, trade reform, and budget allocations (to the social sector).6 Areas such as power sector reform, gas pricing, the sale o f K E S C and Pakistan Telecommunication Limited (PTCL), ie., where the implementation falls under agencies or ministries that are not receiving funding f r o m the loan, have not proceeded as well. In a country as large and complex as Palustan, it may be more appropriate to have smaller, more focused adjustment loans, rather than economy-wide loans. The Northwest Frontier Province SAC and the Sindh SAC were planned as part o f a sequence o f provincial level loans t o support the Government’s program o f devolution. Although progress in financial management has been made, there are still issues o f service at the district level. In addition, since the provinces “own” tax base i s narrow (less than 1 percent o f the Gross Provincial Development Product), the m a i n source (about 80 percent) o f provincial revenues are based o n distributions f r o m the federal government. Hence, the provinces are dependent o n the federal government for revenues, and o n district administration to implement programs effectively. For the purpose o f reviewing the effectiveness o f these projects, progress in areas where there has extensive Bank support (Le., the banking sector) through other loans, or where the country assistance strategy documentation lists accomplishments under ongoing programs that were planned independently o f these loans (such as the Lady Health Workers or the establishment o f the Micro-Credit Bank) are not considered accomplishments o f these projects. 46 Annex 2: Statistical Tables Table 1 Pakistan at a Glance Table 2 K e y Economic and Social Indicators, 1990-2003 Table 3 Consolidated Government Budget Table 4 Pakistan: Presidents and Prime Ministers (September 1978-Present) Table 5 Pakistan: IMF Programs (FY93-03) Table 6 Project Rating by Sector Table 7 Portfolio o f World Bank Lending for Pakistan Table 8 Pakistan: Millennium Development Goals Table 9 Pakistan’s Social Indicators in International Perspective Table 10 Achieving the Millennium Development Goals-Pakistan’s Recent Progress Table 11 Aid Assistance to Pakistan Table 12 W o r l d Bank Senior Management for Pakistan Table 13 Pakistan-Analytical & Advisory Reports (including ESW) Table 14 Pakistan-Natural Gas Prices Table 15 Pakistan-Governance Indicators Compared with South Asia Average Table 16 Pakistan-Private Power Projects 47 Annex Table 1 Pakistan at a glance 12/5/05 POVERTY and SOCIAL South LOW- Development dlamond' Pakistan Asia income 2004 I Population, mid-year (millions) 152.1 1,448 2,338 GNI per capita (Atlas method, US$) 600 590 510 Life expectancy GNI (Atlas method, US$ billions) 90.7 860 1.184 Average annual growth, 1998-04 Population (%) 2.4 1.7 18 30 2.1 21 GNi Gross Labor force (%) per primary Most recent estimate (latest year available, 1998-04) capita enrollment Poverty (% of population below national poverfy line) 33 Urban population (% of total population) 34 28 31 Life expectancy at birth (years) 64 63 58 Infant mortality (per 1,000 live bllths) 74 66 79 Child malnutrition (% of children under 5) 35 48 44 Access to improved water source Access to an improved water source (% ofpopulation) 90 84 75 Literacy (% of population age 75+) 47 59 61 Gross primary enrollment (% ofschool-age population) 68 97 94 ---Pakistan Male 80 105 101 Low-income group Female 57 92 88 KEY ECONOMIC RATIOS and LONG-TERM TRENDS 1984 1994 2003 2004 Economic ratios. GDP (US$ billions) 31.2 51.9 82.3 96 1 Gross capital formation/GDP 18.3 19.5 16.9 17 3 Exports of goods and serviceslGDP 11.1 16.3 16.9 16 0 Trade Gross domestic savings/GDP 6.7 16.8 17.5 18 4 Gross national savings/GDP 25.9 22.2 23.1 23 0 Current account baiance/GDP -2.2 -3.2 5.1 20 Domestic Capital Interest papentslGDP 1.8 1.9 1.o 08 savlngs formation Total debffGDP 39.3 52.8 43.6 37 1 Total debt service/exports 19.5 35.0 16.7 154 Present value of debtlGDP 36.1 Present value of debtlexports 164.4 Indebtedness 1984-94 1994-04 2003 2004 2004-08 (average annual growth) GDP 5.5 3.4 5.0 6.4 6.7 -Pakistan GDP per capita 2.9 0.9 2.5 3.9 Exports of goods and services 10.8 4.7 28.4 -1.5 7.9 STRUCTURE of the ECONOMY 1984 1994 2003 2004 Growth of capital and GDP (%) (% of GDP) 10 Agriculture 27.9 25.6 23.6 22.3 Industry 22.7 24.3 23.0 24.9 5 Manufacturing 16.1 16.8 16.2 17.6 0 Services 49.4 50.2 53.4 52.7 5 Household final consumption expenditure 81.2 71.1 73.6 73.3 .IO General gogt final consumption expenditure 12.1 12.1 8.9 8.4 -GCF -O-GDP Imports of goods and services 22.6 19.0 16.3 14.9 I 1984-94 1994-04 2003 2004 Growth of exports and Imports (Oh) (average annual growth) Agriculture 4.4 3.0 4.1 2.2 45 T Industry 6.7 3.8 4.7 12.0 30 Manufacturing 6.2 4.9 6.9 14.1 15 Services 5.4 4.1 5.2 6.0 0 Household final consumption expenditure 4.2 3.2 0.5 8.2 General goVt final consumption expenditure 5.1 2.2 7.2 2.1 -1 5 Gross capital formation 4.6 0.8 6.3 -3.2 imports of goods and services 3.2 0.1 11.2 -8.6 Note: 2004 data are preliminary estimates. * The diamonds show four key indicators in the country (in bold) compared with its Income-group average. If data are missing, the diamond will be incomolete. 48 1984 1984 200% 2004 i 11 3 31 41 $7 12 a 44 713 172 J 175 1s 1 -2 4 09 11 47 *I4 .I a I 1984 1894 2003 1004 5 885 10 880 If 40 fib5 634 1 1,4b2 1884 1984 2003 3,449 E.3#0 -6813 4,304 1,88G e?? 82 2 11 346 57 6 1984 1894 2003 2004 1 1.247 81 18 57.1 1 -OB4 152 56 354 612 702 0 33# -394 186 1 210 357 421 -212 4% 347 131 BT 458 243 e c 4 ? c